Whole-genome sequencing of mixed OCD–schizophrenia pedigrees characterizes shared and divergent rare-variant architectures

Whole-genome sequencing of mixed OCD–schizophrenia pedigrees characterizes shared and divergent rare-variant architectures | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Whole-genome sequencing of mixed OCD–schizophrenia pedigrees characterizes shared and divergent rare-variant architectures Zeping Xiao, Miaohan Deng, Yuan Wang, Weidi Wang, Yihang Bao, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8943594/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Psychiatric disorders often co-occur and share liability, yet why distinct diagnoses arise under a largely shared familial genetic background remains poorly understood. We assembled 21 Han Chinese mixed pedigrees co-ascertained for obsessive–compulsive disorder (OCD) and schizophrenia (SCZ) (21 OCD, 21 SCZ, and 38 unaffected first-degree relatives) and performed deep (60×) whole-genome sequencing. Leveraging within-family structure as an internal control, we integrated rare coding and regulatory variation across co-segregating and disorder-biased loci and evaluated diagnostic separation using stratified group cross-validation to prevent relatedness-driven leakage. Mutation-burden models distinguished OCD from controls (AUC 0.839), SCZ from controls (AUC 0.749), and affected individuals from controls (AUC 0.792), whereas OCD–SCZ discrimination remained modest (AUC 0.631), consistent with partial genetic sharing. Burden decomposition suggested that coding-region signals accounted for most of the discriminative performance, while non-coding burden provided limited incremental contribution under current annotations. Developmental network mapping nominated temporally stratified prenatal-to-postnatal modules, including a late-pregnancy angiogenesis-related module shared across disorders and SCZ-biased astrocyte/calcium-related programs. Together, these results illustrate how mixed-pedigree WGS can help disentangle convergent versus divergent rare-variant architectures across OCD and SCZ and provide family-grounded, interpretable signatures for future disorder-specific stratification. Biological sciences/Genetics Biological sciences/Molecular biology Biological sciences/Neuroscience Health sciences/Diseases/Psychiatric disorders/Schizophrenia Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Psychiatric diagnoses are largely defined by descriptive symptom constellations rather than etiology, and substantial symptom overlap and comorbidity are common across disorders ( 1 ). Although recent nosologies increasingly incorporate dimensional elements, such as in DSM-5 and ICD-11 ( 2 , 3 ), a central barrier to precision psychiatry remains: we still lack a mechanistic understanding of why clinically distinct disorders share symptoms, risk factors, and familial liability. Convergent genetic and biological evidence suggests that many psychiatric conditions partially share molecular underpinnings, yet the degree of overlap, and critically, the mechanisms that bias individuals toward one diagnostic outcome versus another, remain incompletely characterized ( 1 ). Defining such convergent versus divergent biology is not only essential for refining disease models but also provides a rational basis for improving early stratification and targeted intervention in high-burden families ( 1 ). Motivated by this gap, we initiated the Co mplex F amily Project involving Fi rst-Degree R elatives (CoFFiR), which recruits mixed-pedigree families affected by multiple severe mental disorders to generate family-grounded clinical and genetic evidence. In this first stage, we focus on genetic architecture using deep whole-genome sequencing as a foundation for disentangling shared liability from diagnosis-biased signals. Obsessive–compulsive disorder (OCD) and schizophrenia (SCZ) offer a clinically relevant and biologically informative setting to study such within-family diagnostic divergence. SCZ affects ~ 1% of the population ( 4 ) and OCD affects ~ 2–3% ( 5 ), and OCD/obsessive–compulsive symptoms are frequently observed in individuals with SCZ ( 6 ). Reported prevalence estimates suggest that OCD occurs in ~ 12.1–13.6% of SCZ patients ( 7 ), and obsessive–compulsive symptoms in SCZ can reach ~ 30.7% ( 8 ). Both disorders are complex and heterogeneous, influenced by genetic and environmental factors, with heritability estimates of ~ 80% for SCZ ( 9 ) and 26–45% for OCD ( 10 ). Longitudinal and family studies further support bidirectional aggregation: individuals with OCD show an elevated risk of subsequently developing SCZ (IRR up to 6.9; 11), and 7.8% of OCD patients developed SCZ over an 11-year follow-up ( 12 ). Moreover, multigenerational family studies indicate increased schizophrenia risk among relatives of OCD probands, decreasing with genetic distance ( 13 ). Together, these observations support partially shared genetic components while highlighting an unresolved question of direct relevance to CoFFiR: under a largely shared familial background, what genetic signals are shared across OCD and SCZ, and what signals bias risk toward one diagnostic outcome versus the other? Genetic studies have made substantial progress in mapping shared and disorder-specific risk. Genome-wide association studies (GWAS) have identified many common-risk loci across psychiatric disorders, yet the gap between SNP-based and epidemiological heritability underscores “missing heritability” and motivates attention to rarer variants ( 10 ). Next-generation sequencing (NGS) has therefore become central to psychiatric genetics, particularly for identifying rare coding variation and probing regulatory mechanisms. For OCD, rare-variant studies remain comparatively limited: a large family-based whole-exome sequencing study implicated de novo mutations and prioritized candidate genes ( 14 ), and our prior work using family-based whole-genome sequencing supported an excess burden of rare de novo mutations and suggested potential involvement of regulatory elements in non-coding regions relevant to OCD etiology ( 15 ). In contrast, sequencing studies in SCZ are more extensive, spanning candidate gene discovery, mechanistic insights, cross-ancestry analyses, and broader characterization of genomic architecture ( 16 – 23 ). Despite these advances, cross-disorder sequencing designs that can directly interrogate within-family diagnostic divergence remain scarce. A key reason is methodological: conventional case–control designs typically require very large sample sizes to overcome heterogeneity and background noise, and they are inherently limited in resolving within-family divergence when distinct diagnoses arise within the same pedigree. Family-based designs offer complementary strengths by enriching rare risk alleles, reducing confounding from population stratification, and leveraging shared background to highlight diagnosis-biased signals ( 24 ). Yet such approaches remain underutilized in cross-disorder psychiatric genetics, particularly in designs that co-ascertain multiple disorders within the same families ( 24 ). To our knowledge, few whole-genome sequencing (WGS) studies have leveraged mixed pedigrees co-ascertained for obsessive–compulsive disorder (OCD) and schizophrenia (SCZ) to interrogate within-family diagnostic divergence. Here, we leverage such mixed pedigrees in which OCD and SCZ co-occur to address a question that case–control designs are inherently underpowered to resolve: why distinct psychiatric diagnoses emerge under a largely shared familial genetic background. We performed deep (60×) WGS in 21 OCD–SCZ complex families and integrated rare coding and regulatory variation to (i) prioritize co-segregating and disorder-biased burden patterns, (ii) evaluate disorder separation under a family-aware validation scheme, and (iii) map convergent versus divergent signals onto spatiotemporal neurodevelopmental modules to nominate candidate vulnerability windows and cell-type programs. Methods Subjects Twenty-one families of Chinese Han population with OCD, SCZ patients and healthy controls were recruited for WGS. Each family included one OCD patient, one SCZ patient that met DSM- IV criteria and at least one unaffected family member, resulting in a cohort total of 21 SCZ patients, 21 OCD patients and 38 unaffected controls. All volunteers within the same family were either direct relatives or collateral relatives within three generations. Strict quality control was applied during the process of sample collection. All OCD patients were diagnosed by senior attending psychiatrists or chief psychiatrists. Patients were excluded if they met DSM-IV criteria for any disorders other than OCD or SCZ. The International Neuropsychiatric Interview (M.I.N.I.) was used to screen for DSM-IV Axis I psychiatric diagnoses. Socio-demographic and additional clinical information were collected using a semi-structured interview design by our team. The Yale-Brown Obsessive-compulsive Scale (Y-BOCS) was used to assess OCD symptom severity( 25 ). The Positive and Negative Syndrome Scale (PANSS) was used to assess schizophrenia symptom severity( 26 ). The Hamilton Anxiety Scale (HAMA)( 27 ) and Hamilton Depression Scale (HAMD) ( 28 )were used to assess mood status, such as anxiety and depressive symptoms, respectively. All assessments were conducted by raters trained for this study. This study was approved by the Ethics Committee of Shanghai Mental Health Center, School of Medicine, Shanghai Jiao Tong University. Written informed consent was obtained from each participant. DNA Sequencing DNA was extracted from whole blood. DNA quality was assessed using gel electrophoresis and its concentration was measured using the Qubit Fluorometer. DNBSEQ library (paired-end 150 bp) was constructed and sequencing at 60× coverage was performed using the BGI DNBSEQ platform. Sequencing-derived raw image files were processed by DNBSEQ basecalling software for base-calling with default parameters, and the sequence data for each individual were generated as paired-end reads. Curation of sequencing data and variant calling SOAPnuke v2.1.0 was used for quality control of the raw sequencing data( 29 ). The clean data of each sample were mapped to GRCh37 to obtain an initial alignment file in BAM format using Burrows-Wheeler Aligner (BWA) v0.7.17( 30 ). Samtools v1.3.1 was used to sort and index the SAM files( 31 ). Genome Analysis ToolKit (GATK) v4.1.4.1 was used in the following variant calling. MarkDuplicates was used to mark the duplicate reads, which were ignored in downstream analysis. BaseRecalibrator and ApplyBQSR were used to correct the base quality values. HaplotypeCaller was used to simultaneously detect SNPs and insertion-deletions (InDels). The sample-level variant calling results were stored in gVCF files and subsequent genotyping was conducted using GenotypeGVCFs. After single nucleotide variations (SNVs) and InDels were selected separately for downstream pipelines via SelectVariants, Variant Quality Score Recalibration (VQSR) was used to filter out false mutations. Only the variants indicated as "PASS" were considered credible variant sets and included in following analysis. SNV/Indel annotation We used ANNOVAR (version 2023-08-30) to annotate the SNVs( 32 ). Allele frequency (AF) was annotated based on 1000 Genomes (1000G) and gnomAD (version 2.1.1). Only the variants with the minor allele frequency (MAF) ≤ 0.005 in databases including 1000G_EAS, 1000G_ALL, gnomAD_EAS and gnomAD_ALL were filtered into the following analyses. The functional categories of the SNPs in coding regions were predicted based on RefSeq annotation of hg19 (version 2020-08-17). We annotated the known SNVs in coding regions with ClinVar (released 2024-04-26). Besides, we further assessed the pathogenicity of variants via Franklin ( https://franklin.genoox.com ), pLI (probability of loss-of-function intolerance) score( 33 ), combined annotation dependent depletion (CADD) v1.7( 34 ) as well as MPC( 35 ), MutationTaster( 36 ), LRT( 37 ) and phastCons100way_vertebrate( 38 ) scores included in dbNSFP v4.1( 39 , 40 ). The putative pathogenic variants were manually inspected employing visualization of aligned reads using Integrative Genomics Viewer (IGV)( 41 ). For non-coding region variant annotation, the experiment-derived regulatory features were acquired from PsychEncode( 42 ), which provided the atlas of some regulatory elements of mental disorders. We also employed chromatin states annotations including Roadmap core 15-state model (epigenome data from Brain Dorsolateral Prefrontal Cortex)( 43 ) and Sei model( 44 ), which combined several epigenomic marks in a spatial context, to accurately capture the potential epigenomic function of the variant position. Structural variation /copy number variation annotation BreakDancer v1.4.5( 45 ) was used to detect structural variations (SVs) using default parameter. CNVnator v0.3.2 ( 46 )was used to detect copy number variations (CNVs). The process used standard parameters and settings, and a window length of 100 bp is selected. Ensembl Variant Effect Predictor (VEP) ( 47 ) and annotSV ( 48 , 49 ) were used to annotate the SV/CNV results( 47 ). Co-segregation and disease-specific variant/gene lists In protein-coding regions, the filtered variants were divided into co-segregation variants and disease (OCD or SCZ) specific variants: co-segregation variants were defined as variants detected in both OCD and SCZ patients but not present in any of the controls in the same family; disease-specific variants were defined as variants detected in the patient with the specific disease (OCD or SCZ) but not present in any other family member in the same family. These variants were mapped to the corresponding genes based on RefSeq to imply putative pools of candidate genes in our data. In non-coding regions, to focus on the candidate genes more efficiently, co-segregation variants were enriched by ruling out the variants present in any one of the controls in the whole cohort. Correspondingly, disease-specific variants were enriched by ruling out the variants present in any unaffected individual or individual with the other disorder in the whole cohort. Enriched gene lists were mapped from the enriched variant list. Virtual panel genes/regions from publicly available data We introduced three disease-related virtual panels from publicly available data. To focus on the protein-coding genes detected in brain, we introduced ‘protein-coding genes expressed in brain’ gene list obtained from The Human Protein Atlas (HPA) ( 50 )(n = 15331). The etiology virtual panel including common psychiatric-disorder-related etiological gene sets available from publications: ( 1 ) RBFOX target genes( 51 ); ( 2 ) Nervous developmental disorder genes( 52 ); ( 3 ) LOEUF_TOP20% genes( 53 ); ( 4 ) hPSD genes( 54 ); ( 5 ) FMRP target genes( 55 ); ( 6 ) genes related to chromatin modifiers( 56 ); ( 7 ) CHD8 target genes( 57 ); ( 8 ) reported ASD candidate genes( 58 ). The GWAS_OCD/SCZ virtual panel was calculated from database: Variants with p < 1e-5 from OCD GWAS ( 59 ) and variants with p < 5e-8 from SCZ GWAS( 60 – 64 ) were selected, resulted in 780 variants. Additionally, Clinvar SNVs related to OCD/SCZ, which was in putative risk OCD/SCZ regions nominated by PsychENCODE ( http://resource.psychencode.org/ ) and OMIM ( https://www. omim.org/ ), were added(n = 2358). The combined SNV lists including 3138 SNVs were expanded to all SNVs within the LD block via Haploreg V4.2(65) (r2 > 0.8), resulted in 16235 SNVs. The SNVs were next filtered by DNase I hypersensitive (DHS) peaks in GTEx (including brain-sourced data: cerebellar_cortex, dorsolateral_prefrontal_cortex, frontal_cortex, globus_pallidus, head_of_caudate_nucleus, posterior_cingulate_gyrus and putamen; immune-related data: B_cell, CD14_positive_monocyte, CD4_positive_alpha-beta_T_cell, common_myeloid_progenitor_CD34_positive, naive_B_cell, naive_thymus_derived_CD4-positive_alpha-beta_T_cell, natural_killer_cell, T_cell and T_helper_cell; neuron-sourced data: choroid plexus epithelial cell, smooth muscle cell of the brain, astrocyte of the cerebellum, brain pericyte primary cell, SK-N-DZ cell line, M059J cell line, brain microvascular endothelial cell, BE2C cell line, astrocyte of the hippocampus, Daoy cell line, bipolar neuron in vitro differentiated cells, astrocyte of the spinal cord, and astrocyte primary cell) (n = 5713) and intersected with eQTL locus from GTEx and PsychENCODE. The final eQTL variant list (n = 5682) were mapped to genes based on Refseq(n = 486). Overrepresentation analysis was performed to determine if the overlap between two gene sets was significantly higher than might occur by chance. This analysis was done using the “enrichment” function of the R package clusterProfiler( 66 , 67 ). Coding variants classification and analysis For SNVs, variants were divided into different types for categorized analyses: ( 1 ) Loss of function variants (LOFs), any variants that introduced a stop codon, a shift of the open reading frame or a change at a predicted splice site. ( 2 ) Missense variants (MISs), any single-nucleotide variants changing the amino acid. ( 3 ) synonymous variants, any variants that resulted in none amino acid change. The functional consequences were determined by RefSeq. We further divided LOFs and MISs into following groups according to the pathogenicity score: ( 1 ) LOFs: LofA (pLI > 0.9), LofB (pLI > = 0.5), LofC (pLI = 25 or MPC > = 2), MisB (15 < CADD < 25 or 1 < = MPC < 2), MisC (CADD < 15 and MPC < 1). Among all the groups, LofA and MisA variants were defined as high-impact SNVs, and those located on autosomes were filtered into the following statistical analyses. For Indels, as CADD annotation was inadequate, variants were additionally annotated with VEP. Those indels with CADD > 25 and those with unknown CADD score but with Ensembl Variant Effect Predictor (VEP) annotation of “HIGH” were categorized as high-impact. For SV/CNVs, variants were annotated with AnnotSV and high-impact SVs were defined as: ( 1 ) With AF < 0.005; ( 2 ) Overlap with protein coding regions; ( 3 ) 1) Ranked by AnootSV as 4–5 in pathogenicity (Likely pathogenic/ Pathogenic) OR 2) a. Ranked by AnnotSV as 3 in pathogenicity (VUS) AND b. not overlapped with any of the known benign region ((po_) B_gain, (po_) B_loss, (po_) B_ins, (po_) B_inv). High-impact SV/CNVs were presented to observe the distribution of important variants. Non-coding variants analysis We focus the variants in non-coding regions located in the non-coding co-segregation and disease-specific genes. Fisher’s exact tests were performed to detect the enrichment of regulatory elements on autosomes between groups. P threshold was set as 0.05. FDR threshold was set at 0.2. Association analysis of rare variants Concerning the difficulty of association analysis of rare variants in small-sized family samples( 68 , 69 ), we combined three common gene-based burden analysis approaches, including Fisher’s test, sequence kernel association test (SKAT-O) ( 70 ) and aggregated Cauchy association test-omnibus(ACAT-O) test( 71 ) to detect all the associations contained in the data on autosomes. Only genes that passed Bonferroni correction in any of the test or genes passed all three tests with raw p < 0.05 were reported. In protein coding regions, region-based burden test was implemented in LOFs and MisA variants. In non-coding regions, region-based burden test was performed in possible functional variants, i.e. non-intergenic and non-intronic variants, based on specific gene regions including co-segregation and disease-specific genes. Protein-protein interaction effects evaluation While protein–protein interactions (PPI) influence cellular functions and biological activities in a fundamental way, we evaluated the variant influence on PPI with MIPPI( 72 ). The high-impact missense variants were collected and grouped into SCZ-specific, OCD-specific and control-specific groups. The grouped variants were then mapped to PPI partners of homo sapiens from BioGrid database v4.4.234( 73 ) to predict the category of mutation effect. Selection of high-confidence genes To construct the most affected network in specific diagnosis group, we selected and explicitly reported the high-confidence variants according to the following criteria: For SNVs: ( 1 ) Being the co-segregated variants or disease-specific variants; ( 2 ) 1) Being LofA of MisA; 2) Met 2–3 following criteria: a. predicted by LRT as “U” or “D”; b. predicted by MutationTaster as “A” or “D”; c. phastCons100wat_vertebrate_score > = 0.990. The high-confidence SNVs identified were included in the following network analyses. For SVs: ( 1 ) Being the co-segregated SVs or disease-specific SVs; ( 2 ) With AF < 0.005; ( 3 ) Overlap with protein coding regions; ( 4 ) Ranked by AnootSV ( 48 , 49 ) as 4–5 in pathogenicity (Likely pathogenic/ Pathogenic). Network construction The high-confidence genes derived from high-confidence SNVs and listed in the ‘protein coding genes expressed in brain’ genes from HPA atlas were selected to build the PPI network. PPI networks were constructed by STRING v12.0 ( https://cn.string-db.org/ ) database followed by analyses via Cytoscape 3.10.2( 74 ). Only the direct interactions between seed nodes were presented. Further visualization was done by R package ‘ggnetwork’. Functional enrichment To have a more comprehensively landscape of the target genes as far as possible, we performed functional enrichment analyses by Metascape v3.5.20240101( 75 ), which can incorporate results from several databases at a time and make comparison between two groups. P threshold was set as 0.05 and FDR threshold was set at 0.05. The enriched terms were manually clustered according to biological relatedness, enabling an overall landscape of the biological functions. Single Cell Expression Analysis For specific hub genes, we obtained the single cell expression data from The Allen Brain Cell Atlas ( 76 ), which allowed comprehensive subclusters combining cell types and anatomies, and we performed visualization based on its suggested pipeline. For cluster of target genes, we employed cell type enrichment analysis via R ‘Expression Weighted Cell Type Enrichment’ (EWCE) package( 77 ). The newly published single-nucleus RNA sequencing (snRNA-seq) data from psychencode was used to perform EWCE( 78 ). Weighted correlation network analysis and Developmental Trajectory Analyses As the frontal cortex played a critical role in both SCZ and OCD( 79 ), we made Weighted correlation network analysis (WGCNA) to distinguish different expression patterns between diseases using data from frontal cortex. Expression matrix of developmental transcriptome of frontal cortex from BrainSpan ( https://www.brainspan.org ) was used for WGCNA ( 80 ). Genes with mean RPKM (Reads Per Kilobase of transcript per Million mapped reads) value < 1 across all the developmental stages were discarded. After the detection of co-expression modules, we tested whether specific gene sets were enriched in any of the modules by Fisher’s exact test. The enriched modules were used for the enrichment analysis via Metascape. We also divided the developmental stages as early-pregnancy (0–8 post-conception weeks, pcw), mid-pregnancy (8–26 pcw), late-pregnancy (26–37 pcw) and after-birth to detect the expression and functional characteristics in each developmental stage. Random Forest classifier To classify individuals based on their mutation profiles, we implemented a Random Forest classifier with default hyperparameters. Model performance was rigorously evaluated using a 5-fold stratified group cross-validation protocol, where family identifiers served as the grouping variable to prevent data leakage between folds, ensuring a realistic estimate of the model's ability to generalize to new families. This stratification maintained a consistent case-control ratio across all splits. Within each fold, we assessed performance using the Area Under the Receiver Operating Characteristic Curve (AUC) and Matthews Correlation Coefficient (MCC). For each classification task, we visualized overall performance by plotting the macro-average True Positive Rate (TPR) with a standard deviation confidence band and calculated a comprehensive micro-AUC score from aggregated out-of-fold (OOF) predictions. To interpret the model and identify key predictive features, we employed SHAP (SHapley Additive exPlanations). This analysis was conducted strictly within the cross-validation loop, calculating SHAP values exclusively on the held-out test sets to avoid bias from the training data. The global importance of each feature was determined by its mean absolute SHAP value, and we identified the most consistently impactful predictors by selecting the top 10 features that maintained a high importance ranking across all five cross-validation folds. Results Family-based design enables within-family risk stratification In the first stage of the CoFFiR project (CoFFiR-Ⅰ), we conducted WGS research on OCD-SCZ complex- families, which included 21 SCZ patients, 21 OCD patients and 38 unaffected family members (Figure S1 Methods). As shown in Table 1, PANSS and Y-BOCS scores were different between OCD, SCZ, and control groups (p < 0.001). Then we performed WGS at 60× coverage on the SCZ-OCD complex-families. The average coverage of the sequencing was 99.57%, with 97.5% of bases covered above 20×. 4,116,317 SNVs and 1,026,835 Indels were detected and only those identified as “PASS” by VQSR were considered for downstream analysis. As part of sample quality control, we performed principal components analysis to assess the population stratification of our cohort. The results showed that the samples were primarily stratified based on families and regions, rather than diagnoses (Fig. 2a). For protein-coding regions, variants were stringently filtered based on the criteria detailed in the Methods section. Specifically, single nucleotide variants (SNVs) were filtered by allele frequency (AF) < 0.005 in GnomAD_ALL/EAS and 1000G_ALL/EAS, resulting in 368, 382, 379 variants per sample in OCD, SCZ and control group, respectively(Fig. 2b). Next, we classified the LOF variants and missense variants by pLI score and MPC/CADD, separately (see Methods and Fig. 2b). LofAs and MisAs were designated as high-impact SNVs. Besides, InDels were grouped by CADD phred score and VEP impact annotation. SVs/CNVs were assessed by annotSV (see Methods). The distribution of high-impact variants of all variant types over the whole genome was presented in Fig. 2d. The variants were validated manually using Integrative Genomics Viewer (IGV) (Figure S2 - S3 ). Besides, for non-coding variants, we used the filter criteria described in the Methods and detected 37796, 40628 and 39255 SNVs per sample for OCD, SCZ, and control group, respectively (Fig. 2f). To verify our variant selection and expand the scope of understanding towards the genetic contributions, we curated a SCZ/OCD gene panel by using publicly available GWAS summary statistics (see Methods, Figure S4 a, Table S1 ). We performed enrichment analysis based on KEGG, Gene Ontology (GO), along with selected gene sets, to identify the implications of highlighted pathways and biological functions. These genes were widely distributed in terms related to metabolism, synapse and neurotransmitter, and were overrepresented in gene sets included RBFOX-targets, NDD genes, TOP 20% LOEUF genes, FMRP-targets and ASD-related candidate genes (Figure S4 b, c, d, Table S2 -4). An integrated matrix of high-impact SNV/Indel variants was mapped to the panel genes (Fig. 2e). There were 2818 high-impact variants in 1887 genes detected in our cases (Table S5 ) but only 86 variants in 64 genes were mapped to the panel (Figure S4 a, Fig. 2e). This suggests a large number of potentially novel candidate genes in our data. Design of the disease-specific genetic model Coding-variant burden differs across within-family diagnostic groups We used Fisher's exact test to observe the count distribution of variants between groups (Fig. 2c, Table S6 , details in supplementary note 1). While the divergent results implied the differences between diseases in complex families, we performed segregation analysis on our filtered variants to narrow the range of suspected genes and variants. We identified the variants of co-segregated, OCD-specific, SCZ-specific and control-specific categories within each family (see Methods). We annotated the SNVs and Indels with ClinVar database and the results were presented in Table 2 and Table S7 . We made comprehensive annotation for all kinds of variants, only SNVs in high-impact groups (LofA and MisA) were included in the following analyses (see Methods). Regulatory annotations show disorder-biased enrichment in candidate regions To obtain a comprehensive view of the whole genome, non-coding variants were included in our analyses. We evaluated the distribution of functional elements in non-coding variants using experimentally derived functional element atlas from PsychEncode and the Refseq functional annotation. Chromatin state models including Roadmap core 15 models and Sei were employed to provide a comprehensive landscape (see Methods, Fig. 2a, Figure S5 , details in supplementary note 2). Only non-intronic and non-intergenic variants on autosomes were included. In the primary comparison across the whole genome, we didn’t observe a significant difference between groups (Fig. 2b, Table S8 ). In order to limit the noise in unrelated genomic regions, we focused on comparison within different candidate gene panels including the non-coding co-segregation panel (CSncs)(Table S9 ), non-coding OCD-specific panels (OSncs) (Table S10 ) and non-coding SCZ-specific panels (SSncs) (Table S11 ) (see Methods). From our results, while both OCD and SCZ manifested significant alteration in non-coding regions, OCD showed greater deficit in non-coding regulatory elements within putative risk regions than SCZ, especially in promoter regions and transcription-active regions. Family-aware prediction using burden features To assess the predictive power of mutation burden in selected genomic regions for mental disorders, we developed a series of Random Forest classifiers (Fig. 3a, Table S12 ). The models were trained on mutation counts from each volunteer (Figure S6 ). We employed a StratifiedGroupKFold cross-validation strategy, using family identifiers as groups to prevent data leakage and ensure model generalizability. Our primary model, utilizing features from both coding and non-coding regions, demonstrated strong predictive capability in distinguishing affected individuals from healthy controls. The highest performance was achieved in the OCD vs. Control classification, with a mean AUC of 0.839 ± 0.104. The model also effectively distinguished the combined disease cohort from controls (Disease vs. Control: AUC = 0.792 ± 0.076) and SCZ from controls (SCZ vs. Controls: AUC = 0.749 ± 0.141). As expected, differentiating between the two mental disorders proved more challenging (SCZ vs. OCD: AUC = 0.631 ± 0.135), though the performance remained above chance. To investigate the relative contribution of different genomic regions, we conducted an ablation analysis by training models exclusively on mutations within either coding or non-coding regions. The model using only coding-region features (six features) performed comparably to the full model. For instance, in the SCZ vs. Control classification, its performance was slightly higher (AUC = 0.765 ± 0.097). Conversely, the model trained solely on non-coding features (15 features) showed a marked decrease in predictive accuracy across all classification tasks. For example, its ability to distinguish the general disease cohort from controls dropped significantly (AUC = 0.698 ± 0.102), and performance for SCZ vs. Control fell to an AUC of 0.629 ± 0.176. These results strongly suggest that mutations within the coding regions of the selected loci are the primary drivers of the observed classification performance, while the non-coding mutation burden in these specific areas shows limited incremental value beyond coding burden under current regulatory annotations for these conditions. To further interpret the feature contributions of the full model, we employed SHAP (SHapley Additive exPlanations). The analysis consistently identified MISs and LofC as the most impactful predictors across disease-versus-control comparisons. Specifically, a higher burden of these mutation types strongly increased the model's output towards a disease classification, thereby pinpointing the key genetic markers driving the predictions (Fig. 3b). Based on our results, we hypothesized that differences in the genetic architecture among OCD, SCZ, and control groups could be leveraged to evaluate individual disease risk. To test this hypothesis, we calculated a comprehensive genetic score for OCD and SCZ using the detected variants for each participant within the same family (Table S13 ). The comprehensive score was divided into two components: the OCD score and the SCZ score. To calculate the comprehensive score, we extracted all variants present in each individual from the following sets firstl: 1) Coding regions: variants from co-segregation (Set-Ccds), OCD-specific (Set-Ocds), co-segregation, and SCZ-specific (Set-Scds) panels. 2) Non-coding regions: variants from co-segregation (CSncs), OCD-specific (OSncs), co-segregation, and SCZ-specific (SSncs) panels. Then, each variant was mapped to its corresponding gene. Gene scores were assigned based on established gene-level metrics: the pLI score for coding regions( 33 ), and the average non-coding GERP (ncGERP) score for non-coding regions( 81 ). Variant scores were assigned to each variant according to the variant type. LofA-C and MisA-C in coding regions were assigned scores of 2, 1.5, 1.25, 1, 0.9, or 0.75, whereas non-coding variants annotated as 1_TssA, 4_Tx, 7_Enh, and 15_Quies were assigned scores of 0.5, 0.5, 0.5, and − 0.5, respectively (Table S13 ). For each gene, the gene–variant score was calculated as the product of the gene score and the sum of its variant scores. The OCD comprehensive score for an individual was obtained by summing the gene–variant scores from co-segregation and OCD-specific regions (both coding and non-coding). Similarly, the SCZ comprehensive score was derived from co-segregation and SCZ-specific regions (Table S13 ). This novel scoring framework effectively distinguished the three groups within each family (Fig. 5c, Table 3). A paired Wilcoxon test revealed that OCD patients had significantly higher weights for the OCD score compared to the SCZ score (median OCD score = 40.13 vs. median SCZ score = 31.79, p = 0.009). In contrast, SCZ patients showed the opposite trend (median OCD score = 28.12 vs. median SCZ score = 51.79, p < 0.001). No significant difference was observed in the control group (median OCD score = 25.60 vs. median SCZ score = 25.28, p = 0.122). Biological implication of the genetic model Intra-Family Co-segregation Pinpoints Shared and Disorder-Specific Genetic Hubs As a supplement to the genetic model, we sought to infer etiological relevance from familial genetic data. Regarding the polygenic genomic architecture of OCD and SCZ, we speculate that these diseases may be associated with impairments in a particular functional network. MIPPI( 72 ) was employed and the results suggested a potential larger-scale protein-protein network (PPI) network disturbance in the etiology in OCD and SCZ groups compared with control(Fig. 5a, Table S14 , details in supplementary note 3). Then, we further investigated the shared and specific variants or functional networks of OCD and SCZ patients. Therefore, genes mapped from different high-impact variant groups were pooled to form the gene sets including co-segregated (HI_CScds), OCD-specific (HI_OScds), and SCZ-specific (HI_SScds) gene panels within the protein-coding regions (see Methods, Fig. 3b). We performed an enrichment analysis by Metascape( 75 ) on the HI_OScds and HI_SScds genes, to specify the probable characteristics of functional impairment related to each specific disease (Fig. 5c, Table S15 -16, details in Supplementary note 4). Through the divergent genetic landscape obtained from pathway analyses, we supposed that there were substantial differences etiology reflected by the intra-family phenotypic heterogeneity. To determine the most affected functional PPI network in each group, we selected the high-confidence variants with the strongest evidence of pathogenicity, for each group separately (see Methods, Table 4, Table S17 ). These high-confidence SNVs from the co-segregation, OCD-specific, and SCZ-specific groups were selected to construct the STRING PPI network separately. Results showed that these networks manifested significantly more interactions than expected (co-segregation network: p = 0.047; SCZ-specific network: p = 0.0132; OCD-specific: p = 0.0404). Hub genes of each network was identified to elucidate the core function of different networks (Table S18 ). As a result, microtubule actin crosslinking factor 1 ( MACF1 ) emerged as the hub of the co-segregation network, ryanodine receptor 2 ( RYR2 ) as the hub of SCZ-specific network and filamin A ( FLNA ) as the hub of OCD-specific network (Fig. 5d). The expression data of the hub genes were obtained from Allen Brain atlas that MACF1 plays a substantial role in actin regulation, microtubule arrangement and stabilization, widely distributed in every subcluster of cells (Fig. 5e); RYR2 is a component of a calcium channel that is widely expressed in neurons (Fig. 5e) and FLNA is another actin-binding protein playing a role in cell junctions during the organ development, specifically expressed in certain types of neuron and cells from neurovascular system (Fig. 5e). The core function implied by hub genes was also supported by the network enrichment that the SCZ-specific network was enriched in Uniprot terms “Calmodulin-binding” (q = 0.0064),” Calcium transport” (q = 0.0223) and “Ligand-gated ion channel” (q = 0.0336), while OCD-specific network was enriched in Uniprot terms “actin-binding” (q = 0.0279). To sum up, the PPI networks revealed that the shared genetic basis underlying SCZ and OCD may be the fundamental microtube-actin dynamics. Compared to OCD, SCZ showed more prominent dysfunction in calcium-related pathways. Then, we combined fisher test, SKATO test and ACATO test on high-impact SNVs to perform the region-based rare-variant association test, separately (see Methods). None of the genes reached the Bonferroni threshold in any of the comparisons. However, gene MACF1 was marginally significant in all three tests when we comparing the OCD and control groups. (Table 5). Region-based burden test in non-coding regions revealed the converged network with coding region We replicated the region-based burden test in the non-intronic and non-intergenic variants in non-coding regions. Several genes met the threshold of OR > 1 and p < 0.05 in all the three tests, including CALM2, PLCG2, KCNH5, MMS22L, MCF2L and DNAL1 (Table S19 ). Notably, we found that these genes, including CALM2, PLCG2, MMS22L, were involved in the main cluster of STRING network established by high-confidence variants in coding-region (Figure S7 ). Interestingly, none of the variants in these genes were detected in coding-regions in our cohort. Therefore, our results may indicate convergence in the same functional network of coding and non-coding variants. None of the genes reached the Bonferroni’s threshold (p < 7.66e-6) in all the three test and only gene HLA-DRB5 pass the Bonferroni’s correction in Fisher’s exact test. Temporal Co-expression Networks Map Developmental Divergence and Genetic Convergence To comprehensively explore the developmental trajectory divergence of different diseases, we separately built unsupervised co-expression networks for SCZ and OCD via WGCNA. For each network, we enrolled both coding and non-coding variants, aiming to explore the convergence and divergence in the development of SCZ and OCD. For coding variants, the high-impact variants in co-segregation (Set-Ccds) and OCD-specific (Set-Ocds) panels, as well as co-segregation and SCZ-specific (Set-Scds) genes were used as input representing coding variants for OCD network (Set-HIcds_OCD) and SCZ network (Set-HIcds_SCZ), separately. For non-coding variants, we excluded the intronic and intergenic variants in different disease group firstl. These remaining variants were filtered by the brain-expressed protein-list in the HPA database, and then non-coding panels including CSncs and OSncs, as well as CSncs and SSncs, separately. These selected variants were pooled to form the variant sets in OCD group and in SCZ group. According to the above results of functional element distributions, we further selected the variants which were predicted to be active transcription (simultaneously annotated as “4_Tx” by roadmap and as “TN1 Transcription” by Sei) (set Set-Trans_OSncs and Set-Trans_SSncs) and promoter (simultaneously annotated as “1_TssA” or “11_BivFlnk” by roadmap and as “P Promoter” by Sei) (Set-Pro_OSncs and Set-Pro_SSncs). Finally, we combined the mapped genes from regions of non-coding active transcription and non-coding promoter as input for the OCD (Set-ncsOCD) and SCZ (Set-ncs_SCZncs) group, separately. The aforementioned coding and non-coding gene sets were integrated to construct comprehensive networks for OCD and SCZ. Furthermore, we plotted the average expression curves of the detected modules from the 8th week pcw to 40 years of age, with developmental stages classified according to gestational periods. This allowed us to observe the characteristics of gene expression across different periods. The results demonstrated that there were both similar and distinct expression patterns of gene clusters between groups (Fig. 6a). We prioritized the key modules by performing Fisher’s exact test to observe the distribution of gene sets (Fig. 6b, Table S20 - S21 ) and making EWCE analysis to observe differences of cell type enrichment (Fig. 6c, Table S22 - S23 ). Analysis details were described in Supplementary note 5 and Table S20 -. These results suggested that late-pregnancy modules captured shared OCD–SCZ coding variants associated with angiogenesis. SCZ-specific gene modules were enriched for calcium-related pathways, with peak expression occurring after late pregnancy and continuing into later developmental stages. Modules associated with synaptic plasticity, which comprised both coding and non-coding variants, showed consistent expression across all postnatal developmental stages. Together, these results suggest that OCD and SCZ are shaped by distinct combinations of genomic elements operating across key developmental windows. Discussion In this pilot stage of the CoFFiR project, we performed deep whole-genome sequencing in mixed OCD–SCZ pedigrees to probe convergent and divergent rare-variant signals under a largely shared familial background. By combining within-family co-segregation with burden-based modeling and developmental network mapping, we provide a complementary view to large case–control studies for a question they are often underpowered to resolve: why different psychiatric diagnoses emerge within the same pedigrees. Our study covered both coding and non-coding regions, providing a family-grounded complement to GWAS by interrogating rare coding and regulatory variation with reduced background heterogeneity. After gathering basic functional evidence from burden tests across different genomic regions, we developed a disease-specific genetic risk model. It integrates both coding and non-coding rare variants, finding that combining variants from coding and non-coding regions improved the distinguishing power, highlighting potential implications for future family-based risk stratification, pending replication and prospective validation. The model was supported by pathway analyses across multiple hierarchical levels. Core disease-specific PPI networks revealed that SCZ-specific variants were enriched in pathways related to calmodulin binding, calcium ion transport, and ligand-gated ion channels, while OCD-specific variants were enriched in actin-binding functions. The fact that these signals emerged from different individuals within the same families supports their biological relevance. By introducing BrainSpan developmental data, we mapped the temporal expression patterns of these variants, which served as our attempt to provide a probable biological explanation for the divergence and convergence in the genomic architecture of OCD and SCZ. Genomic Architecture Integrating Stratified Coding Variants and Functional Elements in Non-coding Regions In our study, we incorporated non-coding variants to comprehensively identify the risk factors underlying the etiology of OCD and SCZ. Significant enrichment in strong transcription state and notable trends of enrichment in promoter state were observed in non-coding regions of OCD. The predicted strong transcription state refers to chromatin regions with high enrichment of H3K36me3( 43 , 44 ) and the predicted promoter state indicated enrichment of H3K4me3 (and H3K27me3, which is present in a bivalent state)( 43 , 44 ). Our findings supported previous publications suggesting the involvement of chromatin regulation in SCZ and OCD pathogenesis, especially OCD( 14 , 15 , 82 ). Motivated by the observation that OCD, SCZ, and control groups differ in their proportions of non-coding elements, we proposed that different combinations of functional elements within shared genomic regions could imply different clinical outcomes. This hypothesis was supported by the results of the Random Forest classifiers. By combining coding and non-coding variants, the prediction model achieved the highest accuracy in almost all comparison groups. In addition to underscoring the significance of rare variants in coding regions again, the results highlighted the importance of non-coding variants in OCD and suggested that OCD may involve a greater proportion of non-coding disturbances. Due to the complexity of the non-coding elements and their interactions, the precise genomic architecture still requires further investigation. Despite the modest sample size in our data, the implication from non-coding regions still provided insights into OCD pathogenesis discovery. Different types of variants coordinated in shared and distinct functional networks of OCD and SCZ Both OCD and SCZ are complex mental disorders with poorly characterized etiology. Theories about both conditions include key disturbances in neurotransmitters and disruptions in neurodevelopment( 5 , 83 ). With accumulating evidence from functional genomic research, SCZ is increasingly considered to have a neurodevelopmental origin. Our data demonstrated distinct characteristics of protein dysfunctions and PPI disruptions between OCD and SCZ in a cross-sectional perspective. MIPPI analysis among high-impact variants suggested significant PPI disturbance of OCD and SCZ, indicating probable protein dysfunction in these disorders. Enrichment analysis supported the neurodevelopmental hypothesis of OCD and SCZ as the high-impact variants in co-segregation panels were centered by neurogenesis. SCZ-specific and OCD-specific proteins showed different trends of enrichment, that SCZ-specific genes tended to be enriched in terms related to calcium ion transport, one of the classical pathways in SCZ pathogenicity( 84 ), while OCD-specific genes tended to be enriched in terms related to cell junction, which was also reported but in scattered studies( 85 – 87 ). The results were further validated by the PPI networks formed by selected high-confidence genes and their hub genes. Our results suggest that a bias toward cell junctions may be a feature of the molecular etiology of OCD. In the following rare variants association test combining three different algorithms, MACF1 was the only gene that reached marginal significance in all three tests. It also served as the hub gene in our co-segregated network. Given the modest sample size in our cohort, we think the effect size brought by MACF1 was noteworthy. MACF1 encodes a large and complex protein that is a crucial regulator of the cytoskeleton in neurons and glial cells, and therefore plays prominent roles in multiple cellular functions in brain cells e.g. cell proliferation, migration and neurite development. It has been identified as a candidate gene in psychoses including SCZ( 88 – 90 ) and bipolar disorder( 91 ). In SCZ, it may contribute to cognitive deficits by disrupting the intracellular transport and cytoskeletal stability at synapse( 90 ). It is worth noticing that it has also been reported to be associated with a family suffering from inherited SCZ and schizoaffective disorder( 92 ). We are the first to report high-impact MACF1 variants to be significantly enriched in OCD. So far, combined evidence suggested that it may be a shared candidate gene in OCD, SCZ and bipolar disorder. Since the detailed mechanism of its pathological function remains largely unknown, further studies are in need to deepen our understanding of it on a mechanistic or disease-specific level. Given the limited power of rare variant association analysis in the family-based design, false negatives may be present in our results. Future studies with larger family sample sizes and enhanced family-based algorithms will be essential. We also performed multiple association tests in non-coding regions. Several genes (CALM2, PLCG2, MMS22L, KCNH5, MCF2L, DNAL1) were identified as key results. Among the results, CALM2, PLCG2, MMS22L were found to be integral to the main PPI network constructed by coding variants. CALM2 was mostly studied in cardiac arrhythmias( 93 ). A few studies have linked altered CALM2 expression with SCZ( 93 – 96 ). Additionally, abnormal DNA methylation of CALM2 has been observed in postnatal malnourished mice( 97 ), indicating that adversity may affect CALM2 through epigenomic mechanisms. PLCG2 acts as a downstream molecule in BDNF/TrkB pathway, and was mostly been reported to contribute to Alzheimer’s disease by affecting synaptic function( 98 ). Its mutations have been detected in psychoses( 99 ), autism( 98 ) and pediatric acute onset neuropsychiatric syndrome( 100 ). Cross-species brain transcriptomic analyses suggested that it exhibited conserved gene expression patterns in affective disorders( 101 ). Given their significant enrichment in the whole disease group, and their central position in the network, our results suggest that CALM2 and PLCG2 are candidate genes for both schizophrenia and OCD. Concerning the complex effects of non-coding variants, more accurate annotations and quantified predictions of non-coding variants, as well as more powerful gene-based algorithms are in need to further explain the effects of these cryptic variants in a broader range of brain disease. WGCNA sketched the comprehensive temporal landscape of variants in OCD and SCZ Early-and-mid-pregnancy stage captured epigenomic changes and divergence between disease In order to provide biological explanation for the disease-specific score, we incorporated both coding and non-coding variants into WGCNA to gain insights into the dynamic changes of high-impact variants in OCD and SCZ. In early-pregnancy i.e., embryonic stage, we detected overlapping and similarly enriched terms pertaining to cellular reproduction, epigenomic modification and so on between OCD and SCZ. These results may be consistent with previous publications implying chromatin modification ( 14 , 15 , 82 ) and isoform-level dysregulation( 42 , 102 ) in OCD and SCZ. According to the cellular and molecular landscape of developing human brain, mid-pregnancy is the stage characterized by the onset of neurogenesis and neuronal migration( 103 ). Results in this stage suggests potential differences in noncoding regions between OCD and SCZ. During the mid-pregnancy, we detected enrichment trends of non-coding variants, especially in SCZ. This result was in accordance with previous studies that the fetal brain development was shown to harbor greater enrichment for SCZ GWAS signal( 104 ). Significant enrichment was detected in terms related to epigenetic and post-transcription regulation, as well as terms related to cytoskeleton and cell structure in OCD. Broader disturbed functions were detected in SCZ. Noticeably, while a broad class of inhibitory neurons in SCZ exhibited trends of enrichment in this stage compared to OCD, Lamp5 and Pvalb emerged as the most prominent. Pvalb, the largest class of inhibitory neurons, provide feedforward and feedback synaptic inhibition to a variety of neurons( 105 ), which allows the elaboration on a broader, more coordinated, and more delicate firing patterns( 106 ). A strong body of evidence has linked Pvalb to SCZ ( 106 ). Meta-analysis of post-mortem studies suggested a significant reduction of Pvalb cell density in the prefrontal cortical regions of SCZ( 107 ). Animal experiments demonstrated that disruption of Pvalb development, particularly before puberty, may exacerbate the neuropathological consequence of adversity e.g. social isolation, in subsequent developmental stage( 108 ), and thus leading to various pathophysiological manifestations in SCZ patients( 108 , 109 ). The trajectory in our data implied the idea that this disruption may root in the genomic alterations, with primary effects occurring as early as in the mid-pregnancy, leading to delayed but prolonged synaptic effects on these late-maturing GABAergic neurons( 110 ). In contrast to Pvalb, Lamp5 neurons are less well-studied. They are thought to play a role in GABAergic neurotransmission within specific neuronal subpopulations, affecting short-term synaptic plasticity( 111 ). Recent evidence from snRNA-seq( 112 ) and ALLEN BRAIN MAP( 113 ) suggested Lamp5 and Pvalb neurons among the most altered neurons types in SCZ post-mortem samples, although information about Lamp5 remains limited. Angiogenesis in late-pregnancy was the core affected functional pathway shared by diseases The late-pregnancy and early postnatal stage are considered as the onset of astrogliogenesis, oligodendrogenesis and synaptogenesis( 103 ). Our results indicated this stage as the core period, with co-segregation coding variants significantly enriched in both the OCD and SCZ groups. The enrichment terms strongly supported the alteration of angiogenesis and blood-brain-barrier (BBB) integrity as a shared genetic etiology of OCD and SCZ. Many studies supported the brain microvascular endothelial cell and BBB dysfunction in SCZ. However, it remained unsolved whether or not the deficits were primary or compensatory( 114 , 115 ). Alterations in adhesion and angiogenesis molecules in OCD have been reported in a few studies, but the evidence remains limited( 86 , 116 – 119 ). The brain angiogenesis and BBB formation are thought to start at the embryonic stage. BBB is formed on the basis of neurovascular units (NVU) ( 120 ). We detected the enrichment of several components in an NVU in this stage. The neurovascular system plays a critical role selectively and dynamically regulating the molecular transport between the bloodstream and brain parenchyma ( 115 ). This regulation is essential to maintain the integrity of the brain parenchyma and preserve subtle neurochemical homeostasis from external environment( 121 ). Deficits in NVU may lead to disturbed neurotransmitter transport, disrupted neuronal signaling, or impaired neurotoxin elimination e.g. cytokines, which may ultimately contribute to the clinical manifestations of SCZ and OCD( 121 ). Recent literature reported the abnormalities in gene expression, microstructure and paracellular permeability in endothelial cells of SCZ patient-derived 3D cerebral organoids( 120 ). These findings support the early origins of endothelial alteration and BBB dysfunction in SCZ( 120 ). It also provided explanation for the disturbance of immune-inflammatory system observed in SCZ( 120 ). Impaired BBB in OCD was most noticed in the context of pediatric acute-onset neuropsychiatric syndrome (PANS), and pediatric autoimmune neuropsychiatric disorders associated with streptococcal infection (PANDAS). PANS/PANDAS sufferers usually present with OCS, and the prevailing etiology model includes the breaching of BBB by the infection-triggered antibodies( 122 , 123 ). It was reported that the first-degree relatives of PANS/PANDAS sufferers have up to 10-fold risks of developing OCD, which suggested significant genetic overlap( 122 ). Autoantibodies directed against basal ganglia was also detected in the cerebrospinal fluid (CSF) of drug-naïve OCD patients ( 124 ). Multifaceted evidence in our study indicated the NVU-based BBB formation as primary pathologies shared by SCZ and OCD. In fact, there existed a further aspect underscoring the close structural and developmental link between blood vessels and nerves due to the tight coupling between neuronal activity and blood flow( 125 ). Abnormalities in vascular network and blood dynamics may exert more prolonged impact on synaptic plasticity and neuronal metabolism in SCZ and OCD pathology( 114 , 125 ). Further elucidation of neurovascular system may offer new insights into the abnormal brain development and functioning underlying OCD and SCZ. In schizophrenia (SCZ), various human studies have indicated neurovascular system abnormalities. Postmortem analyses have shown structural changes in brain capillaries, such as reduced density, smaller diameters, endothelial degeneration, and extracellular matrix buildup, particularly in the prefrontal and visual cortices ( 126 ). Imaging studies, including 7T-MRI, have revealed altered volumes in small cerebral arteries, potentially explaining SCZ-related gray matter loss ( 126 ). Diffusion-prepared arterial spin labeling (DP-ASL) MRI also showed reduced neurovascular water exchange in SCZ-spectrum patients ( 127 ). Cerebrospinal fluid studies found elevated blood-brain barrier (BBB) permeability markers like Qalb and S100β in SCZ, suggesting compromised BBB integrity( 126 ). In contrast, only few studies have identified altered levels of related molecules, such as angiopoietin (ANG), cadherin-5 (CDH5), and ICAM-3, in peripheral blood of OCD patients ( 86 , 119 ), hinting at possible neurovascular dysregulation. Our study found significantly more OCD-associated genes than SCZ-associated genes in the relevant module, highlighting this pathway's potential importance in OCD pathophysiology. SCZ-specific module highlighted the critical role of astrocytes and calcium signaling Compared with OCD, there also existed another black module enriched with SCZ-specific coding variants. Cell types including astrocytes and pericytes, as well as terms related to ion calcium signaling pathways, RHOa GTPase cycle, clathrin-mediated endocytosis and transport of small molecules were enriched. It reflected the affected pathways detected in the SCZ-specific PPI network. These results marked it as another astrocyte-centric module, with sustained high expression during the whole postanal stage. Emerging genetic, pathological and serological evidence increasingly highlighted the critical role of astrocytes in SCZ ( 128 ). Astrocytes are not only structural components of BBB, but also play key roles in glutamate metabolism, maintenance of synaptic networks, and regulation of synaptic plasticity ( 121 , 129 ). Growing genetic, pathological, and serological evidence highlights their critical involvement in SCZ ( 128 ). Postmortem studies from SCZ patients have revealed altered astrocyte density and morphology, along with dysregulated expression of markers like aquaporin-4 (AQP-4). Animal models have shown that similar changes in the prefrontal cortex can lead to SCZ-related cognitive impairments ( 130 ). Transcriptomic analyses of developmental data show that proliferation-regulating genes are downregulated during astrocyte maturation, while genes involved in neurotransmitter transport (glutamate, GABA), connexins (Cx30, Cx43), and potassium channels (Kir4.1) are upregulated ( 130 ). These upregulated genes support synaptic homeostasis and astrocyte-synapse interactions, suggesting that astrocyte maturation parallels increased synaptic activity. In contrast, SCZ astrocytes may exhibit immature-like phenotypes with reduced expression of Kir channels and Cx30, which align with SCZ features in transgenic mouse models ( 130 ). Therefore, the astrocyte and small molecule transport, especially calcium-related, abnormalities observed in this module may reflect SCZ-specific functional changes, distinct from OCD, likely involving disrupted astrocyte development and related synaptic functions. Since astrocytes mediate interactions between the BBB and neurons, disturbances in the blood–astrocyte–neuron axis may also occur. Further investigation into astrocyte development and their role in neurotransmission could help explain core SCZ symptoms and guide new treatment strategies ( 130 , 131 ). Astrocytes serve as the interface between BBB and the neurons, and mediate interactions between vascular dynamics and neurotransmitter system. They are not only a structural component of BBB, but also a key partitioner in a variety of neurobiological functions including glutamate metabolism, synaptic network structuring and maintenance, as well as synaptic plasticity regulation( 121 , 129 ). Our results implied a more prominent alteration in molecular transportation and potential disruption in blood-glia-neuro crosstalk via astrocytes and pericytes as featured functional impairments in SCZ compared with OCD. Shared and distinct neuronal types were found disturbed in modules about synaptic plasticity In addition to the variant-type-enriched modules, we also identified the brown modules in OCD and SCZ to contain shared and distinct components. The brown modules in both groups were characterized by the increasing expression starting from synaptogenesis time in late-pregnancy( 103 )and maintained a consistently high expression since late infancy. Functional enrichment revealed significantly enriched terms related to synaptic plasticity. These findings are highly consistent with previous GWAS and are further supported by prior evidence implicating specific allele sets—particularly genes expressed in neurons within defined brain regions and gene sets intolerant to PTVs or involved in synaptic function—in shared genetic and biological mechanisms underlying both common and rare variants in SCZ ( 22 ). Additionally, enrichment in terms related to metabolism and membrane localization points to processes involved in maintaining neuronal system homeostasis. While not central to earlier GWAS results, these processes align with emerging insights from metabolomics studies and transcriptome-wide association studies (TWAS) in SCZ( 132 – 134 ), suggesting they may serve as phenotypic modifiers or contribute to disease expression. In the broader context of cross-disorder research, our findings provide new evidence supporting shared physiological pathways across neuropsychiatric conditions, offering further insight into the overlapping molecular underpinnings of these complex disorders. Whereas genes in both OCD and SCZ groups were enriched in types of neurons, of particular interest, some genes in OCD group were specifically enriched in Pvalb inhibitory neurons, while some genes in SCZ group were specifically enriched in oligodendrocytes. Pvalb was scarcely studied in OCD. As we have detected enrichment of Pvalb in SCZ during mid-pregnancy, the enrichment of Pvalb in OCD suggested that the potential Pvalb deficits in OCD may be more closely related to the process of neurotransmitter and synaptic function. Pvalb was reported in OCD in striatum, where continuous optogenetic stimulation of Pvalb in the striatal areas connecting to lateral orbital-frontal-cortex can reduce compulsive-like grooming behaviors in Sapap3-KO mice( 135 ). Knocking down Pvalb in pre-frontal cortex was found to reduce cognitive flexibility( 136 , 137 ), one of the core cognitive impairments in both SCZ( 136 ) and OCD( 138 , 139 ). The neuropathological effects of Pvalb alteration may vary across regions( 140 ) and developmental stages( 141 ). As evidence was limited, more research investigating the role of Pvalb in the whole classical cortico-striato-thalamo-cortical circuits in different developmental stages was needed. Besides Pvalb in OCD, enrichment of oligodendrocytes was found in SCZ in our study. Recent research has increasingly supported the role of oligodendrocytes in SCZ, with accumulating evidence from imaging, histopathological, and molecular studies( 142 ). Abundant evidence came from functional neuroimaging studies suggested SCZ as a disorder of dysconnectivity in brain( 143 ). Oligodendrocytes are cells that form myelin sheaths around multiple axons, facilitating rapid conduction of electrical signals and maintaining axonal integrity at the cellular level. Oligodendrogenesis was thought to begin at late-pregnancy and followed by myelination throughout the whole childhood( 103 ). However, few studies have investigated the specific impact of oligodendrocytes in SCZ-related induced pluripotent stem cell (iPSC) models ( 142 ). The exact mechanism by which oligodendrocytes affect SCZ remains uncertain. Possible impacts include altered synaptic plasticity, axonal degeneration, conduction velocity, neuronal circuitry or neuronal signaling( 144 ). Our results revealed the enrichment of oligodendrocytes as featured impaired cell type of SCZ, highlighting the need for further investigation into the complex intercellular interactions of oligodendrocytes with neurons. Enlightenment brought by integrative genomic structure and developmental trajectory The phenotypic divergence observed within complex families, ranging from unaffected individuals to those diagnosed with OCD or SCZ, may be explained by the specific configuration of rare inherited variants, their regulatory loci, and the timing of their activity during neurodevelopment. Our study indicates that while some family members inherit high-impact coding or noncoding variants active in disorder-specific pathways, others either lack such variants or possess variants that fall outside critical spatiotemporal windows. This may account for the absence of clinical symptoms in certain individuals. The "two-hit" or even “multi-hit” hypothesis in SCZ posits that multiple genetic and environmental factors must converge during key periods of neurodevelopment to precipitate the disorder in genetically susceptible individuals( 145 , 146 ). A similarly complex architecture may underlie OCD etiology. While blood–brain barrier (BBB) dysfunction and neurovascular developmental alteration has been linked to several pathogenic mechanisms, including oxidative stress, inflammation and synaptic development( 120 ), our findings suggest that they may represent a core primary defect tied to genetic susceptibility in both SCZ and OCD, particularly in OCD, where environmental modulation appears less influential. Notably, deficits related to calcium or ion channels were more prominently associated with SCZ, indicating a higher pathogenic burden in this disorder. Furthermore,, the strong enrichment of noncoding variants in early-to-mid pregnancy modules, alongside moderate involvement in postnatal modules, supports the hypothesis that cell proliferation and synaptic elimination represent key vulnerability windows in SCZ ( 146 , 147 ). These critical developmental stages may therefore constitute promising targets for preventive strategies and epigenomic interventions. Environmental risk factors, such as prenatal vitamin D deficiency, viral infections, poor nutrition, postnatal social adversity, and childhood trauma, may exert cumulative and interactive effects, influencing disease onset by acting during different critical windows of neurodevelopment( 148 ). While prior studies highlight the transdiagnostic impact of such exposures across mental disorders ( 148 ), our prediction model points out that the trajectory toward a specific disorder is strongly rooted in genomic architecture. Given the reversibility of epigenetic modifications, future research should focus on clarifying how these risk factors interact with specific genes, pathways, or cell types at defined developmental stages. Such insights may ultimately facilitate targeted interventions aimed at correcting noncoding regulatory disruptions or epigenetic defects( 148 ). Together, our findings offer a mechanistic framework for understanding phenotypic divergence within genetically similar individuals. Despite a shared genetic background, the presence of distinct rare variants, whether in coding regions or regulatory elements, appears to direct individuals along divergent neurodevelopmental trajectories, resulting in variable clinical outcomes. Toward Disorder-specific Stratification within High-burden Families Our results suggest that mixed-pedigree designs can move cross-disorder genetics beyond population-level overlap toward within-family divergence, where individuals share substantial background yet exhibit distinct diagnoses. In this setting, combinations of rare coding and regulatory variants may bias spatiotemporal neurodevelopmental programs toward OCD, SCZ, or resilience. While the current cohort size limits definitive modeling of intra-family outcomes, the framework illustrates how family-grounded burden profiles and developmental context can be integrated to generate interpretable, disorder-weighted signatures. Future studies in larger mixed-pedigree cohorts and complementary functional systems will be required to validate specific mechanisms and assess clinical generalizability. Limitations Several limitations should be noted. First, the strict ascertainment criteria required for mixed OCD–SCZ pedigrees constrained sample size, which reduces power for gene-level inference and limits the stability of disorder-specific estimates. Second, regulatory annotations for non-coding variants remain imperfect, and the interpretability of non-coding burden is therefore conditional on current epigenomic resources and priors. Third, our conclusions are based on genomic and in silico network/developmental analyses; experimental validation in cellular or animal systems will be necessary to test causal mechanisms and pinpoint functional consequences of prioritized variants and pathways. Conclusion In stage I of the CoFFiR project, we used deep whole-genome sequencing of mixed OCD–SCZ pedigrees to dissect convergent and divergent rare-variant architectures under a largely shared familial background. Integrating coding and regulatory variation, we derived family-aware burden signatures that separated affected individuals from controls and provided interpretable disorder-weighted patterns consistent with partial genetic sharing between OCD and SCZ. Developmental network mapping further nominated temporally stratified modules that may underlie shared susceptibility as well as disorder-biased programs. Together, these findings support mixed-pedigree WGS as a complementary strategy for cross-disorder psychiatric genetics and provide a framework for mechanistic prioritization and future replication. Declarations Acknowledgements We thanked for the substantial support from all the patients and healthy volunteers. We thanked every member in our group exploring the mechanism of OCD and better strategy of OCD treatment. Conflict of Interest The authors declare that they have no competing interests. Ethics statement This study was approved by Ethics Committee of Shanghai Mental Health Center, School of Medicine, Shanghai Jiao Tong University. Written informed consent was obtained from each participant. Data available statement The data that support the findings of this study are available from the corresponding author upon reasonable request. Author contributions MD is responsible for original draft, visualization and data curation, YW is responsible for resources collecting and project administration, WW is responsible for methodology and project administration, YB and ZL are responsible for methodology, QF, LW, SS and ZW are responsible or resources collecting, XL and GNL are responsible for project administration, review, editing, and funding acquisition, ZX is responsible for conceptualization, supervision and funding acquisition. Funding This work was supported by National Natural Science Foundation of China (81971261, 82071518, 32200924, 82571771), the Brain Science and Brain-like Intelligence Technology-National Science and Technology Major Project (No. 2022ZD 0209100), Shanghai Science and Technology Committee (22YF1439000), Natural Science Foundation of Shanghai (no: 25ZR1401167), Hospital Project of Shanghai Mental Health Center (2019-YJ15) in the sample collection and genotyping. References P. H. Lee, Y. A. Feng, J. W. Smoller. Pleiotropy and Cross-Disorder Genetics Among Psychiatric Disorders. Biol Psychiatry. 2021;89(1):20–31. p. American Psychiatric Association, D. S. M. T. F. a. American Psychiatric Association. Diagnostic and statistical manual of mental disorders : DSM-5™. 5th edition. ed. Washington, DC ;: American Psychiatric Publishing, a division of American Psychiatric Association; 2013. World Health Organization. ICD-11 revision [cited 2026 24th January]. Available from: https://icd.who.int/en. C. Wander. Schizophrenia: opportunities to improve outcomes and reduce economic burden through managed care. Am J Manag Care. 2020;26(3 Suppl):S62–s8. D. J. Stein, D. L. C. Costa, C. Lochner, E. C. Miguel, Y. C. J. Reddy, R. G. Shavitt, et al. Obsessive-compulsive disorder. Nat Rev Dis Primers. 2019;5(1):52. P. H. Lee, V. Anttila, H. Won, Y.-C. A. Feng, J. Rosenthal, Z. Zhu, et al. Genomic Relationships, Novel Loci, and Pleiotropic Mechanisms across Eight Psychiatric Disorders. Cell. 2019;179(7):1469–82.e11. A. M. Achim, M. Maziade, E. Raymond, D. Olivier, C. Merette, M. A. Roy. How prevalent are anxiety disorders in schizophrenia? A meta-analysis and critical review on a significant association. Schizophr Bull. 2011;37(4):811–21. M. Swets, J. Dekker, K. van Emmerik-van Oortmerssen, G. E. Smid, F. Smit, L. de Haan, et al. The obsessive compulsive spectrum in schizophrenia, a meta-analysis and meta-regression exploring prevalence rates. Schizophr Res. 2014;152(2-3):458–68. M. J. Owen, A. Sawa, P. B. Mortensen. Schizophrenia. Lancet. 2016;388(10039):86–97. J. W. Smoller, O. A. Andreassen, H. J. Edenberg, S. V. Faraone, S. J. Glatt, K. S. Kendler. Psychiatric genetics and the structure of psychopathology. Mol Psychiatry. 2019;24(3):409–20. S. M. Meier, L. Petersen, M. G. Pedersen, M. C. Arendt, P. R. Nielsen, M. Mattheisen, et al. Obsessive-compulsive disorder as a risk factor for schizophrenia: a nationwide study. Jama Psychiatry. 2014;71(11):1215–21. M. H. Chen, S. J. Tsai, C. S. Liang, C. M. Cheng, T. P. Su, T. J. Chen, et al. Diagnostic progression to schizophrenia in 35,255 patients with obsessive-compulsive disorder: a longitudinal follow-up study. Eur Arch Psychiatry Clin Neurosci. 2023;273(3):541–51. M. Cederlof, P. Lichtenstein, H. Larsson, M. Boman, C. Ruck, M. Landen, et al. Obsessive-Compulsive Disorder, Psychosis, and Bipolarity: A Longitudinal Cohort and Multigenerational Family Study. Schizophr Bull. 2015;41(5):1076–83. C. Cappi, M. E. Oliphant, Z. Peter, G. Zai, M. Conceicao do Rosario, C. A. W. Sullivan, et al. De Novo Damaging DNA Coding Mutations Are Associated With Obsessive-Compulsive Disorder and Overlap With Tourette's Disorder and Autism. Biol Psychiatry. 2020;87(12):1035–44. G. N. Lin, W. Song, W. Wang, P. Wang, H. Yu, W. Cai, et al. De novo mutations identified by whole-genome sequencing implicate chromatin modifications in obsessive-compulsive disorder. Sci Adv. 2022;8(2):eabi6180. M. Fromer, A. J. Pocklington, D. H. Kavanagh, H. J. Williams, S. Dwyer, P. Gormley, et al. De novo mutations in schizophrenia implicate synaptic networks. Nature. 2014;506(7487):179–84. S. E. McCarthy, J. Gillis, M. Kramer, J. Lihm, S. Yoon, Y. Berstein, et al. De novo mutations in schizophrenia implicate chromatin remodeling and support a genetic overlap with autism and intellectual disability. Mol Psychiatry. 2014;19(6):652–8. A. Takata, B. Xu, I. Ionita-Laza, J. L. Roos, J. A. Gogos, M. Karayiorgou. Loss-of-function variants in schizophrenia risk and SETD1A as a candidate susceptibility gene. Neuron. 2014;82(4):773–80. M. Halvorsen, R. Huh, N. Oskolkov, J. Wen, S. Netotea, P. Giusti-Rodriguez, et al. Increased burden of ultra-rare structural variants localizing to boundaries of topologically associated domains in schizophrenia. Nat Commun. 2020;11(1):1842. E. Rees, H. D. J. Creeth, H. G. Hwu, W. J. Chen, M. Tsuang, S. J. Glatt, et al. Schizophrenia, autism spectrum disorders and developmental disorders share specific disruptive coding mutations. Nat Commun. 2021;12(1):5353. D. Liu, D. Meyer, B. Fennessy, C. Feng, E. Cheng, J. S. Johnson, et al. Schizophrenia risk conferred by rare protein-truncating variants is conserved across diverse human populations. Nat Genet. 2023;55(3):369–76. W. A. Akingbuwa, A. R. Hammerschlag, M. Bartels, M. G. Nivard, C. M. Middeldorp. Ultra-rare and common genetic variant analysis converge to implicate negative selection and neuronal processes in the aetiology of schizophrenia. Mol Psychiatry. 2022;27(9):3699–707. P. F. Sullivan, S. Yao, J. Hjerling-Leffler. Schizophrenia genomics: genetic complexity and functional insights. Arch Gen Psychiatry. 2024. J. Ott, Y. Kamatani, M. Lathrop. Family-based designs for genome-wide association studies. Nat Rev Genet. 2011;12(7):465–74. W. K. Goodman, L. H. Price, S. A. Rasmussen, C. Mazure, R. L. Fleischmann, C. L. Hill, et al. The Yale-Brown Obsessive Compulsive Scale: I. Development, Use, and Reliability. Archives of General Psychiatry. 1989;46(11):1006–11. S. R. Kay, A. Fiszbein, L. A. Opler. The Positive and Negative Syndrome Scale (PANSS) for Schizophrenia. Schizophr Bull. 1987;13(2):261–76. M. Hamilton. The assessment of anxiety states by rating. Br J Med Psychol. 1959;32(1):50–5. M. Hamilton. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol. 1967;6(4):278–96. Y. Chen, Y. Chen, C. Shi, Z. Huang, Y. Zhang, S. Li, et al. SOAPnuke: a MapReduce acceleration-supported software for integrated quality control and preprocessing of high-throughput sequencing data. Gigascience. 2018;7(1):1–6. H. Li, R. Durbin. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25(14):1754–60. P. Danecek, J. K. Bonfield, J. Liddle, J. Marshall, V. Ohan, M. O. Pollard, et al. Twelve years of SAMtools and BCFtools. Gigascience. 2021;10(2):giab008. K. Wang, M. Li, H. Hakonarson. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010;38(16):e164–e. M. Lek, K. J. Karczewski, E. V. Minikel, K. E. Samocha, E. Banks, T. Fennell, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536(7616):285–91. M. Schubach, T. Maass, L. Nazaretyan, S. Röner, M. Kircher. CADD v1.7: using protein language models, regulatory CNNs and other nucleotide-level scores to improve genome-wide variant predictions. Nucleic Acids Res. 2024;52(D1):D1143–D54. K. E. Samocha, J. A. Kosmicki, K. J. Karczewski, A. H. O’Donnell-Luria, E. Pierce-Hoffman, D. G. MacArthur, et al. Regional missense constraint improves variant deleteriousness prediction. bioRxiv. 2017:148353. R. Steinhaus, S. Proft, M. Schuelke, D. N. Cooper, Jana M. Schwarz, D. Seelow. MutationTaster2021. Nucleic Acids Res. 2021;49(W1):W446–W51. S. Chun, J. C. Fay. Identification of deleterious mutations within three human genomes. Genome Res. 2009;19(9):1553–61. A. Siepel, G. Bejerano, J. S. Pedersen, A. S. Hinrichs, M. Hou, K. Rosenbloom, et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 2005;15(8):1034–50. X. Liu, X. Jian, E. Boerwinkle. dbNSFP: A lightweight database of human nonsynonymous SNPs and their functional predictions. Hum Mutat. 2011;32(8):894–9. X. Liu, C. Li, C. Mou, Y. Dong, Y. Tu. dbNSFP v4: a comprehensive database of transcript-specific functional predictions and annotations for human nonsynonymous and splice-site SNVs. Genome Med. 2020;12(1):103. J. T. Robinson, H. Thorvaldsdóttir, W. Winckler, M. Guttman, E. S. Lander, G. Getz, et al. Integrative genomics viewer. Nat Biotechnol. 2011;29(1):24–6. M. J. Gandal, P. Zhang, E. Hadjimichael, R. L. Walker, C. Chen, S. Liu, et al. Transcriptome-wide isoform-level dysregulation in ASD, schizophrenia, and bipolar disorder. Science. 2018;362(6420). C. Roadmap Epigenomics, A. Kundaje, W. Meuleman, J. Ernst, M. Bilenky, A. Yen, et al. Integrative analysis of 111 reference human epigenomes. Nature. 2015;518(7539):317–30. K. M. Chen, A. K. Wong, O. G. Troyanskaya, J. Zhou. A sequence-based global map of regulatory activity for deciphering human genetics. Nat Genet. 2022;54(7):940–9. K. Chen, J. W. Wallis, M. D. McLellan, D. E. Larson, J. M. Kalicki, C. S. Pohl, et al. BreakDancer: an algorithm for high-resolution mapping of genomic structural variation. Nat Methods. 2009;6(9):677–81. A. Abyzov, A. E. Urban, M. Snyder, M. Gerstein. CNVnator: an approach to discover, genotype, and characterize typical and atypical CNVs from family and population genome sequencing. Genome Res. 2011;21(6):974–84. W. McLaren, L. Gil, S. E. Hunt, H. S. Riat, G. R. S. Ritchie, A. Thormann, et al. The Ensembl Variant Effect Predictor. Genome Biol. 2016;17(1):122. V. Geoffroy, J.-B. Lamouche, T. Guignard, S. Nicaise, A. Kress, S. Scheidecker, et al. The AnnotSV webserver in 2023: updated visualization and ranking. Nucleic Acids Res. 2023;51(W1):W39–W45. V. Geoffroy, Y. Herenger, A. Kress, C. Stoetzel, A. Piton, H. Dollfus, et al. AnnotSV: an integrated tool for structural variations annotation. Bioinformatics. 2018;34(20):3572–4. E. Sjöstedt, W. Zhong, L. Fagerberg, M. Karlsson, N. Mitsios, C. Adori, et al. An atlas of the protein-coding genes in the human, pig, and mouse brain. Science. 2020;367(6482). S. M. Weyn-Vanhentenryck, A. Mele, Q. Yan, S. Sun, N. Farny, Z. Zhang, et al. HITS-CLIP and integrative modeling define the Rbfox splicing-regulatory network linked to brain development and autism. Cell Rep. 2014;6(6):1139–52. Y. Lv, L. Wen, W. J. Hu, C. Deng, H. W. Ren, Y. N. Bao, et al. Schizophrenia in the genetic era: a review from development history, clinical features and genomic research approaches to insights of susceptibility genes. Metab Brain Dis. 2024;39(1):147–71. K. J. Karczewski, L. C. Francioli, G. Tiao, B. B. Cummings, J. Alföldi, Q. Wang, et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020;581(7809):434–43. A. Bayes, L. N. van de Lagemaat, M. O. Collins, M. D. Croning, I. R. Whittle, J. S. Choudhary, et al. Characterization of the proteome, diseases and evolution of the human postsynaptic density. Nat Neurosci. 2011;14(1):19–21. J. C. Darnell, S. J. Van Driesche, C. Zhang, K. Y. Hung, A. Mele, C. E. Fraser, et al. FMRP stalls ribosomal translocation on mRNAs linked to synaptic function and autism. Cell. 2011;146(2):247–61. S. Chen, R. Fragoza, L. Klei, Y. Liu, J. Wang, K. Roeder, et al. An interactome perturbation framework prioritizes damaging missense mutations for developmental disorders. Nat Genet. 2018;50(7):1032–40. A. Sugathan, M. Biagioli, C. Golzio, S. Erdin, I. Blumenthal, P. Manavalan, et al. CHD8 regulates neurodevelopmental pathways associated with autism spectrum disorder in neural progenitors. Proc Natl Acad Sci U S A. 2014;111(42):E4468–77. M. Abdi, E. Aliyev, B. Trost, M. Kohailan, W. Aamer, N. Syed, et al. Genomic architecture of autism spectrum disorder in Qatar: The BARAKA-Qatar Study. Genome Med. 2023;15(1):81. C. International Obsessive Compulsive Disorder Foundation Genetics, O. C. D. C. G. A. Studies. Revealing the complex genetic architecture of obsessive-compulsive disorder using meta-analysis. Mol Psychiatry. 2018;23(5):1181–8. S. Ripke, C. O'Dushlaine, K. Chambert, J. L. Moran, A. K. Kahler, S. Akterin, et al. Genome-wide association analysis identifies 13 new risk loci for schizophrenia. Nat Genet. 2013;45(10):1150–9. C. Schizophrenia Working Group of the Psychiatric Genomics. Biological insights from 108 schizophrenia-associated genetic loci. Nature. 2014;511(7510):421–7. A. F. Pardinas, P. Holmans, A. J. Pocklington, V. Escott-Price, S. Ripke, N. Carrera, et al. Common schizophrenia alleles are enriched in mutation-intolerant genes and in regions under strong background selection. Nat Genet. 2018;50(3):381–9. M. Lam, C. Y. Chen, Z. Li, A. R. Martin, J. Bryois, X. Ma, et al. Comparative genetic architectures of schizophrenia in East Asian and European populations. Nat Genet. 2019;51(12):1670–8. V. Trubetskoy, A. F. Pardinas, T. Qi, G. Panagiotaropoulou, S. Awasthi, T. B. Bigdeli, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. 2022;604(7906):502–8. L. D. Ward, M. Kellis. HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res. 2012;40(D1):D930–D4. T. Wu, E. Hu, S. Xu, M. Chen, P. Guo, Z. Dai, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. The Innovation. 2021;2(3). G. Yu, L.-G. Wang, Y. Han, Q.-Y. He. clusterProfiler: an R Package for Comparing Biological Themes Among Gene Clusters. OMICS: A Journal of Integrative Biology. 2012;16(5):284–7. M. V. Fernández, J. Budde, J. L. Del-Aguila, L. Ibañez, Y. Deming, O. Harari, et al. Evaluation of Gene-Based Family-Based Methods to Detect Novel Genes Associated With Familial Late Onset Alzheimer Disease. Front Neurosci. 2018;12. J. Hecker, F. W. Townes, P. Kachroo, C. Laurie, J. Lasky-Su, J. Ziniti, et al. A unifying framework for rare variant association testing in family-based designs, including higher criticism approaches, SKATs, and burden tests. Bioinformatics. 2021;36(22-23):5432–8. S. Lee, M. J. Emond, M. J. Bamshad, K. C. Barnes, M. J. Rieder, D. A. Nickerson, et al. Optimal unified approach for rare-variant association testing with application to small-sample case-control whole-exome sequencing studies. Am J Hum Genet. 2012;91(2):224–37. Y. Liu, S. Chen, Z. Li, A. C. Morrison, E. Boerwinkle, X. Lin. ACAT: A Fast and Powerful p Value Combination Method for Rare-Variant Analysis in Sequencing Studies. Am J Hum Genet. 2019;104(3):410–21. Z. Liu, W. Qian, W. Cai, W. Song, W. Wang, D. T. Maharjan, et al. Inferring the Effects of Protein Variants on Protein–Protein Interactions with Interpretable Transformer Representations. Research.6:0219. R. Oughtred, J. Rust, C. Chang, B. J. Breitkreutz, C. Stark, A. Willems, et al. The BioGRID database: A comprehensive biomedical resource of curated protein, genetic, and chemical interactions. Protein Sci. 2021;30(1):187–200. P. Shannon, A. Markiel, O. Ozier, N. S. Baliga, J. T. Wang, D. Ramage, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498–504. Y. Zhou, B. Zhou, L. Pache, M. Chang, A. H. Khodabakhshi, O. Tanaseichuk, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523. K. Siletti, R. Hodge, A. Mossi Albiach, K. W. Lee, S. L. Ding, L. Hu, et al. Transcriptomic diversity of cell types across the adult human brain. Science. 2023;382(6667):eadd7046. N. G. Skene, S. G. Grant. Identification of Vulnerable Cell Types in Major Brain Disorders Using Single Cell Transcriptomes and Expression Weighted Cell Type Enrichment. Front Neurosci. 2016;10:16. P. S. Emani, J. J. Liu, D. Clarke, M. Jensen, J. Warrell, C. Gupta, et al. Single-cell genomics and regulatory networks for 388 human brains. Science. 2024;384(6698):eadi5199. S. E. Ahmari, S. L. Rauch. The prefrontal cortex and OCD. Neuropsychopharmacology. 2022;47(1):211–24. P. Langfelder, S. Horvath. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9(1):559. S. Petrovski, A. B. Gussow, Q. Wang, M. Halvorsen, Y. Han, W. H. Weir, et al. The Intolerance of Regulatory Sequence to Genetic Variation Predicts Gene Dosage Sensitivity. PLoS Genet. 2015;11(9):e1005492. N. Delphin, C. Aust, L. Griffiths, F. Fernandez. Epigenetic Regulation in Schizophrenia: Focus on Methylation and Histone Modifications in Human Studies. Genes (Basel). 2024;15(3). M. G. Butler, A. B. McGuire, H. Masoud, A. M. Manzardo. Currently recognized genes for schizophrenia: High-resolution chromosome ideogram representation. Am J Med Genet B Neuropsychiatr Genet. 2016;171B(2):181–202. N. E. Clifton, L. Collado-Torres, E. E. Burke, A. F. Pardinas, J. C. Harwood, A. Di Florio, et al. Developmental Profile of Psychiatric Risk Associated With Voltage-Gated Cation Channel Activity. Biol Psychiatry. 2021;90(6):399–408. R. Zhai, G. Tong, Z. Li, W. Song, Y. Hu, S. Xu, et al. Rhesus monkeys exhibiting spontaneous ritualistic behaviors resembling obsessive-compulsive disorder. Natl Sci Rev. 2023;10(11):nwad312. M. Deng, X. Li, D. Shi, Q. Fan, H. Zhang, Z. Wang, et al. iTRAQ-Based Serum Proteomic Analysis Reveals Multifactorial Cellular Function Impairment and Aggravated Systematic Inflammation in Drug-free Obsessive-Compulsive Disorders. ACS Chem Neurosci. 2024. K. C. de Oliveira, C. Camilo, V. D. Gastaldi, A. Sant'Anna Feltrin, B. C. G. Lisboa, V. de Jesus Rodrigues de Paula, et al. Brain areas involved with obsessive-compulsive disorder present different DNA methylation modulation. BMC Genom Data. 2021;22(1):45. J. J. Moffat, M. Ka, E. M. Jung, A. L. Smith, W. Y. Kim. The role of MACF1 in nervous system development and maintenance. Semin Cell Dev Biol. 2017;69:9–17. Q. Wang, M. Li, Z. Yang, X. Hu, H. M. Wu, P. Ni, et al. Increased co-expression of genes harboring the damaging de novo mutations in Chinese schizophrenic patients during prenatal development. Sci Rep. 2015;5:18209. P. J. Harrison, D. R. Weinberger. Schizophrenia genes, gene expression, and neuropathology: on the matter of their convergence. Mol Psychiatry. 2005;10(1):40–68. D. Salem, R. J. Fecek. Role of microtubule actin crosslinking factor 1 (MACF1) in bipolar disorder pathophysiology and potential in lithium therapeutic mechanism. Transl Psychiatry. 2023;13(1):221. J. Pol-Fuster, F. Canellas, L. Ruiz-Guerra, A. Medina-Dols, B. Bisbal-Carrio, V. Asensio, et al. Familial Psychosis Associated With a Missense Mutation at MACF1 Gene Combined With the Rare Duplications DUP3p26.3 and DUP16q23.3, Affecting the CNTN6 and CDH13 Genes. Front Genet. 2021;12:622886. H. H. Jensen, M. Brohus, J. W. Hussey, A.-O. Busuioc, E. D. Iversen, F. Darki, et al. Calmodulin variants in schizophrenia patients display gain-of-function or loss-of-function effects. medRxiv. 2024:2024.05.22.24307674. F. Vidal-Domenech, G. Riquelme, R. Pinacho, R. Rodriguez-Mias, A. Vera, A. Monje, et al. Calcium-binding proteins are altered in the cerebellum in schizophrenia. PLoS One. 2020;15(7):e0230400. E. Parnell, L. Culotta, M. P. Forrest, H. A. Jalloul, B. L. Eckman, D. D. Loizzo, et al. Excitatory Dysfunction Drives Network and Calcium Handling Deficits in 16p11.2 Duplication Schizophrenia Induced Pluripotent Stem Cell-Derived Neurons. Biol Psychiatry. 2023;94(2):153–63. K.-C. Huang, K.-C. Yang, H. Lin, T. T.-H. Tsao, S.-A. Lee. Transcriptome alterations of mitochondrial and coagulation function in schizophrenia by cortical sequencing analysis. BMC Genomics. 2014;15(9):S6. X. Weng, D. Zhou, F. Liu, H. Zhang, J. Ye, Z. Zhang, et al. DNA methylation profiling in the thalamus and hippocampus of postnatal malnourished mice, including effects related to long-term potentiation. BMC Neurosci. 2014;15(1):31. K. H. Lim, S. Yang, S. H. Kim, E. Ko, M. Kang, J. Y. Joo. Cryptic mutations of PLC family members in brain disorders: recent discoveries and a deep-learning-based approach. Brain. 2023;146(4):1267–80. D. Malhotra, S. McCarthy, J. J. Michaelson, V. Vacic, K. E. Burdick, S. Yoon, et al. High frequencies of de novo CNVs in bipolar disorder and schizophrenia. Neuron. 2011;72(6):951–63. R. Trifiletti, H. M. Lachman, O. Manusama, D. Zheng, A. Spalice, P. Chiurazzi, et al. Identification of ultra-rare genetic variants in pediatric acute onset neuropsychiatric syndrome (PANS) by exome and whole genome sequencing. Sci Rep. 2022;12(1):11106. K. A. Demin, N. A. Krotova, N. P. Ilyin, D. S. Galstyan, T. O. Kolesnikova, T. Strekalova, et al. Evolutionarily conserved gene expression patterns for affective disorders revealed using cross-species brain transcriptomic analyses in humans, rats and zebrafish. Sci Rep. 2022;12(1):20836. A. Patowary, P. Zhang, C. Jops, C. K. Vuong, X. Ge, K. Hou, et al. Developmental isoform diversity in the human neocortex informs neuropsychiatric risk mechanisms. Science. 2024;384(6698):eadh7688. J. C. Silbereis, S. Pochareddy, Y. Zhu, M. Li, N. Sestan. The Cellular and Molecular Landscapes of the Developing Human Central Nervous System. Neuron. 2016;89(2):248–68. R. Birnbaum, D. R. Weinberger. The Genesis of Schizophrenia: An Origin Story. Am J Psychiatry. 2024;181(6):482–92. X. Mao, J. F. Staiger. Multimodal cortical neuronal cell type classification. Pflugers Arch. 2024;476(5):721–33. L. Nahar, B. M. Delacroix, H. W. Nam. The Role of Parvalbumin Interneurons in Neurotransmitter Balance and Neurological Disease. Front Psychiatry. 2021;12:679960. S. J. Kaar, I. Angelescu, T. R. Marques, O. D. Howes. Pre-frontal parvalbumin interneurons in schizophrenia: a meta-analysis of post-mortem studies. J Neural Transm (Vienna). 2019;126(12):1637–51. J. A. Cale, E. J. Chauhan, J. J. Cleaver, A. R. Fusciardi, S. McCann, H. C. Waters, et al. GABAergic and inflammatory changes in the frontal cortex following neonatal PCP plus isolation rearing, as a dual-hit neurodevelopmental model for schizophrenia. Mol Neurobiol. 2024. O. Marin. Parvalbumin interneuron deficits in schizophrenia. Eur Neuropsychopharmacol. 2024;82:44–52. A. J. Price, A. E. Jaffe, D. R. Weinberger. Cortical cellular diversity and development in schizophrenia. Mol Psychiatry. 2021;26(1):203–17. M. C. Tiveron, C. Beurrier, C. Ceni, N. Andriambao, A. Combes, M. Koehl, et al. LAMP5 Fine-Tunes GABAergic Synaptic Transmission in Defined Circuits of the Mouse Brain. PLoS One. 2016;11(6):e0157052. M. Y. Batiuk, T. Tyler, K. Dragicevic, S. Mei, R. Rydbirk, V. Petukhov, et al. Upper cortical layer–driven network impairment in schizophrenia. Science Advances.8(41):eabn8367. W. Cai, W. Song, Z. Liu, D. T. Maharjan, J. Liang, G. N. Lin. An Integrative Analysis of Identified Schizophrenia-Associated Brain Cell Types and Gene Expression Changes. Int J Mol Sci. 2022;23(19). A. Rampino, T. Annese, S. Torretta, R. Tamma, R. M. Falcone, D. Ribatti. Involvement of vascular endothelial growth factor in schizophrenia. Neurosci Lett. 2021;760:136093. S. Pong, R. Karmacharya, M. Sofman, J. R. Bishop, P. Lizano. The Role of Brain Microvascular Endothelial Cell and Blood-Brain Barrier Dysfunction in Schizophrenia. Complex Psychiatry. 2020;6(1-2):30–46. S. Zengil, E. Laloglu. Evaluation of serum zonulin and occludin levels in obsessive-compulsive disorder and the effect of major depressive disorder comorbidity. Front Psychiatry. 2024;15:1395235. N. W. McGregor, C. Lochner, D. J. Stein, S. M. Hemmings. Polymorphisms within the neuronal cadherin (CDH2) gene are associated with obsessive-compulsive disorder (OCD) in a South African cohort. Metab Brain Dis. 2016;31(1):191–6. M. Zamanian-Azodi, M. Rezaei-Tavirani, M. Rezaei Tavirani. Obsessive-Compulsive Disorder Interactome Profile Analysis: A Perspective From Molecular Mechanism. Basic Clin Neurosci. 2021;12(2):187–98. A. Bilgiç, B. Akbaş İleri, İ. Kılınç, Ö. F. Akça. Decreased plasma intercellular adhesion molecule-3 levels are associated with obsessive compulsive disorder in medication-free children. J. Obsessive Compuls. Relat. Disord. 2020;27. I. Stankovic, M. Notaras, P. Wolujewicz, T. Lu, R. Lis, M. E. Ross, et al. Schizophrenia endothelial cells exhibit higher permeability and altered angiogenesis patterns in patient-derived organoids. Transl Psychiatry. 2024;14(1):53. S. Stanca, M. Rossetti, L. Bokulic Panichi, P. Bongioanni. The Cellular Dysfunction of the Brain-Blood Barrier from Endothelial Cells to Astrocytes: The Pathway towards Neurotransmitter Impairment in Schizophrenia. Int J Mol Sci. 2024;25(2). A. Vreeland, D. Calaprice, N. Or-Geva, R. E. Frye, D. Agalliu, H. M. Lachman, et al. Postinfectious Inflammation, Autoimmunity, and Obsessive-Compulsive Disorder: Sydenham Chorea, Pediatric Autoimmune Neuropsychiatric Disorder Associated with Streptococcal Infection, and Pediatric Acute-Onset Neuropsychiatric Disorder. Dev Neurosci. 2023;45(6):361–74. D. Endres, T. A. Pollak, K. Bechter, D. Denzel, K. Pitsch, K. Nickel, et al. Immunological causes of obsessive-compulsive disorder: is it time for the concept of an "autoimmune OCD" subtype? Transl Psychiatry. 2022;12(1):5. S. Bhattacharyya, S. Khanna, K. Chakrabarty, A. Mahadevan, R. Christopher, S. K. Shankar. Anti-brain autoantibodies and altered excitatory neurotransmitters in obsessive-compulsive disorder. Neuropsychopharmacology. 2009;34(12):2489–96. D. Ribatti, D. Guidolin. Morphogenesis of vascular and neuronal networks and the relationships between their remodeling processes. Brain Res Bull. 2022;186:62–9. J. Kealy, C. Greene, M. Campbell. Blood-brain barrier regulation in psychiatric disorders. Neurosci Lett. 2020;726:133664. E. L. Goldwaser, D. J. J. Wang, B. M. Adhikari, J. Chiappelli, X. Shao, J. Yu, et al. Evidence of Neurovascular Water Exchange and Endothelial Vascular Dysfunction in Schizophrenia: An Exploratory Study. Schizophr Bull. 2023;49(5):1325–35. C. Y. Chang, D. Z. Luo, J. C. Pei, M. C. Kuo, Y. C. Hsieh, W. S. Lai. Not Just a Bystander: The Emerging Role of Astrocytes and Research Tools in Studying Cognitive Dysfunctions in Schizophrenia. Int J Mol Sci. 2021;22(10). S. Kiarash, B. Mohammad, A. Banafshe, P. Zeynab, E. Zahra, A. Shiva, et al. Astrocytic Abnormalities in Schizophrenia. In: Thomas H, editor. Neurophysiology. Rijeka: IntechOpen; 2022. p. Ch. 12. E. C. de Oliveira Figueiredo, C. Cali, F. Petrelli, P. Bezzi. Emerging evidence for astrocyte dysfunction in schizophrenia. Glia. 2022;70(9):1585–604. V. V. Tarasov, A. A. Svistunov, V. N. Chubarev, S. S. Sologova, P. Mukhortova, D. Levushkin, et al. Alterations of Astrocytes in the Context of Schizophrenic Dementia. Front Pharmacol. 2019;10:1612. D. Senko, O. Efimova, M. Osetrova, N. Anikanov, M. Boyko, M. Sharaev, et al. White matter lipidome alterations in the schizophrenia brain. Schizophrenia (Heidelb). 2024;10(1):123. M. Valdés-Tovar, A. M. Rodríguez-Ramírez, L. Rodríguez-Cárdenas, C. E. Sotelo-Ramírez, B. Camarena, M. A. Sanabrais-Jiménez, et al. Insights into myelin dysfunction in schizophrenia and bipolar disorder. World J Psychiatry. 2022;12(2):264–85. L. Bast, S. Yao, J. A. Martinez-Lopez, F. Memic, H. French, M. Valiukonyte, et al. Transcriptomic and genetic analysis suggests a role for mitochondrial dysregulation in schizophrenia. medRxiv. 2025. S. L. Mondragon-Gonzalez, C. Schreiweis, E. Burguiere. Closed-loop recruitment of striatal interneurons prevents compulsive-like grooming behaviors. Nat Neurosci. 2024;27(6):1148–56. S. M. Perez, A. Boley, D. J. Lodge. Region specific knockdown of Parvalbumin or Somatostatin produces neuronal and behavioral deficits consistent with those observed in schizophrenia. Transl Psychiatry. 2019;9(1):264. A. J. Murray, M. U. Woloszynowska-Fraser, L. Ansel-Bollepalli, K. L. Cole, A. Foggetti, B. Crouch, et al. Parvalbumin-positive interneurons of the prefrontal cortex support working memory and cognitive flexibility. Sci Rep. 2015;5:16778. A. M. Frota Lisboa Pereira de Souza, L. Pellegrini, N. A. Fineberg. Cognitive inflexibility, obsessive-compulsive symptoms and traits and poor post-pandemic adjustment. Neuroscience Applied. 2024;3. P. Gruner, C. Pittenger. Cognitive inflexibility in Obsessive-Compulsive Disorder. Neuroscience. 2017;345:243–55. P. Monteiro, B. Barak, Y. Zhou, R. McRae, D. Rodrigues, I. R. Wickersham, et al. Dichotomous parvalbumin interneuron populations in dorsolateral and dorsomedial striatum. J Physiol. 2018;596(16):3695–707. A. Caballero, E. Flores-Barrera, D. R. Thomases, K. Y. Tseng. Downregulation of parvalbumin expression in the prefrontal cortex during adolescence causes enduring prefrontal disinhibition in adulthood. Neuropsychopharmacology. 2020;45(9):1527–35. F. J. Raabe, L. Slapakova, M. J. Rossner, L. Cantuti-Castelvetri, M. Simons, P. G. Falkai, et al. Oligodendrocytes as A New Therapeutic Target in Schizophrenia: From Histopathological Findings to Neuron-Oligodendrocyte Interaction. Cells. 2019;8(12). P. Falkai, M. J. Rossner, F. J. Raabe, E. Wagner, D. Keeser, I. Maurus, et al. Disturbed Oligodendroglial Maturation Causes Cognitive Dysfunction in Schizophrenia: A New Hypothesis. Schizophr Bull. 2023;49(6):1614–24. N. Takahashi, T. Sakurai, K. L. Davis, J. D. Buxbaum. Linking oligodendrocyte and myelin dysfunction to neurocircuitry abnormalities in schizophrenia. Prog Neurobiol. 2011;93(1):13–24. J. Davis, H. Eyre, F. N. Jacka, S. Dodd, O. Dean, S. McEwen, et al. A review of vulnerability and risks for schizophrenia: Beyond the two hit hypothesis. Neurosci Biobehav Rev. 2016;65:185–94. O. D. Howes, E. C. Onwordi. The synaptic hypothesis of schizophrenia version III: a master mechanism. Mol Psychiatry. 2023;28(5):1843–56. L. D. Selemon, N. Zecevic. Schizophrenia: a tale of two critical periods for prefrontal cortical development. Transl Psychiatry. 2015;5(8):e623. E. J. Nestler, C. J. Pena, M. Kundakovic, A. Mitchell, S. Akbarian. Epigenetic Basis of Mental Illness. Neuroscientist. 2016;22(5):447–63. Tables Tables 1 to 5 are available in the supplementary files section Additional Declarations The authors have declared there is NO conflict of interest to disclose Supplementary Files Table1.docx Table2.docx Table3.docx Table4.docx Table5.docx TableS4.csv Table S4 TableS5.csv Table S5 TableS8.csv Table S8 TableS9.csv Table S9 TableS24.csv Table S24 TableS3.csv Table S3 TableS14.csv Table S14 TableS20.csv Table S20 TableS28.csv Table S28 TableS16.csv Table S16 TableS31.csv Table S31 TableS25.csv Table S25 TableS2.csv Table S2 TableS23.csv Table S23 TableS18.csv Table S18 TableS29.csv Table S29 TableS27.csv Table S27 TableS1.csv Table S1 TableS12.csv Table S12 TableS32.csv Table S32 TableS11.csv Table S11 TableS13.csv Table S13 TableS10.csv Table S10 TableS17.csv Table S17 Supplemantaryinformation.docx Supplemantary information TableS26.csv Table S26 TableS19.csv Table S19 TableS30.csv Table S30 TableS6.csv Table S6 TableS22.csv Table S22 TableS15.csv Table S15 TableS21.csv Table S21 TableS33.csv Table S33 SupplementaryFigures.docx Supplementary Figures TableS7.csv Table S7 Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8943594","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":607207488,"identity":"b78e42ce-23d9-4031-87ad-6d500e3bf99d","order_by":0,"name":"Zeping Xiao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA7ElEQVRIiWNgGAWjYDACZgjFA2QdfPDBwMaOFC1syYYzCtKSSbGPx0ya58MhxgZC6gyOMz98+KPijow5/7JkYxuDA8wM7IePbsCnRbKZzdhA4swzHssZjw8+zjG4w8fAk5Z2A58WfmYGMwnDtsM8BjeOJRvnGDxjZpDgMcOrhY2Z/ZtE4j+QljNm0hYGhxkbCGnhZ+YxkzjYANRyvsdMmoEYLZLNPMWGDceeAW0BBnKPQVoyGyG/GJw/vvHhj5o79gbnDx988OOPjR0/++FjeLVAwQEGBokEqO+IUA7Vwn+ASLWjYBSMglEw4gAAntBLjOds5qoAAAAASUVORK5CYII=","orcid":"","institution":"Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine","correspondingAuthor":true,"prefix":"","firstName":"Zeping","middleName":"","lastName":"Xiao","suffix":""},{"id":607207489,"identity":"7633c9f4-10c8-4313-b800-d002f0dfe679","order_by":1,"name":"Miaohan Deng","email":"","orcid":"https://orcid.org/0000-0003-2058-8764","institution":"Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Miaohan","middleName":"","lastName":"Deng","suffix":""},{"id":607207490,"identity":"64117d4f-912f-4a3d-af7a-8732ff28d704","order_by":2,"name":"Yuan Wang","email":"","orcid":"","institution":"Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yuan","middleName":"","lastName":"Wang","suffix":""},{"id":607207491,"identity":"2a222eb1-91c0-46f5-9654-db30b5891e13","order_by":3,"name":"Weidi Wang","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Weidi","middleName":"","lastName":"Wang","suffix":""},{"id":607207492,"identity":"feada93d-3308-4ba9-a392-b3c232e57c57","order_by":4,"name":"Yihang Bao","email":"","orcid":"","institution":"Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yihang","middleName":"","lastName":"Bao","suffix":""},{"id":607207493,"identity":"e75554b9-ff1f-4948-a3fa-afedca3539ee","order_by":5,"name":"Zhe Liu","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Zhe","middleName":"","lastName":"Liu","suffix":""},{"id":607207494,"identity":"9194020c-8164-46ae-8b0b-279cf91cf8ee","order_by":6,"name":"Qing Fan","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Qing","middleName":"","lastName":"Fan","suffix":""},{"id":607207495,"identity":"1e3fab47-5877-477a-9a7a-1aff97ccba0f","order_by":7,"name":"Lihua Wang","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Lihua","middleName":"","lastName":"Wang","suffix":""},{"id":607207496,"identity":"878378c6-e9d0-4d79-9497-71ad2ce59d6e","order_by":8,"name":"Shanshan Su","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Shanshan","middleName":"","lastName":"Su","suffix":""},{"id":607207497,"identity":"9625adce-7e4a-4cf6-93bb-5f48bc1aa1db","order_by":9,"name":"Zhen Wang","email":"","orcid":"https://orcid.org/0000-0003-4319-5314","institution":"Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine","correspondingAuthor":false,"prefix":"","firstName":"Zhen","middleName":"","lastName":"Wang","suffix":""},{"id":607207498,"identity":"5eef8034-9133-43ec-95f0-29ec18601a6d","order_by":10,"name":"Xia Li","email":"","orcid":"","institution":"Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai","correspondingAuthor":false,"prefix":"","firstName":"Xia","middleName":"","lastName":"Li","suffix":""},{"id":607207499,"identity":"6f225951-1856-4428-a4f0-f75d298b8e4a","order_by":11,"name":"Guan Lin","email":"","orcid":"https://orcid.org/0000-0001-9496-0149","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Guan","middleName":"","lastName":"Lin","suffix":""}],"badges":[],"createdAt":"2026-02-23 06:30:17","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8943594/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8943594/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106401418,"identity":"053e3bb1-9ed0-45d3-bb40-5e08c637742e","added_by":"auto","created_at":"2026-04-08 08:49:29","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":97257,"visible":true,"origin":"","legend":"\u003cp\u003eWork flow of the study. We applied 60× WGS in 21 OCD-SCZ complex families to investigate the shared and distinct genetic components between OCD and SCZ, aiming to uncover the convergence and divergence in their etiology. In coding regions, we found that co-segregating genes and OCD-specific genes were associated with basic microtube-actin functional pathways, while SCZ-specific genes were linked to calcium ion-related pathways. In the non-coding regions within putative candidate gene windows shared by SCZ and OCD families, active transcription and promoter states were enriched, with enrichment being more pronounced in OCD. We identified candidate genes of OCD or SCZ including \u003cem\u003eMACF1\u003c/em\u003e, \u003cem\u003eCALM2\u003c/em\u003e, \u003cem\u003eMMS22L\u003c/em\u003ethrough multiple rare variants association tests. By combining variants from coding and non-coding regions, we identified temporal functional modules composed of these two types of variants in OCD and SCZ, analyzed separately. The module trajectories revealed distinct functional modules involved at different developmental stages in the etiology of OCD and SCZ. These include the prominence of epigenetic modification pathways, such as chromatin modification, during early and mid-pregnancy; angiogenesis during late pregnancy; and calcium-related or neuron- and synapse-related pathways after late pregnancy. Based on these results, we brought up the concept of the model of disease-specific genetic risk model, aiming to manifest the family-level prediction value of the comprehensive genetic architecture.\u003c/p\u003e","description":"","filename":"Binder11.png","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/7e62fcfebd112488a66ea592.png"},{"id":105034040,"identity":"75cca945-a893-4772-b537-2e92ea0e63e2","added_by":"auto","created_at":"2026-03-20 07:22:29","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":146078,"visible":true,"origin":"","legend":"\u003cp\u003eSummary of variants in protein coding region\u003c/p\u003e\n\u003cp\u003ea. Principal component analysis (PCA) of samples in the whole cohort\u003c/p\u003e\n\u003cp\u003eb. Average count of stratified variants per sample.\u003c/p\u003e\n\u003cp\u003ec. Distrubution of stratified variants calculated by two-sided Fisher’s exact test.\u003c/p\u003e\n\u003cp\u003ed. Distribution of high-impact variants in different variants types across the whole genome. The variant type referred from the outer circle to the inner was SNVs, indels, CNVs, SVs, separately. Density of variants were mapped with a gradient blue color.\u003c/p\u003e\n\u003cp\u003ee. Integrated matrix of high-impact SNVs/indels in the OCD/SCZ_GWAS panels and 42 cases with detected variants. Rows are panel genes detected in the cases, and columns are the patients. The upper panels show patient diagnoses, gender and family number. The right panels show the pLI, RVIS rank and the variant counts of the gene in all the cases.\u003c/p\u003e\n\u003cp\u003e† S, schizophrenia; O, obsessive-compulsive disorder; N, healthy control.\u003c/p\u003e\n\u003cp\u003e† † “*”referred to p\u0026lt;0.05.\u003c/p\u003e","description":"","filename":"Binder12.png","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/18f304d74722b12fbee543dd.png"},{"id":104859794,"identity":"caf0adf2-0e9f-4a1f-a5ac-960ee70d411c","added_by":"auto","created_at":"2026-03-18 04:56:12","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":98959,"visible":true,"origin":"","legend":"\u003cp\u003eBurden test of regulatory functions in non-coding co-segregated regions\u003c/p\u003e\n\u003cp\u003ea. A sample of a loci annotated as “P Promoter”while annotated by roadmap as “1_TssA”(chr7_5553429)\u003c/p\u003e\n\u003cp\u003eb. Enrichment of functional elements predicted by Roadmap model in non-coding region across the whole genome.\u003c/p\u003e\n\u003cp\u003ec. Enrichment of functional elements predicted by Roadmap model in non-coding region within CS_NCS regions.\u003c/p\u003e\n\u003cp\u003ed. Enrichment of functional elements predicted by Roadmap model in non-coding region within OS_NCS regions.\u003c/p\u003e\n\u003cp\u003ee. Enrichment of functional elements predicted by Roadmap model in non-coding region within SS_NCS regions.\u003c/p\u003e\n\u003cp\u003e† S, schizophrenia; O, obsessive-compulsive disorder; N, healthy control.\u003c/p\u003e\n\u003cp\u003e† †“*”indicated p\u0026lt;0.05,“**”indicated p\u0026lt;0.005,“***”indicated p\u0026lt;0.001\u003c/p\u003e","description":"","filename":"Binder13.png","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/a447b9bb1eb97ffa54ec9d9c.png"},{"id":105751919,"identity":"d0fc4f00-a35d-4201-8158-59aeaf3b2052","added_by":"auto","created_at":"2026-03-30 15:50:47","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":346023,"visible":true,"origin":"","legend":"\u003cp\u003ePrediction model combining non-coding and coding functions in the cohort\u003c/p\u003e\n\u003cp\u003ea. ROC curves of Random Forest classifiers combining Refseq annotation in coding region and roadmap annotation in non-coding region.\u003c/p\u003e\n\u003cp\u003eb. Feature contributions prioritized by SHAP.\u003c/p\u003e\n\u003cp\u003ec. Caculated disease-specific genetic scores based on the disease-specific architecture.\u003c/p\u003e\n\u003cp\u003e† C, schizophrenia+ obsessive-compulsive disorder; S, schizophrenia; O, obsessive-compulsive disorder; N, healthy control.\u003c/p\u003e","description":"","filename":"Binder14.png","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/3d0b1e8957a20da4aa3fd567.png"},{"id":104859796,"identity":"51de3aa8-4ad6-4b4d-a004-e288822e4ccb","added_by":"auto","created_at":"2026-03-18 04:56:12","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":590759,"visible":true,"origin":"","legend":"\u003cp\u003eCo-segregation and disease-specific genes, pathways and networks\u003c/p\u003e\n\u003cp\u003ea. The distribution of the prediction results of MIPPI across different disease-specific groups. Comparisons were made based on the “no effect ” class. “***” referred to the significance of p\u0026lt;0.001.\u003c/p\u003e\n\u003cp\u003eb. The overlap among high-impact co-segregation genes (HI_CSCDS), OCD-specific genes (HI_OSCDS), SCZ-specific genes (HI_SSCDS) in coding region. The inner circle represents gene lists, where hits are arranged along the arc. Genes that hit multiple lists are colored in dark gold, and genes unique to a list are shown in light gold. The blue curves link genes that belong to the same enriched ontology term.\u003c/p\u003e\n\u003cp\u003ec. Heatmap comparing the functional enrichment between HI_OSCDS and HI_SSCDS. Effect size is given as the normalized log10(pvalue) and mapped to the color gradient from red to green.\u003c/p\u003e\n\u003cp\u003ed. Protein-protein network constructed by the high-confidence genes in HI_CSCDS, HI_OSCDS, HI_SSCDS genes. The largest cluster in the original PPI were colored and the other scattered cluster were grey. The nodes in the SCZ-specific network were colored blue, the nodes in the OCD-specific network were colored green, the nodes in the cosegregation network were colored purple, and the nodes calculated to be hub genes were colored orange . Then the main cluster in the three networks were merged, aiming to observe the relationship between networks. The nodes connecting two distinct network were colored pink.\u003c/p\u003e\n\u003cp\u003ee. The snc-RNA cluster information of the hub genes acquire from ABC atlas from Allen Brain Atlas.\u003c/p\u003e\n\u003cp\u003e† S, schizophrenia; O, obsessive-compulsive disorder; N, healthy control.\u003c/p\u003e\n\u003cp\u003e† † “***” referred to the significance of p\u0026lt;0.001.\u003c/p\u003e","description":"","filename":"Binder15.png","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/4f1c2eb2f5247ef826e7867b.png"},{"id":105033671,"identity":"3dc9e108-dcc6-4f0e-840b-90ea5e25c7d7","added_by":"auto","created_at":"2026-03-20 07:21:10","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":159269,"visible":true,"origin":"","legend":"\u003cp\u003eWGCNA and functional enrichment\u003c/p\u003e\n\u003cp\u003ea. Trajectories of modules divided by weighted gene coexpression network analysis (WGCNA) . The top panel showed modules in schizophrenia (SCZ) networks and the bottom panel showed modules in obsessive-compulsive disorder (OCD) networks. The x-axis was mapped to the sample ages in the original data and the y-axis was mapped to the average expression level measured by RPKM of genes within the module. The grey line showed the average expression level of all the included genes.\u003c/p\u003e\n\u003cp\u003eb. Enrichment of input genesets of each modules, calculated by fisher’s exact test. The top figure was in SCZ and the bottom one was in OCD. Those modules with enriched OR\u0026gt;1 were colored with blue. Color scale represents –log10(p) for gene set enrichment. “*”indicated p\u0026lt;0.05,“**”indicated p\u0026lt;0.005,“***”indicated p\u0026lt;0.001\u003c/p\u003e\n\u003cp\u003ec. Expression Weighted Cell Type Enrichment (EWCE) results of each modules. The left figure was in SCZ and the right one was in OCD. Only celltype-module pairs with enriched p\u0026lt;0.05 were presented. Only celltype-module pairs with q\u0026lt;0.05 were marked with “*”. “*”indicated q\u0026lt;0.05,“**”indicated q\u0026lt;0.005,“***”indicated q\u0026lt;0.001\u003c/p\u003e\n\u003cp\u003ed. Paired metascape enrichment of OCD.green module and SCZ.yellowgreen module.\u003c/p\u003e\n\u003cp\u003ee. Paired metascape enrichment of OCD.brown module and SCZ.brown module.\u003c/p\u003e\n\u003cp\u003ef. Metascape enrichment of SCZ.black module.\u003c/p\u003e\n\u003cp\u003eg. Paired metascape enrichment of early-pregnancy modules (pink, turquoise and yellow moudles) between SCZ and OCD.\u003c/p\u003e\n\u003cp\u003eh. Paired metascape enrichment of mid-pregnancy modules (SCZ.green and SCZ.blue modules versus OCD.blue and OCD.black modules) between SCZ and OCD.\u003c/p\u003e\n\u003cp\u003e† SCZ, schizophrenia; OCD, obsessive-compulsive disorder\u003c/p\u003e","description":"","filename":"Binder16.png","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/ed7df6088ef32ce0eb7b6d5a.png"},{"id":109000283,"identity":"333ded8c-d69d-4764-b172-dfa8d1790bf1","added_by":"auto","created_at":"2026-05-11 14:53:06","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1733631,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/88815973-d1cb-4975-8e3d-dd95bf2221be.pdf"},{"id":105033891,"identity":"62059bb8-58ca-4445-a813-18b130f47798","added_by":"auto","created_at":"2026-03-20 07:22:04","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":16788,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/2ffa95421e16dd5049663d87.docx"},{"id":104859833,"identity":"f8398e7f-d797-4844-a41d-fe6fec114011","added_by":"auto","created_at":"2026-03-18 04:56:15","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":20324,"visible":true,"origin":"","legend":"","description":"","filename":"Table2.docx","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/1bc5a145392de613ca957064.docx"},{"id":104859799,"identity":"0fd53f9b-8f21-455c-aafb-14334930c25f","added_by":"auto","created_at":"2026-03-18 04:56:12","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":16957,"visible":true,"origin":"","legend":"","description":"","filename":"Table3.docx","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/955c6e845ead397ee6252d65.docx"},{"id":105034091,"identity":"73d4e7cc-9f3a-49b3-865f-72b96377deef","added_by":"auto","created_at":"2026-03-20 07:22:38","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":24211,"visible":true,"origin":"","legend":"","description":"","filename":"Table4.docx","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/480d59b4ce51695a79c2114a.docx"},{"id":104859805,"identity":"4469339c-965f-4d2e-8fa5-1d38ed0ebb4e","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"docx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":16936,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cbr\u003e\u003c/p\u003e","description":"","filename":"Table5.docx","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/7f37a2bbc06c6f88bab5caed.docx"},{"id":105034033,"identity":"7d1f5f04-f5a0-4330-a2dc-358b005f2aee","added_by":"auto","created_at":"2026-03-20 07:22:28","extension":"csv","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":3459,"visible":true,"origin":"","legend":"\u003cp\u003eTable S4\u003c/p\u003e","description":"","filename":"TableS4.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/cdca2deac97c7e8ffe3d6e46.csv"},{"id":104859802,"identity":"dcc75027-e94f-4eca-b110-65ffc4f774d5","added_by":"auto","created_at":"2026-03-18 04:56:12","extension":"csv","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":160096,"visible":true,"origin":"","legend":"\u003cp\u003eTable S5\u003c/p\u003e","description":"","filename":"TableS5.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/9869210cbd6e254650a766e3.csv"},{"id":105034118,"identity":"84977237-20de-4846-9aad-5514d9271531","added_by":"auto","created_at":"2026-03-20 07:22:41","extension":"csv","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":24266,"visible":true,"origin":"","legend":"\u003cp\u003eTable S8\u003c/p\u003e","description":"","filename":"TableS8.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/a0fc7bef075b3ad06e99be87.csv"},{"id":105034053,"identity":"c0fce654-ae3c-4ddd-a61a-6fd4aa6c5733","added_by":"auto","created_at":"2026-03-20 07:22:32","extension":"csv","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":23516,"visible":true,"origin":"","legend":"\u003cp\u003eTable S9\u003c/p\u003e","description":"","filename":"TableS9.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/a18133f334ff8a5cf7f6dc8f.csv"},{"id":105033929,"identity":"4eba4d29-07d1-43ec-9ef5-6157270068d0","added_by":"auto","created_at":"2026-03-20 07:22:10","extension":"csv","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":6514,"visible":true,"origin":"","legend":"\u003cp\u003eTable S24\u003c/p\u003e","description":"","filename":"TableS24.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/2d510b6adbe194c86f94ca18.csv"},{"id":105034146,"identity":"24dc1132-b1b3-4b80-a112-5759843421c7","added_by":"auto","created_at":"2026-03-20 07:22:45","extension":"csv","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":1285,"visible":true,"origin":"","legend":"\u003cp\u003eTable S3\u003c/p\u003e","description":"","filename":"TableS3.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/f65578aa29ea1f7951e8ee92.csv"},{"id":104859811,"identity":"cd8c6728-df8d-4225-9507-1601b9073d5c","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":1985,"visible":true,"origin":"","legend":"\u003cp\u003eTable S14\u003c/p\u003e","description":"","filename":"TableS14.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/1160c152f91da5d122358b1a.csv"},{"id":104859808,"identity":"8bd09f51-2a48-40a1-8140-f1bb228dcd85","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":13,"title":"","display":"","copyAsset":false,"role":"supplement","size":808,"visible":true,"origin":"","legend":"\u003cp\u003eTable S20\u003c/p\u003e","description":"","filename":"TableS20.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/e7bcaf87af2bc85de0c4e42b.csv"},{"id":105034021,"identity":"0d12c388-1e22-4466-b4f9-7bbc3714f6a1","added_by":"auto","created_at":"2026-03-20 07:22:27","extension":"csv","order_by":14,"title":"","display":"","copyAsset":false,"role":"supplement","size":5002,"visible":true,"origin":"","legend":"\u003cp\u003eTable S28\u003c/p\u003e","description":"","filename":"TableS28.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/d3e0e95433c57bf400e80784.csv"},{"id":105562601,"identity":"d0fa2a27-9a7e-4f54-90b1-52c89e5d4cc6","added_by":"auto","created_at":"2026-03-27 12:43:28","extension":"csv","order_by":15,"title":"","display":"","copyAsset":false,"role":"supplement","size":8101,"visible":true,"origin":"","legend":"Table S16","description":"","filename":"TableS16.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/9fbcda7acccb7bcb254a4f49.csv"},{"id":104859806,"identity":"c89d02b1-1355-47f3-93fa-30996924cb4e","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":16,"title":"","display":"","copyAsset":false,"role":"supplement","size":13843,"visible":true,"origin":"","legend":"Table S31","description":"","filename":"TableS31.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/6b0d594a143b4ab6dbbaf980.csv"},{"id":105034071,"identity":"d92b31e2-4d4d-4d1f-8955-d6553e5c6c45","added_by":"auto","created_at":"2026-03-20 07:22:36","extension":"csv","order_by":17,"title":"","display":"","copyAsset":false,"role":"supplement","size":3548,"visible":true,"origin":"","legend":"Table S25","description":"","filename":"TableS25.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/b09ef4805eaedab7fe6a4e6d.csv"},{"id":105033782,"identity":"5dd1502d-6d39-4554-bf76-e235e99ccff1","added_by":"auto","created_at":"2026-03-20 07:21:38","extension":"csv","order_by":18,"title":"","display":"","copyAsset":false,"role":"supplement","size":57005,"visible":true,"origin":"","legend":"Table S2","description":"","filename":"TableS2.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/0565a834e57f99bd5df1f608.csv"},{"id":104859820,"identity":"f1d7ed35-447a-443d-8ca0-c1189752ba49","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":19,"title":"","display":"","copyAsset":false,"role":"supplement","size":23799,"visible":true,"origin":"","legend":"Table S23","description":"","filename":"TableS23.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/876a7a03013bb13666b6726b.csv"},{"id":105034022,"identity":"ee521eca-9181-4f7f-b302-f0fa853bf746","added_by":"auto","created_at":"2026-03-20 07:22:27","extension":"csv","order_by":20,"title":"","display":"","copyAsset":false,"role":"supplement","size":27346,"visible":true,"origin":"","legend":"Table S18","description":"","filename":"TableS18.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/955770c862ce1db269d268bd.csv"},{"id":104859815,"identity":"6a067546-6808-4617-8c1e-0ace946e0bb2","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":21,"title":"","display":"","copyAsset":false,"role":"supplement","size":1885,"visible":true,"origin":"","legend":"Table S29","description":"","filename":"TableS29.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/d842d2a4a80d94b30327a650.csv"},{"id":104859832,"identity":"daf8c9c7-5a64-48d1-bd4f-a587ecb39c16","added_by":"auto","created_at":"2026-03-18 04:56:14","extension":"csv","order_by":22,"title":"","display":"","copyAsset":false,"role":"supplement","size":6553,"visible":true,"origin":"","legend":"Table S27","description":"","filename":"TableS27.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/fcbb9074b5cfe614186dda68.csv"},{"id":105034090,"identity":"d689c774-5518-4bf0-ab26-380aa06f17fe","added_by":"auto","created_at":"2026-03-20 07:22:38","extension":"csv","order_by":23,"title":"","display":"","copyAsset":false,"role":"supplement","size":3789,"visible":true,"origin":"","legend":"Table S1","description":"","filename":"TableS1.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/10fda043258bdd7980314fb5.csv"},{"id":104859827,"identity":"875a44bb-7359-4c46-b37c-f8dda6e948b2","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":24,"title":"","display":"","copyAsset":false,"role":"supplement","size":806,"visible":true,"origin":"","legend":"Table S12","description":"","filename":"TableS12.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/a6ebe67168531da0e6678ef6.csv"},{"id":105034088,"identity":"5748d489-2b3d-4433-9d64-9a8fccf38fbc","added_by":"auto","created_at":"2026-03-20 07:22:38","extension":"csv","order_by":25,"title":"","display":"","copyAsset":false,"role":"supplement","size":9387,"visible":true,"origin":"","legend":"Table S32","description":"","filename":"TableS32.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/ddfce92b456449510a81be6a.csv"},{"id":104859821,"identity":"eb8dc6e1-8ee0-4b2f-894f-c1b90f3e20fb","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":26,"title":"","display":"","copyAsset":false,"role":"supplement","size":21169,"visible":true,"origin":"","legend":"Table S11","description":"","filename":"TableS11.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/ff92f59770ad0c5bf62309ff.csv"},{"id":104859823,"identity":"b9904982-17e8-4d9a-9abc-16deca9df6ff","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":27,"title":"","display":"","copyAsset":false,"role":"supplement","size":1333,"visible":true,"origin":"","legend":"Table S13","description":"","filename":"TableS13.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/7b3b551721aa588ca6d26058.csv"},{"id":104859818,"identity":"904f9e60-be52-4514-bc18-451ebcf25d20","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":28,"title":"","display":"","copyAsset":false,"role":"supplement","size":21979,"visible":true,"origin":"","legend":"Table S10","description":"","filename":"TableS10.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/652549f9b959ed5e7059e234.csv"},{"id":105033873,"identity":"37c61b0a-12d5-44a3-8644-0942afc6a17f","added_by":"auto","created_at":"2026-03-20 07:22:00","extension":"csv","order_by":29,"title":"","display":"","copyAsset":false,"role":"supplement","size":46349,"visible":true,"origin":"","legend":"Table S17","description":"","filename":"TableS17.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/2ec4813cfab6272a77e372ca.csv"},{"id":104859828,"identity":"85f59456-1b69-4517-9873-5119c4e485db","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"docx","order_by":30,"title":"","display":"","copyAsset":false,"role":"supplement","size":25990,"visible":true,"origin":"","legend":"Supplemantary information","description":"","filename":"Supplemantaryinformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/6b921e4b8e7948a60d631b7b.docx"},{"id":105034144,"identity":"d5554c65-97ae-4ec7-a561-0048b99f4118","added_by":"auto","created_at":"2026-03-20 07:22:45","extension":"csv","order_by":31,"title":"","display":"","copyAsset":false,"role":"supplement","size":9999,"visible":true,"origin":"","legend":"Table S26","description":"","filename":"TableS26.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/8abfd0df35e19707daab9b45.csv"},{"id":104859825,"identity":"25184cfe-2537-4c2a-9d76-082500ba0845","added_by":"auto","created_at":"2026-03-18 04:56:13","extension":"csv","order_by":32,"title":"","display":"","copyAsset":false,"role":"supplement","size":942,"visible":true,"origin":"","legend":"Table S19","description":"","filename":"TableS19.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/2c4310a75e11bda0f635401a.csv"},{"id":105033895,"identity":"d6ff7ed4-1422-4fd7-8046-73ddd7aa568e","added_by":"auto","created_at":"2026-03-20 07:22:05","extension":"csv","order_by":33,"title":"","display":"","copyAsset":false,"role":"supplement","size":12479,"visible":true,"origin":"","legend":"Table S30","description":"","filename":"TableS30.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/eb14c9fe9c26acbf5c0d856f.csv"},{"id":104859829,"identity":"f6f64765-ceb7-4102-89c1-a885c76b4ac3","added_by":"auto","created_at":"2026-03-18 04:56:14","extension":"csv","order_by":34,"title":"","display":"","copyAsset":false,"role":"supplement","size":2308,"visible":true,"origin":"","legend":"Table S6","description":"","filename":"TableS6.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/616c2df00383d97ad5d7b624.csv"},{"id":105034089,"identity":"b319de25-08ae-4da0-9117-d3640097604e","added_by":"auto","created_at":"2026-03-20 07:22:38","extension":"csv","order_by":35,"title":"","display":"","copyAsset":false,"role":"supplement","size":16995,"visible":true,"origin":"","legend":"Table S22","description":"","filename":"TableS22.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/6449078bbf35f116d5608d8e.csv"},{"id":104859830,"identity":"36c9b200-c91e-41e4-8f5e-00ca82fd7eda","added_by":"auto","created_at":"2026-03-18 04:56:14","extension":"csv","order_by":36,"title":"","display":"","copyAsset":false,"role":"supplement","size":6597,"visible":true,"origin":"","legend":"Table S15","description":"","filename":"TableS15.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/f93cc3f61615b66d3b45855f.csv"},{"id":105034082,"identity":"047efcc0-e2a3-4377-a068-73becc586132","added_by":"auto","created_at":"2026-03-20 07:22:36","extension":"csv","order_by":37,"title":"","display":"","copyAsset":false,"role":"supplement","size":1071,"visible":true,"origin":"","legend":"Table S21","description":"","filename":"TableS21.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/b357d0cce7f5538d3cefec65.csv"},{"id":104859831,"identity":"544e627e-65f7-456b-ab6b-7a19e53e3253","added_by":"auto","created_at":"2026-03-18 04:56:14","extension":"csv","order_by":38,"title":"","display":"","copyAsset":false,"role":"supplement","size":7962,"visible":true,"origin":"","legend":"Table S33","description":"","filename":"TableS33.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/4d8ff7d3b6ac9818f7b5e606.csv"},{"id":105033804,"identity":"8db368eb-fd89-45dc-9207-95430a706467","added_by":"auto","created_at":"2026-03-20 07:21:45","extension":"docx","order_by":39,"title":"","display":"","copyAsset":false,"role":"supplement","size":4132823,"visible":true,"origin":"","legend":"Supplementary Figures","description":"","filename":"SupplementaryFigures.docx","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/6b7fcdd97b92c5d56a269a7b.docx"},{"id":105034081,"identity":"257676cb-c0d7-4d07-8ea4-42a4db065600","added_by":"auto","created_at":"2026-03-20 07:22:36","extension":"csv","order_by":40,"title":"","display":"","copyAsset":false,"role":"supplement","size":1293,"visible":true,"origin":"","legend":"Table S7","description":"","filename":"TableS7.csv","url":"https://assets-eu.researchsquare.com/files/rs-8943594/v1/0cdc046ef464a0421b1a7414.csv"}],"financialInterests":"The authors have declared there is \u003cb\u003eNO\u003c/b\u003e conflict of interest to disclose","formattedTitle":"Whole-genome sequencing of mixed OCD–schizophrenia pedigrees characterizes shared and divergent rare-variant architectures","fulltext":[{"header":"Introduction","content":"\u003cp\u003ePsychiatric diagnoses are largely defined by descriptive symptom constellations rather than etiology, and substantial symptom overlap and comorbidity are common across disorders (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Although recent nosologies increasingly incorporate dimensional elements, such as in DSM-5 and ICD-11 (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e), a central barrier to precision psychiatry remains: we still lack a mechanistic understanding of why clinically distinct disorders share symptoms, risk factors, and familial liability. Convergent genetic and biological evidence suggests that many psychiatric conditions partially share molecular underpinnings, yet the degree of overlap, and critically, the mechanisms that bias individuals toward one diagnostic outcome versus another, remain incompletely characterized (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Defining such convergent versus divergent biology is not only essential for refining disease models but also provides a rational basis for improving early stratification and targeted intervention in high-burden families (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e). Motivated by this gap, we initiated the \u003cb\u003eCo\u003c/b\u003emplex \u003cb\u003eF\u003c/b\u003eamily Project involving \u003cb\u003eFi\u003c/b\u003erst-Degree \u003cb\u003eR\u003c/b\u003eelatives (CoFFiR), which recruits mixed-pedigree families affected by multiple severe mental disorders to generate family-grounded clinical and genetic evidence. In this first stage, we focus on genetic architecture using deep whole-genome sequencing as a foundation for disentangling shared liability from diagnosis-biased signals.\u003c/p\u003e \u003cp\u003eObsessive\u0026ndash;compulsive disorder (OCD) and schizophrenia (SCZ) offer a clinically relevant and biologically informative setting to study such within-family diagnostic divergence. SCZ affects\u0026thinsp;~\u0026thinsp;1% of the population (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) and OCD affects\u0026thinsp;~\u0026thinsp;2\u0026ndash;3% (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e), and OCD/obsessive\u0026ndash;compulsive symptoms are frequently observed in individuals with SCZ (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e). Reported prevalence estimates suggest that OCD occurs in ~\u0026thinsp;12.1\u0026ndash;13.6% of SCZ patients (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e), and obsessive\u0026ndash;compulsive symptoms in SCZ can reach\u0026thinsp;~\u0026thinsp;30.7% (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e). Both disorders are complex and heterogeneous, influenced by genetic and environmental factors, with heritability estimates of ~\u0026thinsp;80% for SCZ (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e) and 26\u0026ndash;45% for OCD (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Longitudinal and family studies further support bidirectional aggregation: individuals with OCD show an elevated risk of subsequently developing SCZ (IRR up to 6.9; 11), and 7.8% of OCD patients developed SCZ over an 11-year follow-up (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e). Moreover, multigenerational family studies indicate increased schizophrenia risk among relatives of OCD probands, decreasing with genetic distance (\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e). Together, these observations support partially shared genetic components while highlighting an unresolved question of direct relevance to CoFFiR: under a largely shared familial background, what genetic signals are shared across OCD and SCZ, and what signals bias risk toward one diagnostic outcome versus the other?\u003c/p\u003e \u003cp\u003eGenetic studies have made substantial progress in mapping shared and disorder-specific risk. Genome-wide association studies (GWAS) have identified many common-risk loci across psychiatric disorders, yet the gap between SNP-based and epidemiological heritability underscores \u0026ldquo;missing heritability\u0026rdquo; and motivates attention to rarer variants (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e). Next-generation sequencing (NGS) has therefore become central to psychiatric genetics, particularly for identifying rare coding variation and probing regulatory mechanisms. For OCD, rare-variant studies remain comparatively limited: a large family-based whole-exome sequencing study implicated de novo mutations and prioritized candidate genes (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e), and our prior work using family-based whole-genome sequencing supported an excess burden of rare de novo mutations and suggested potential involvement of regulatory elements in non-coding regions relevant to OCD etiology (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e). In contrast, sequencing studies in SCZ are more extensive, spanning candidate gene discovery, mechanistic insights, cross-ancestry analyses, and broader characterization of genomic architecture (\u003cspan additionalcitationids=\"CR17 CR18 CR19 CR20 CR21 CR22\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e). Despite these advances, cross-disorder sequencing designs that can directly interrogate within-family diagnostic divergence remain scarce.\u003c/p\u003e \u003cp\u003eA key reason is methodological: conventional case\u0026ndash;control designs typically require very large sample sizes to overcome heterogeneity and background noise, and they are inherently limited in resolving within-family divergence when distinct diagnoses arise within the same pedigree. Family-based designs offer complementary strengths by enriching rare risk alleles, reducing confounding from population stratification, and leveraging shared background to highlight diagnosis-biased signals (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e). Yet such approaches remain underutilized in cross-disorder psychiatric genetics, particularly in designs that co-ascertain multiple disorders within the same families (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo our knowledge, few whole-genome sequencing (WGS) studies have leveraged mixed pedigrees co-ascertained for obsessive\u0026ndash;compulsive disorder (OCD) and schizophrenia (SCZ) to interrogate within-family diagnostic divergence. Here, we leverage such mixed pedigrees in which OCD and SCZ co-occur to address a question that case\u0026ndash;control designs are inherently underpowered to resolve: why distinct psychiatric diagnoses emerge under a largely shared familial genetic background. We performed deep (60\u0026times;) WGS in 21 OCD\u0026ndash;SCZ complex families and integrated rare coding and regulatory variation to (i) prioritize co-segregating and disorder-biased burden patterns, (ii) evaluate disorder separation under a family-aware validation scheme, and (iii) map convergent versus divergent signals onto spatiotemporal neurodevelopmental modules to nominate candidate vulnerability windows and cell-type programs.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eSubjects\u003c/h2\u003e \u003cp\u003eTwenty-one families of Chinese Han population with OCD, SCZ patients and healthy controls were recruited for WGS. Each family included one OCD patient, one SCZ patient that met DSM- IV criteria and at least one unaffected family member, resulting in a cohort total of 21 SCZ patients, 21 OCD patients and 38 unaffected controls. All volunteers within the same family were either direct relatives or collateral relatives within three generations.\u003c/p\u003e \u003cp\u003eStrict quality control was applied during the process of sample collection. All OCD patients were diagnosed by senior attending psychiatrists or chief psychiatrists. Patients were excluded if they met DSM-IV criteria for any disorders other than OCD or SCZ. The International Neuropsychiatric Interview (M.I.N.I.) was used to screen for DSM-IV Axis I psychiatric diagnoses.\u003c/p\u003e \u003cp\u003eSocio-demographic and additional clinical information were collected using a semi-structured interview design by our team. The Yale-Brown Obsessive-compulsive Scale (Y-BOCS) was used to assess OCD symptom severity(\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e). The Positive and Negative Syndrome Scale (PANSS) was used to assess schizophrenia symptom severity(\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e). The Hamilton Anxiety Scale (HAMA)(\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e) and Hamilton Depression Scale (HAMD) (\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e)were used to assess mood status, such as anxiety and depressive symptoms, respectively. All assessments were conducted by raters trained for this study.\u003c/p\u003e \u003cp\u003e This study was approved by the Ethics Committee of Shanghai Mental Health Center, School of Medicine, Shanghai Jiao Tong University. Written informed consent was obtained from each participant.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eDNA Sequencing\u003c/h3\u003e\n\u003cp\u003eDNA was extracted from whole blood. DNA quality was assessed using gel electrophoresis and its concentration was measured using the Qubit Fluorometer. DNBSEQ library (paired-end 150 bp) was constructed and sequencing at 60\u0026times; coverage was performed using the BGI DNBSEQ platform. Sequencing-derived raw image files were processed by DNBSEQ basecalling software for base-calling with default parameters, and the sequence data for each individual were generated as paired-end reads.\u003c/p\u003e\n\u003ch3\u003eCuration of sequencing data and variant calling\u003c/h3\u003e\n\u003cp\u003eSOAPnuke v2.1.0 was used for quality control of the raw sequencing data(\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e). The clean data of each sample were mapped to GRCh37 to obtain an initial alignment file in BAM format using Burrows-Wheeler Aligner (BWA) v0.7.17(\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e). Samtools v1.3.1 was used to sort and index the SAM files(\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e). Genome Analysis ToolKit (GATK) v4.1.4.1 was used in the following variant calling. MarkDuplicates was used to mark the duplicate reads, which were ignored in downstream analysis. BaseRecalibrator and ApplyBQSR were used to correct the base quality values. HaplotypeCaller was used to simultaneously detect SNPs and insertion-deletions (InDels). The sample-level variant calling results were stored in gVCF files and subsequent genotyping was conducted using GenotypeGVCFs. After single nucleotide variations (SNVs) and InDels were selected separately for downstream pipelines via SelectVariants, Variant Quality Score Recalibration (VQSR) was used to filter out false mutations. Only the variants indicated as \"PASS\" were considered credible variant sets and included in following analysis.\u003c/p\u003e\n\u003ch3\u003eSNV/Indel annotation\u003c/h3\u003e\n\u003cp\u003eWe used ANNOVAR (version 2023-08-30) to annotate the SNVs(\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e). Allele frequency (AF) was annotated based on 1000 Genomes (1000G) and gnomAD (version 2.1.1). Only the variants with the minor allele frequency (MAF)\u0026thinsp;\u0026le;\u0026thinsp;0.005 in databases including 1000G_EAS, 1000G_ALL, gnomAD_EAS and gnomAD_ALL were filtered into the following analyses. The functional categories of the SNPs in coding regions were predicted based on RefSeq annotation of hg19 (version 2020-08-17). We annotated the known SNVs in coding regions with ClinVar (released 2024-04-26). Besides, we further assessed the pathogenicity of variants via Franklin (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://franklin.genoox.com\u003c/span\u003e\u003cspan address=\"https://franklin.genoox.com\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), pLI (probability of loss-of-function intolerance) score(\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e), combined annotation dependent depletion (CADD) v1.7(\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e) as well as MPC(\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e), MutationTaster(\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e), LRT(\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e) and phastCons100way_vertebrate(\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e) scores included in dbNSFP v4.1(\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe putative pathogenic variants were manually inspected employing visualization of aligned reads using Integrative Genomics Viewer (IGV)(\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e). For non-coding region variant annotation, the experiment-derived regulatory features were acquired from PsychEncode(\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e), which provided the atlas of some regulatory elements of mental disorders. We also employed chromatin states annotations including Roadmap core 15-state model (epigenome data from Brain Dorsolateral Prefrontal Cortex)(\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e) and Sei model(\u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e), which combined several epigenomic marks in a spatial context, to accurately capture the potential epigenomic function of the variant position.\u003c/p\u003e\n\u003ch3\u003eStructural variation /copy number variation annotation\u003c/h3\u003e\n\u003cp\u003eBreakDancer v1.4.5(\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e) was used to detect structural variations (SVs) using default parameter. CNVnator v0.3.2 (\u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e)was used to detect copy number variations (CNVs). The process used standard parameters and settings, and a window length of 100 bp is selected. Ensembl Variant Effect Predictor (VEP) (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e) and annotSV (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e) were used to annotate the SV/CNV results(\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e).\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eCo-segregation and disease-specific variant/gene lists\u003c/h2\u003e \u003cp\u003eIn protein-coding regions, the filtered variants were divided into co-segregation variants and disease (OCD or SCZ) specific variants: co-segregation variants were defined as variants detected in both OCD and SCZ patients but not present in any of the controls in the same family; disease-specific variants were defined as variants detected in the patient with the specific disease (OCD or SCZ) but not present in any other family member in the same family. These variants were mapped to the corresponding genes based on RefSeq to imply putative pools of candidate genes in our data.\u003c/p\u003e \u003cp\u003eIn non-coding regions, to focus on the candidate genes more efficiently, co-segregation variants were enriched by ruling out the variants present in any one of the controls in the whole cohort. Correspondingly, disease-specific variants were enriched by ruling out the variants present in any unaffected individual or individual with the other disorder in the whole cohort. Enriched gene lists were mapped from the enriched variant list.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eVirtual panel genes/regions from publicly available data\u003c/h3\u003e\n\u003cp\u003eWe introduced three disease-related virtual panels from publicly available data.\u003c/p\u003e \u003cp\u003eTo focus on the protein-coding genes detected in brain, we introduced \u0026lsquo;protein-coding genes expressed in brain\u0026rsquo; gene list obtained from The Human Protein Atlas (HPA) (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e)(n\u0026thinsp;=\u0026thinsp;15331).\u003c/p\u003e \u003cp\u003eThe etiology virtual panel including common psychiatric-disorder-related etiological gene sets available from publications: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) RBFOX target genes(\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e51\u003c/span\u003e); (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) Nervous developmental disorder genes(\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e52\u003c/span\u003e); (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) LOEUF_TOP20% genes(\u003cspan citationid=\"CR53\" class=\"CitationRef\"\u003e53\u003c/span\u003e); (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) hPSD genes(\u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e54\u003c/span\u003e); (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e) FMRP target genes(\u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e55\u003c/span\u003e); (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e) genes related to chromatin modifiers(\u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e56\u003c/span\u003e); (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e) CHD8 target genes(\u003cspan citationid=\"CR57\" class=\"CitationRef\"\u003e57\u003c/span\u003e); (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e) reported ASD candidate genes(\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e58\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThe GWAS_OCD/SCZ virtual panel was calculated from database: Variants with p\u0026thinsp;\u0026lt;\u0026thinsp;1e-5 from OCD GWAS (\u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e59\u003c/span\u003e) and variants with p\u0026thinsp;\u0026lt;\u0026thinsp;5e-8 from SCZ GWAS(\u003cspan additionalcitationids=\"CR61 CR62 CR63\" citationid=\"CR60\" class=\"CitationRef\"\u003e60\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e64\u003c/span\u003e) were selected, resulted in 780 variants. Additionally, Clinvar SNVs related to OCD/SCZ, which was in putative risk OCD/SCZ regions nominated by PsychENCODE (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://resource.psychencode.org/\u003c/span\u003e\u003cspan address=\"http://resource.psychencode.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and OMIM (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.\u003c/span\u003e\u003cspan address=\"https://www.\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003eomim.org/\u003c/span\u003e\u003cspan address=\"http://omim.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), were added(n\u0026thinsp;=\u0026thinsp;2358). The combined SNV lists including 3138 SNVs were expanded to all SNVs within the LD block via Haploreg V4.2(65) (r2\u0026thinsp;\u0026gt;\u0026thinsp;0.8), resulted in 16235 SNVs. The SNVs were next filtered by DNase I hypersensitive (DHS) peaks in GTEx (including brain-sourced data: cerebellar_cortex, dorsolateral_prefrontal_cortex, frontal_cortex, globus_pallidus, head_of_caudate_nucleus, posterior_cingulate_gyrus and putamen; immune-related data: B_cell, CD14_positive_monocyte, CD4_positive_alpha-beta_T_cell, common_myeloid_progenitor_CD34_positive, naive_B_cell, naive_thymus_derived_CD4-positive_alpha-beta_T_cell, natural_killer_cell, T_cell and T_helper_cell; neuron-sourced data: choroid plexus epithelial cell, smooth muscle cell of the brain, astrocyte of the cerebellum, brain pericyte primary cell, SK-N-DZ cell line, M059J cell line, brain microvascular endothelial cell, BE2C cell line, astrocyte of the hippocampus, Daoy cell line, bipolar neuron in vitro differentiated cells, astrocyte of the spinal cord, and astrocyte primary cell) (n\u0026thinsp;=\u0026thinsp;5713) and intersected with eQTL locus from GTEx and PsychENCODE. The final eQTL variant list (n\u0026thinsp;=\u0026thinsp;5682) were mapped to genes based on Refseq(n\u0026thinsp;=\u0026thinsp;486).\u003c/p\u003e \u003cp\u003eOverrepresentation analysis was performed to determine if the overlap between two gene sets was significantly higher than might occur by chance. This analysis was done using the \u0026ldquo;enrichment\u0026rdquo; function of the R package clusterProfiler(\u003cspan citationid=\"CR66\" class=\"CitationRef\"\u003e66\u003c/span\u003e, \u003cspan citationid=\"CR67\" class=\"CitationRef\"\u003e67\u003c/span\u003e).\u003c/p\u003e\n\u003ch3\u003eCoding variants classification and analysis\u003c/h3\u003e\n\u003cp\u003eFor SNVs, variants were divided into different types for categorized analyses: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) Loss of function variants (LOFs), any variants that introduced a stop codon, a shift of the open reading frame or a change at a predicted splice site. (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) Missense variants (MISs), any single-nucleotide variants changing the amino acid. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) synonymous variants, any variants that resulted in none amino acid change. The functional consequences were determined by RefSeq.\u0026nbsp;We further divided LOFs and MISs into following groups according to the pathogenicity score: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) LOFs: LofA (pLI\u0026thinsp;\u0026gt;\u0026thinsp;0.9), LofB (pLI\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;0.5), LofC (pLI\u0026thinsp;\u0026lt;\u0026thinsp;0.5); (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) MISs: MisA (CADD\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;25 or MPC\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;2), MisB (15\u0026thinsp;\u0026lt;\u0026thinsp;CADD\u0026thinsp;\u0026lt;\u0026thinsp;25 or 1\u0026thinsp;\u0026lt;\u0026thinsp;=\u0026thinsp;MPC\u0026thinsp;\u0026lt;\u0026thinsp;2), MisC (CADD\u0026thinsp;\u0026lt;\u0026thinsp;15 and MPC\u0026thinsp;\u0026lt;\u0026thinsp;1). Among all the groups, LofA and MisA variants were defined as high-impact SNVs, and those located on autosomes were filtered into the following statistical analyses.\u003c/p\u003e \u003cp\u003eFor Indels, as CADD annotation was inadequate, variants were additionally annotated with VEP. Those indels with CADD\u0026thinsp;\u0026gt;\u0026thinsp;25 and those with unknown CADD score but with Ensembl Variant Effect Predictor (VEP) annotation of \u0026ldquo;HIGH\u0026rdquo; were categorized as high-impact.\u003c/p\u003e \u003cp\u003eFor SV/CNVs, variants were annotated with AnnotSV and high-impact SVs were defined as: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) With AF\u0026thinsp;\u0026lt;\u0026thinsp;0.005; (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) Overlap with protein coding regions; (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) 1) Ranked by AnootSV as 4\u0026ndash;5 in pathogenicity (Likely pathogenic/ Pathogenic) OR 2) a. Ranked by AnnotSV as 3 in pathogenicity (VUS) AND b. not overlapped with any of the known benign region ((po_) B_gain, (po_) B_loss, (po_) B_ins, (po_) B_inv). High-impact SV/CNVs were presented to observe the distribution of important variants.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eNon-coding variants analysis\u003c/h2\u003e \u003cp\u003eWe focus the variants in non-coding regions located in the non-coding co-segregation and disease-specific genes. Fisher\u0026rsquo;s exact tests were performed to detect the enrichment of regulatory elements on autosomes between groups. P threshold was set as 0.05. FDR threshold was set at 0.2.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eAssociation analysis of rare variants\u003c/h2\u003e \u003cp\u003eConcerning the difficulty of association analysis of rare variants in small-sized family samples(\u003cspan citationid=\"CR68\" class=\"CitationRef\"\u003e68\u003c/span\u003e, \u003cspan citationid=\"CR69\" class=\"CitationRef\"\u003e69\u003c/span\u003e), we combined three common gene-based burden analysis approaches, including Fisher\u0026rsquo;s test, sequence kernel association test (SKAT-O) (\u003cspan citationid=\"CR70\" class=\"CitationRef\"\u003e70\u003c/span\u003e) and aggregated Cauchy association test-omnibus(ACAT-O) test(\u003cspan citationid=\"CR71\" class=\"CitationRef\"\u003e71\u003c/span\u003e) to detect all the associations contained in the data on autosomes. Only genes that passed Bonferroni correction in any of the test or genes passed all three tests with raw p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 were reported. In protein coding regions, region-based burden test was implemented in LOFs and MisA variants. In non-coding regions, region-based burden test was performed in possible functional variants, i.e. non-intergenic and non-intronic variants, based on specific gene regions including co-segregation and disease-specific genes.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eProtein-protein interaction effects evaluation\u003c/h2\u003e \u003cp\u003eWhile protein\u0026ndash;protein interactions (PPI) influence cellular functions and biological activities in a fundamental way, we evaluated the variant influence on PPI with MIPPI(\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e). The high-impact missense variants were collected and grouped into SCZ-specific, OCD-specific and control-specific groups. The grouped variants were then mapped to PPI partners of homo sapiens from BioGrid database v4.4.234(\u003cspan citationid=\"CR73\" class=\"CitationRef\"\u003e73\u003c/span\u003e) to predict the category of mutation effect.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eSelection of high-confidence genes\u003c/h2\u003e \u003cp\u003eTo construct the most affected network in specific diagnosis group, we selected and explicitly reported the high-confidence variants according to the following criteria:\u003c/p\u003e \u003cp\u003eFor SNVs: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) Being the co-segregated variants or disease-specific variants; (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) 1) Being LofA of MisA; 2) Met 2\u0026ndash;3 following criteria: a. predicted by LRT as \u0026ldquo;U\u0026rdquo; or \u0026ldquo;D\u0026rdquo;; b. predicted by MutationTaster as \u0026ldquo;A\u0026rdquo; or \u0026ldquo;D\u0026rdquo;; c. phastCons100wat_vertebrate_score\u0026thinsp;\u0026gt;\u0026thinsp;=\u0026thinsp;0.990. The high-confidence SNVs identified were included in the following network analyses.\u003c/p\u003e \u003cp\u003eFor SVs: (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e) Being the co-segregated SVs or disease-specific SVs; (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e) With AF\u0026thinsp;\u0026lt;\u0026thinsp;0.005; (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e) Overlap with protein coding regions; (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e) Ranked by AnootSV (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e, \u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e) as 4\u0026ndash;5 in pathogenicity (Likely pathogenic/ Pathogenic).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eNetwork construction\u003c/h2\u003e \u003cp\u003eThe high-confidence genes derived from high-confidence SNVs and listed in the \u0026lsquo;protein coding genes expressed in brain\u0026rsquo; genes from HPA atlas were selected to build the PPI network.\u003c/p\u003e \u003cp\u003ePPI networks were constructed by STRING v12.0 (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cn.string-db.org/\u003c/span\u003e\u003cspan address=\"https://cn.string-db.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) database followed by analyses via Cytoscape 3.10.2(\u003cspan citationid=\"CR74\" class=\"CitationRef\"\u003e74\u003c/span\u003e). Only the direct interactions between seed nodes were presented. Further visualization was done by R package \u0026lsquo;ggnetwork\u0026rsquo;.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eFunctional enrichment\u003c/h2\u003e \u003cp\u003eTo have a more comprehensively landscape of the target genes as far as possible, we performed functional enrichment analyses by Metascape v3.5.20240101(\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e), which can incorporate results from several databases at a time and make comparison between two groups. P threshold was set as 0.05 and FDR threshold was set at 0.05. The enriched terms were manually clustered according to biological relatedness, enabling an overall landscape of the biological functions.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eSingle Cell Expression Analysis\u003c/h2\u003e \u003cp\u003eFor specific hub genes, we obtained the single cell expression data from The Allen Brain Cell Atlas (\u003cspan citationid=\"CR76\" class=\"CitationRef\"\u003e76\u003c/span\u003e), which allowed comprehensive subclusters combining cell types and anatomies, and we performed visualization based on its suggested pipeline.\u003c/p\u003e \u003cp\u003eFor cluster of target genes, we employed cell type enrichment analysis via R \u0026lsquo;Expression Weighted Cell Type Enrichment\u0026rsquo; (EWCE) package(\u003cspan citationid=\"CR77\" class=\"CitationRef\"\u003e77\u003c/span\u003e). The newly published single-nucleus RNA sequencing (snRNA-seq) data from psychencode was used to perform EWCE(\u003cspan citationid=\"CR78\" class=\"CitationRef\"\u003e78\u003c/span\u003e).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eWeighted correlation network analysis and Developmental Trajectory Analyses\u003c/h2\u003e \u003cp\u003eAs the frontal cortex played a critical role in both SCZ and OCD(\u003cspan citationid=\"CR79\" class=\"CitationRef\"\u003e79\u003c/span\u003e), we made Weighted correlation network analysis (WGCNA) to distinguish different expression patterns between diseases using data from frontal cortex. Expression matrix of developmental transcriptome of frontal cortex from BrainSpan (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.brainspan.org\u003c/span\u003e\u003cspan address=\"https://www.brainspan.org\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was used for WGCNA (\u003cspan citationid=\"CR80\" class=\"CitationRef\"\u003e80\u003c/span\u003e). Genes with mean RPKM (Reads Per Kilobase of transcript per Million mapped reads) value\u0026thinsp;\u0026lt;\u0026thinsp;1 across all the developmental stages were discarded. After the detection of co-expression modules, we tested whether specific gene sets were enriched in any of the modules by Fisher\u0026rsquo;s exact test. The enriched modules were used for the enrichment analysis via Metascape. We also divided the developmental stages as early-pregnancy (0\u0026ndash;8 post-conception weeks, pcw), mid-pregnancy (8\u0026ndash;26 pcw), late-pregnancy (26\u0026ndash;37 pcw) and after-birth to detect the expression and functional characteristics in each developmental stage.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eRandom Forest classifier\u003c/h2\u003e \u003cp\u003eTo classify individuals based on their mutation profiles, we implemented a Random Forest classifier with default hyperparameters. Model performance was rigorously evaluated using a 5-fold stratified group cross-validation protocol, where family identifiers served as the grouping variable to prevent data leakage between folds, ensuring a realistic estimate of the model's ability to generalize to new families. This stratification maintained a consistent case-control ratio across all splits. Within each fold, we assessed performance using the Area Under the Receiver Operating Characteristic Curve (AUC) and Matthews Correlation Coefficient (MCC). For each classification task, we visualized overall performance by plotting the macro-average True Positive Rate (TPR) with a standard deviation confidence band and calculated a comprehensive micro-AUC score from aggregated out-of-fold (OOF) predictions. To interpret the model and identify key predictive features, we employed SHAP (SHapley Additive exPlanations). This analysis was conducted strictly within the cross-validation loop, calculating SHAP values exclusively on the held-out test sets to avoid bias from the training data. The global importance of each feature was determined by its mean absolute SHAP value, and we identified the most consistently impactful predictors by selecting the top 10 features that maintained a high importance ranking across all five cross-validation folds.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eFamily-based design enables within-family risk stratification\u003c/h2\u003e \u003cp\u003eIn the first stage of the CoFFiR project (CoFFiR-Ⅰ), we conducted WGS research on OCD-SCZ complex- families, which included 21 SCZ patients, 21 OCD patients and 38 unaffected family members (Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e Methods). As shown in Table\u0026nbsp;1, PANSS and Y-BOCS scores were different between OCD, SCZ, and control groups (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Then we performed WGS at 60\u0026times; coverage on the SCZ-OCD complex-families. The average coverage of the sequencing was 99.57%, with 97.5% of bases covered above 20\u0026times;. 4,116,317 SNVs and 1,026,835 Indels were detected and only those identified as \u0026ldquo;PASS\u0026rdquo; by VQSR were considered for downstream analysis. As part of sample quality control, we performed principal components analysis to assess the population stratification of our cohort. The results showed that the samples were primarily stratified based on families and regions, rather than diagnoses (Fig.\u0026nbsp;2a).\u003c/p\u003e \u003cp\u003eFor protein-coding regions, variants were stringently filtered based on the criteria detailed in the Methods section. Specifically, single nucleotide variants (SNVs) were filtered by allele frequency (AF)\u0026thinsp;\u0026lt;\u0026thinsp;0.005 in GnomAD_ALL/EAS and 1000G_ALL/EAS, resulting in 368, 382, 379 variants per sample in OCD, SCZ and control group, respectively(Fig.\u0026nbsp;2b). Next, we classified the LOF variants and missense variants by pLI score and MPC/CADD, separately (see Methods and Fig.\u0026nbsp;2b). LofAs and MisAs were designated as high-impact SNVs. Besides, InDels were grouped by CADD phred score and VEP impact annotation. SVs/CNVs were assessed by annotSV (see Methods). The distribution of high-impact variants of all variant types over the whole genome was presented in Fig.\u0026nbsp;2d. The variants were validated manually using Integrative Genomics Viewer (IGV) (Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e-\u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e). Besides, for non-coding variants, we used the filter criteria described in the Methods and detected 37796, 40628 and 39255 SNVs per sample for OCD, SCZ, and control group, respectively (Fig.\u0026nbsp;2f).\u003c/p\u003e \u003cp\u003eTo verify our variant selection and expand the scope of understanding towards the genetic contributions, we curated a SCZ/OCD gene panel by using publicly available GWAS summary statistics (see Methods, Figure \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003ea, Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). We performed enrichment analysis based on KEGG, Gene Ontology (GO), along with selected gene sets, to identify the implications of highlighted pathways and biological functions. These genes were widely distributed in terms related to metabolism, synapse and neurotransmitter, and were overrepresented in gene sets included RBFOX-targets, NDD genes, TOP 20% LOEUF genes, FMRP-targets and ASD-related candidate genes (Figure \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003eb, c, d, Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e-4). An integrated matrix of high-impact SNV/Indel variants was mapped to the panel genes (Fig.\u0026nbsp;2e). There were 2818 high-impact variants in 1887 genes detected in our cases (Table \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e) but only 86 variants in 64 genes were mapped to the panel (Figure \u003cspan refid=\"MOESM4\" class=\"InternalRef\"\u003eS4\u003c/span\u003ea, Fig.\u0026nbsp;2e). This suggests a large number of potentially novel candidate genes in our data.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eDesign of the disease-specific genetic model\u003c/h2\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eCoding-variant burden differs across within-family diagnostic groups\u003c/h2\u003e \u003cp\u003eWe used Fisher's exact test to observe the count distribution of variants between groups (Fig.\u0026nbsp;2c, Table \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e, details in supplementary note 1). While the divergent results implied the differences between diseases in complex families, we performed segregation analysis on our filtered variants to narrow the range of suspected genes and variants. We identified the variants of co-segregated, OCD-specific, SCZ-specific and control-specific categories within each family (see Methods). We annotated the SNVs and Indels with ClinVar database and the results were presented in Table\u0026nbsp;2 and Table \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e. We made comprehensive annotation for all kinds of variants, only SNVs in high-impact groups (LofA and MisA) were included in the following analyses (see Methods).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eRegulatory annotations show disorder-biased enrichment in candidate regions\u003c/h2\u003e \u003cp\u003eTo obtain a comprehensive view of the whole genome, non-coding variants were included in our analyses. We evaluated the distribution of functional elements in non-coding variants using experimentally derived functional element atlas from PsychEncode and the Refseq functional annotation. Chromatin state models including Roadmap core 15 models and Sei were employed to provide a comprehensive landscape (see Methods, Fig.\u0026nbsp;2a, Figure \u003cspan refid=\"MOESM5\" class=\"InternalRef\"\u003eS5\u003c/span\u003e, details in supplementary note 2). Only non-intronic and non-intergenic variants on autosomes were included. In the primary comparison across the whole genome, we didn\u0026rsquo;t observe a significant difference between groups (Fig.\u0026nbsp;2b, Table \u003cspan refid=\"MOESM8\" class=\"InternalRef\"\u003eS8\u003c/span\u003e). In order to limit the noise in unrelated genomic regions, we focused on comparison within different candidate gene panels including the non-coding co-segregation panel (CSncs)(Table \u003cspan refid=\"MOESM9\" class=\"InternalRef\"\u003eS9\u003c/span\u003e), non-coding OCD-specific panels (OSncs) (Table \u003cspan refid=\"MOESM10\" class=\"InternalRef\"\u003eS10\u003c/span\u003e) and non-coding SCZ-specific panels (SSncs) (Table \u003cspan refid=\"MOESM11\" class=\"InternalRef\"\u003eS11\u003c/span\u003e) (see Methods). From our results, while both OCD and SCZ manifested significant alteration in non-coding regions, OCD showed greater deficit in non-coding regulatory elements within putative risk regions than SCZ, especially in promoter regions and transcription-active regions.\u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003eFamily-aware prediction using burden features\u003c/h2\u003e \u003cp\u003eTo assess the predictive power of mutation burden in selected genomic regions for mental disorders, we developed a series of Random Forest classifiers (Fig.\u0026nbsp;3a, Table \u003cspan refid=\"MOESM12\" class=\"InternalRef\"\u003eS12\u003c/span\u003e). The models were trained on mutation counts from each volunteer (Figure \u003cspan refid=\"MOESM6\" class=\"InternalRef\"\u003eS6\u003c/span\u003e). We employed a StratifiedGroupKFold cross-validation strategy, using family identifiers as groups to prevent data leakage and ensure model generalizability. Our primary model, utilizing features from both coding and non-coding regions, demonstrated strong predictive capability in distinguishing affected individuals from healthy controls. The highest performance was achieved in the OCD vs. Control classification, with a mean AUC of 0.839\u0026thinsp;\u0026plusmn;\u0026thinsp;0.104. The model also effectively distinguished the combined disease cohort from controls (Disease vs. Control: AUC\u0026thinsp;=\u0026thinsp;0.792\u0026thinsp;\u0026plusmn;\u0026thinsp;0.076) and SCZ from controls (SCZ vs. Controls: AUC\u0026thinsp;=\u0026thinsp;0.749\u0026thinsp;\u0026plusmn;\u0026thinsp;0.141). As expected, differentiating between the two mental disorders proved more challenging (SCZ vs. OCD: AUC\u0026thinsp;=\u0026thinsp;0.631\u0026thinsp;\u0026plusmn;\u0026thinsp;0.135), though the performance remained above chance. To investigate the relative contribution of different genomic regions, we conducted an ablation analysis by training models exclusively on mutations within either coding or non-coding regions. The model using only coding-region features (six features) performed comparably to the full model. For instance, in the SCZ vs. Control classification, its performance was slightly higher (AUC\u0026thinsp;=\u0026thinsp;0.765\u0026thinsp;\u0026plusmn;\u0026thinsp;0.097). Conversely, the model trained solely on non-coding features (15 features) showed a marked decrease in predictive accuracy across all classification tasks. For example, its ability to distinguish the general disease cohort from controls dropped significantly (AUC\u0026thinsp;=\u0026thinsp;0.698\u0026thinsp;\u0026plusmn;\u0026thinsp;0.102), and performance for SCZ vs. Control fell to an AUC of 0.629\u0026thinsp;\u0026plusmn;\u0026thinsp;0.176. These results strongly suggest that mutations within the coding regions of the selected loci are the primary drivers of the observed classification performance, while the non-coding mutation burden in these specific areas shows limited incremental value beyond coding burden under current regulatory annotations for these conditions.\u003c/p\u003e \u003cp\u003eTo further interpret the feature contributions of the full model, we employed SHAP (SHapley Additive exPlanations). The analysis consistently identified MISs and LofC as the most impactful predictors across disease-versus-control comparisons. Specifically, a higher burden of these mutation types strongly increased the model's output towards a disease classification, thereby pinpointing the key genetic markers driving the predictions (Fig.\u0026nbsp;3b).\u003c/p\u003e \u003cp\u003eBased on our results, we hypothesized that differences in the genetic architecture among OCD, SCZ, and control groups could be leveraged to evaluate individual disease risk. To test this hypothesis, we calculated a comprehensive genetic score for OCD and SCZ using the detected variants for each participant within the same family (Table \u003cspan refid=\"MOESM13\" class=\"InternalRef\"\u003eS13\u003c/span\u003e). The comprehensive score was divided into two components: the OCD score and the SCZ score. To calculate the comprehensive score, we extracted all variants present in each individual from the following sets firstl: 1) Coding regions: variants from co-segregation (Set-Ccds), OCD-specific (Set-Ocds), co-segregation, and SCZ-specific (Set-Scds) panels. 2) Non-coding regions: variants from co-segregation (CSncs), OCD-specific (OSncs), co-segregation, and SCZ-specific (SSncs) panels. Then, each variant was mapped to its corresponding gene. Gene scores were assigned based on established gene-level metrics: the pLI score for coding regions(\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e), and the average non-coding GERP (ncGERP) score for non-coding regions(\u003cspan citationid=\"CR81\" class=\"CitationRef\"\u003e81\u003c/span\u003e). Variant scores were assigned to each variant according to the variant type. LofA-C and MisA-C in coding regions were assigned scores of 2, 1.5, 1.25, 1, 0.9, or 0.75, whereas non-coding variants annotated as 1_TssA, 4_Tx, 7_Enh, and 15_Quies were assigned scores of 0.5, 0.5, 0.5, and \u0026minus;\u0026thinsp;0.5, respectively (Table \u003cspan refid=\"MOESM13\" class=\"InternalRef\"\u003eS13\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eFor each gene, the gene\u0026ndash;variant score was calculated as the product of the gene score and the sum of its variant scores. The OCD comprehensive score for an individual was obtained by summing the gene\u0026ndash;variant scores from co-segregation and OCD-specific regions (both coding and non-coding). Similarly, the SCZ comprehensive score was derived from co-segregation and SCZ-specific regions (Table \u003cspan refid=\"MOESM13\" class=\"InternalRef\"\u003eS13\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eThis novel scoring framework effectively distinguished the three groups within each family (Fig.\u0026nbsp;5c, Table\u0026nbsp;3). A paired Wilcoxon test revealed that OCD patients had significantly higher weights for the OCD score compared to the SCZ score (median OCD score\u0026thinsp;=\u0026thinsp;40.13 vs. median SCZ score\u0026thinsp;=\u0026thinsp;31.79, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.009). In contrast, SCZ patients showed the opposite trend (median OCD score\u0026thinsp;=\u0026thinsp;28.12 vs. median SCZ score\u0026thinsp;=\u0026thinsp;51.79, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001). No significant difference was observed in the control group (median OCD score\u0026thinsp;=\u0026thinsp;25.60 vs. median SCZ score\u0026thinsp;=\u0026thinsp;25.28, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.122).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003eBiological implication of the genetic model\u003c/h2\u003e \u003cdiv id=\"Sec27\" class=\"Section4\"\u003e \u003ch2\u003eIntra-Family Co-segregation Pinpoints Shared and Disorder-Specific Genetic Hubs\u003c/h2\u003e \u003cp\u003eAs a supplement to the genetic model, we sought to infer etiological relevance from familial genetic data. Regarding the polygenic genomic architecture of OCD and SCZ, we speculate that these diseases may be associated with impairments in a particular functional network. MIPPI(\u003cspan citationid=\"CR72\" class=\"CitationRef\"\u003e72\u003c/span\u003e) was employed and the results suggested a potential larger-scale protein-protein network (PPI) network disturbance in the etiology in OCD and SCZ groups compared with control(Fig.\u0026nbsp;5a, Table \u003cspan refid=\"MOESM14\" class=\"InternalRef\"\u003eS14\u003c/span\u003e, details in supplementary note 3).\u003c/p\u003e \u003cp\u003eThen, we further investigated the shared and specific variants or functional networks of OCD and SCZ patients. Therefore, genes mapped from different high-impact variant groups were pooled to form the gene sets including co-segregated (HI_CScds), OCD-specific (HI_OScds), and SCZ-specific (HI_SScds) gene panels within the protein-coding regions (see Methods, Fig.\u0026nbsp;3b). We performed an enrichment analysis by Metascape(\u003cspan citationid=\"CR75\" class=\"CitationRef\"\u003e75\u003c/span\u003e) on the HI_OScds and HI_SScds genes, to specify the probable characteristics of functional impairment related to each specific disease (Fig.\u0026nbsp;5c, Table \u003cspan refid=\"MOESM15\" class=\"InternalRef\"\u003eS15\u003c/span\u003e-16, details in Supplementary note 4).\u003c/p\u003e \u003cp\u003eThrough the divergent genetic landscape obtained from pathway analyses, we supposed that there were substantial differences etiology reflected by the intra-family phenotypic heterogeneity. To determine the most affected functional PPI network in each group, we selected the high-confidence variants with the strongest evidence of pathogenicity, for each group separately (see Methods, Table\u0026nbsp;4, Table \u003cspan refid=\"MOESM17\" class=\"InternalRef\"\u003eS17\u003c/span\u003e). These high-confidence SNVs from the co-segregation, OCD-specific, and SCZ-specific groups were selected to construct the STRING PPI network separately. Results showed that these networks manifested significantly more interactions than expected (co-segregation network: p\u0026thinsp;=\u0026thinsp;0.047; SCZ-specific network: p\u0026thinsp;=\u0026thinsp;0.0132; OCD-specific: p\u0026thinsp;=\u0026thinsp;0.0404). Hub genes of each network was identified to elucidate the core function of different networks (Table \u003cspan refid=\"MOESM18\" class=\"InternalRef\"\u003eS18\u003c/span\u003e). As a result, \u003cem\u003emicrotubule actin crosslinking factor 1\u003c/em\u003e (\u003cem\u003eMACF1\u003c/em\u003e) emerged as the hub of the co-segregation network, \u003cem\u003eryanodine receptor 2\u003c/em\u003e(\u003cem\u003eRYR2\u003c/em\u003e) as the hub of SCZ-specific network and \u003cem\u003efilamin A\u003c/em\u003e (\u003cem\u003eFLNA\u003c/em\u003e) as the hub of OCD-specific network (Fig.\u0026nbsp;5d). The expression data of the hub genes were obtained from Allen Brain atlas that MACF1 plays a substantial role in actin regulation, microtubule arrangement and stabilization, widely distributed in every subcluster of cells (Fig.\u0026nbsp;5e); RYR2 is a component of a calcium channel that is widely expressed in neurons (Fig.\u0026nbsp;5e) and FLNA is another actin-binding protein playing a role in cell junctions during the organ development, specifically expressed in certain types of neuron and cells from neurovascular system (Fig.\u0026nbsp;5e). The core function implied by hub genes was also supported by the network enrichment that the SCZ-specific network was enriched in Uniprot terms \u0026ldquo;Calmodulin-binding\u0026rdquo; (q\u0026thinsp;=\u0026thinsp;0.0064),\u0026rdquo; Calcium transport\u0026rdquo; (q\u0026thinsp;=\u0026thinsp;0.0223) and \u0026ldquo;Ligand-gated ion channel\u0026rdquo; (q\u0026thinsp;=\u0026thinsp;0.0336), while OCD-specific network was enriched in Uniprot terms \u0026ldquo;actin-binding\u0026rdquo; (q\u0026thinsp;=\u0026thinsp;0.0279).\u003c/p\u003e \u003cp\u003eTo sum up, the PPI networks revealed that the shared genetic basis underlying SCZ and OCD may be the fundamental microtube-actin dynamics. Compared to OCD, SCZ showed more prominent dysfunction in calcium-related pathways. Then, we combined fisher test, SKATO test and ACATO test on high-impact SNVs to perform the region-based rare-variant association test, separately (see Methods). None of the genes reached the Bonferroni threshold in any of the comparisons. However, gene \u003cem\u003eMACF1\u003c/em\u003e was marginally significant in all three tests when we comparing the OCD and control groups. (Table\u0026nbsp;5).\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003eRegion-based burden test in non-coding regions revealed the converged network with coding region\u003c/h2\u003e \u003cp\u003eWe replicated the region-based burden test in the non-intronic and non-intergenic variants in non-coding regions. Several genes met the threshold of OR\u0026thinsp;\u0026gt;\u0026thinsp;1 and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 in all the three tests, including CALM2, PLCG2, KCNH5, MMS22L, MCF2L and DNAL1 (Table \u003cspan refid=\"MOESM19\" class=\"InternalRef\"\u003eS19\u003c/span\u003e). Notably, we found that these genes, including CALM2, PLCG2, MMS22L, were involved in the main cluster of STRING network established by high-confidence variants in coding-region (Figure \u003cspan refid=\"MOESM7\" class=\"InternalRef\"\u003eS7\u003c/span\u003e). Interestingly, none of the variants in these genes were detected in coding-regions in our cohort. Therefore, our results may indicate convergence in the same functional network of coding and non-coding variants. None of the genes reached the Bonferroni\u0026rsquo;s threshold (p\u0026thinsp;\u0026lt;\u0026thinsp;7.66e-6) in all the three test and only gene HLA-DRB5 pass the Bonferroni\u0026rsquo;s correction in Fisher\u0026rsquo;s exact test.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec29\" class=\"Section2\"\u003e \u003ch2\u003eTemporal Co-expression Networks Map Developmental Divergence and Genetic Convergence\u003c/h2\u003e \u003cp\u003eTo comprehensively explore the developmental trajectory divergence of different diseases, we separately built unsupervised co-expression networks for SCZ and OCD via WGCNA. For each network, we enrolled both coding and non-coding variants, aiming to explore the convergence and divergence in the development of SCZ and OCD. For coding variants, the high-impact variants in co-segregation (Set-Ccds) and OCD-specific (Set-Ocds) panels, as well as co-segregation and SCZ-specific (Set-Scds) genes were used as input representing coding variants for OCD network (Set-HIcds_OCD) and SCZ network (Set-HIcds_SCZ), separately. For non-coding variants, we excluded the intronic and intergenic variants in different disease group firstl. These remaining variants were filtered by the brain-expressed protein-list in the HPA database, and then non-coding panels including CSncs and OSncs, as well as CSncs and SSncs, separately. These selected variants were pooled to form the variant sets in OCD group and in SCZ group. According to the above results of functional element distributions, we further selected the variants which were predicted to be active transcription (simultaneously annotated as \u0026ldquo;4_Tx\u0026rdquo; by roadmap and as \u0026ldquo;TN1 Transcription\u0026rdquo; by Sei) (set Set-Trans_OSncs and Set-Trans_SSncs) and promoter (simultaneously annotated as \u0026ldquo;1_TssA\u0026rdquo; or \u0026ldquo;11_BivFlnk\u0026rdquo; by roadmap and as \u0026ldquo;P Promoter\u0026rdquo; by Sei) (Set-Pro_OSncs and Set-Pro_SSncs). Finally, we combined the mapped genes from regions of non-coding active transcription and non-coding promoter as input for the OCD (Set-ncsOCD) and SCZ (Set-ncs_SCZncs) group, separately. The aforementioned coding and non-coding gene sets were integrated to construct comprehensive networks for OCD and SCZ. Furthermore, we plotted the average expression curves of the detected modules from the 8th week pcw to 40 years of age, with developmental stages classified according to gestational periods. This allowed us to observe the characteristics of gene expression across different periods. The results demonstrated that there were both similar and distinct expression patterns of gene clusters between groups (Fig.\u0026nbsp;6a). We prioritized the key modules by performing Fisher\u0026rsquo;s exact test to observe the distribution of gene sets (Fig.\u0026nbsp;6b, Table \u003cspan refid=\"MOESM20\" class=\"InternalRef\"\u003eS20\u003c/span\u003e-\u003cspan refid=\"MOESM21\" class=\"InternalRef\"\u003eS21\u003c/span\u003e) and making EWCE analysis to observe differences of cell type enrichment (Fig.\u0026nbsp;6c, Table \u003cspan refid=\"MOESM22\" class=\"InternalRef\"\u003eS22\u003c/span\u003e-\u003cspan refid=\"MOESM23\" class=\"InternalRef\"\u003eS23\u003c/span\u003e). Analysis details were described in Supplementary note 5 and Table \u003cspan refid=\"MOESM20\" class=\"InternalRef\"\u003eS20\u003c/span\u003e-. These results suggested that late-pregnancy modules captured shared OCD\u0026ndash;SCZ coding variants associated with angiogenesis. SCZ-specific gene modules were enriched for calcium-related pathways, with peak expression occurring after late pregnancy and continuing into later developmental stages. Modules associated with synaptic plasticity, which comprised both coding and non-coding variants, showed consistent expression across all postnatal developmental stages. Together, these results suggest that OCD and SCZ are shaped by distinct combinations of genomic elements operating across key developmental windows.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this pilot stage of the CoFFiR project, we performed deep whole-genome sequencing in mixed OCD\u0026ndash;SCZ pedigrees to probe convergent and divergent rare-variant signals under a largely shared familial background. By combining within-family co-segregation with burden-based modeling and developmental network mapping, we provide a complementary view to large case\u0026ndash;control studies for a question they are often underpowered to resolve: why different psychiatric diagnoses emerge within the same pedigrees.\u003c/p\u003e \u003cp\u003eOur study covered both coding and non-coding regions, providing a family-grounded complement to GWAS by interrogating rare coding and regulatory variation with reduced background heterogeneity. After gathering basic functional evidence from burden tests across different genomic regions, we developed a disease-specific genetic risk model. It integrates both coding and non-coding rare variants, finding that combining variants from coding and non-coding regions improved the distinguishing power, highlighting potential implications for future family-based risk stratification, pending replication and prospective validation. The model was supported by pathway analyses across multiple hierarchical levels. Core disease-specific PPI networks revealed that SCZ-specific variants were enriched in pathways related to calmodulin binding, calcium ion transport, and ligand-gated ion channels, while OCD-specific variants were enriched in actin-binding functions. The fact that these signals emerged from different individuals within the same families supports their biological relevance. By introducing BrainSpan developmental data, we mapped the temporal expression patterns of these variants, which served as our attempt to provide a probable biological explanation for the divergence and convergence in the genomic architecture of OCD and SCZ.\u003c/p\u003e \u003cdiv id=\"Sec31\" class=\"Section2\"\u003e \u003ch2\u003eGenomic Architecture Integrating Stratified Coding Variants and Functional Elements in Non-coding Regions\u003c/h2\u003e \u003cp\u003eIn our study, we incorporated non-coding variants to comprehensively identify the risk factors underlying the etiology of OCD and SCZ. Significant enrichment in strong transcription state and notable trends of enrichment in promoter state were observed in non-coding regions of OCD. The predicted strong transcription state refers to chromatin regions with high enrichment of H3K36me3(\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e) and the predicted promoter state indicated enrichment of H3K4me3 (and H3K27me3, which is present in a bivalent state)(\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e). Our findings supported previous publications suggesting the involvement of chromatin regulation in SCZ and OCD pathogenesis, especially OCD(\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e). Motivated by the observation that OCD, SCZ, and control groups differ in their proportions of non-coding elements, we proposed that different combinations of functional elements within shared genomic regions could imply different clinical outcomes. This hypothesis was supported by the results of the Random Forest classifiers. By combining coding and non-coding variants, the prediction model achieved the highest accuracy in almost all comparison groups. In addition to underscoring the significance of rare variants in coding regions again, the results highlighted the importance of non-coding variants in OCD and suggested that OCD may involve a greater proportion of non-coding disturbances. Due to the complexity of the non-coding elements and their interactions, the precise genomic architecture still requires further investigation. Despite the modest sample size in our data, the implication from non-coding regions still provided insights into OCD pathogenesis discovery.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec32\" class=\"Section2\"\u003e \u003ch2\u003eDifferent types of variants coordinated in shared and distinct functional networks of OCD and SCZ\u003c/h2\u003e \u003cp\u003eBoth OCD and SCZ are complex mental disorders with poorly characterized etiology. Theories about both conditions include key disturbances in neurotransmitters and disruptions in neurodevelopment(\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR83\" class=\"CitationRef\"\u003e83\u003c/span\u003e). With accumulating evidence from functional genomic research, SCZ is increasingly considered to have a neurodevelopmental origin. Our data demonstrated distinct characteristics of protein dysfunctions and PPI disruptions between OCD and SCZ in a cross-sectional perspective. MIPPI analysis among high-impact variants suggested significant PPI disturbance of OCD and SCZ, indicating probable protein dysfunction in these disorders. Enrichment analysis supported the neurodevelopmental hypothesis of OCD and SCZ as the high-impact variants in co-segregation panels were centered by neurogenesis. SCZ-specific and OCD-specific proteins showed different trends of enrichment, that SCZ-specific genes tended to be enriched in terms related to calcium ion transport, one of the classical pathways in SCZ pathogenicity(\u003cspan citationid=\"CR84\" class=\"CitationRef\"\u003e84\u003c/span\u003e), while OCD-specific genes tended to be enriched in terms related to cell junction, which was also reported but in scattered studies(\u003cspan additionalcitationids=\"CR86\" citationid=\"CR85\" class=\"CitationRef\"\u003e85\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR87\" class=\"CitationRef\"\u003e87\u003c/span\u003e). The results were further validated by the PPI networks formed by selected high-confidence genes and their hub genes. Our results suggest that a bias toward cell junctions may be a feature of the molecular etiology of OCD.\u003c/p\u003e \u003cp\u003eIn the following rare variants association test combining three different algorithms, MACF1 was the only gene that reached marginal significance in all three tests. It also served as the hub gene in our co-segregated network. Given the modest sample size in our cohort, we think the effect size brought by \u003cem\u003eMACF1\u003c/em\u003e was noteworthy. \u003cem\u003eMACF1\u003c/em\u003e encodes a large and complex protein that is a crucial regulator of the cytoskeleton in neurons and glial cells, and therefore plays prominent roles in multiple cellular functions in brain cells e.g. cell proliferation, migration and neurite development. It has been identified as a candidate gene in psychoses including SCZ(\u003cspan additionalcitationids=\"CR89\" citationid=\"CR88\" class=\"CitationRef\"\u003e88\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e) and bipolar disorder(\u003cspan citationid=\"CR91\" class=\"CitationRef\"\u003e91\u003c/span\u003e). In SCZ, it may contribute to cognitive deficits by disrupting the intracellular transport and cytoskeletal stability at synapse(\u003cspan citationid=\"CR90\" class=\"CitationRef\"\u003e90\u003c/span\u003e). It is worth noticing that it has also been reported to be associated with a family suffering from inherited SCZ and schizoaffective disorder(\u003cspan citationid=\"CR92\" class=\"CitationRef\"\u003e92\u003c/span\u003e). We are the first to report high-impact \u003cem\u003eMACF1\u003c/em\u003e variants to be significantly enriched in OCD. So far, combined evidence suggested that it may be a shared candidate gene in OCD, SCZ and bipolar disorder. Since the detailed mechanism of its pathological function remains largely unknown, further studies are in need to deepen our understanding of it on a mechanistic or disease-specific level. Given the limited power of rare variant association analysis in the family-based design, false negatives may be present in our results. Future studies with larger family sample sizes and enhanced family-based algorithms will be essential.\u003c/p\u003e \u003cp\u003eWe also performed multiple association tests in non-coding regions. Several genes (CALM2, PLCG2, MMS22L, KCNH5, MCF2L, DNAL1) were identified as key results. Among the results, CALM2, PLCG2, MMS22L were found to be integral to the main PPI network constructed by coding variants. CALM2 was mostly studied in cardiac arrhythmias(\u003cspan citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e). A few studies have linked altered CALM2 expression with SCZ(\u003cspan additionalcitationids=\"CR94 CR95\" citationid=\"CR93\" class=\"CitationRef\"\u003e93\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR96\" class=\"CitationRef\"\u003e96\u003c/span\u003e). Additionally, abnormal DNA methylation of CALM2 has been observed in postnatal malnourished mice(\u003cspan citationid=\"CR97\" class=\"CitationRef\"\u003e97\u003c/span\u003e), indicating that adversity may affect CALM2 through epigenomic mechanisms. PLCG2 acts as a downstream molecule in BDNF/TrkB pathway, and was mostly been reported to contribute to Alzheimer\u0026rsquo;s disease by affecting synaptic function(\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e). Its mutations have been detected in psychoses(\u003cspan citationid=\"CR99\" class=\"CitationRef\"\u003e99\u003c/span\u003e), autism(\u003cspan citationid=\"CR98\" class=\"CitationRef\"\u003e98\u003c/span\u003e) and pediatric acute onset neuropsychiatric syndrome(\u003cspan citationid=\"CR100\" class=\"CitationRef\"\u003e100\u003c/span\u003e). Cross-species brain transcriptomic analyses suggested that it exhibited conserved gene expression patterns in affective disorders(\u003cspan citationid=\"CR101\" class=\"CitationRef\"\u003e101\u003c/span\u003e). Given their significant enrichment in the whole disease group, and their central position in the network, our results suggest that CALM2 and PLCG2 are candidate genes for both schizophrenia and OCD. Concerning the complex effects of non-coding variants, more accurate annotations and quantified predictions of non-coding variants, as well as more powerful gene-based algorithms are in need to further explain the effects of these cryptic variants in a broader range of brain disease.\u003c/p\u003e \u003cdiv id=\"Sec33\" class=\"Section3\"\u003e \u003ch2\u003eWGCNA sketched the comprehensive temporal landscape of variants in OCD and SCZ\u003c/h2\u003e \u003c/div\u003e \u003cdiv id=\"Sec34\" class=\"Section3\"\u003e \u003ch2\u003eEarly-and-mid-pregnancy stage captured epigenomic changes and divergence between disease\u003c/h2\u003e \u003cp\u003eIn order to provide biological explanation for the disease-specific score, we incorporated both coding and non-coding variants into WGCNA to gain insights into the dynamic changes of high-impact variants in OCD and SCZ. In early-pregnancy i.e., embryonic stage, we detected overlapping and similarly enriched terms pertaining to cellular reproduction, epigenomic modification and so on between OCD and SCZ. These results may be consistent with previous publications implying chromatin modification (\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR82\" class=\"CitationRef\"\u003e82\u003c/span\u003e) and isoform-level dysregulation(\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR102\" class=\"CitationRef\"\u003e102\u003c/span\u003e) in OCD and SCZ.\u003c/p\u003e \u003cp\u003eAccording to the cellular and molecular landscape of developing human brain, mid-pregnancy is the stage characterized by the onset of neurogenesis and neuronal migration(\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e). Results in this stage suggests potential differences in noncoding regions between OCD and SCZ. During the mid-pregnancy, we detected enrichment trends of non-coding variants, especially in SCZ. This result was in accordance with previous studies that the fetal brain development was shown to harbor greater enrichment for SCZ GWAS signal(\u003cspan citationid=\"CR104\" class=\"CitationRef\"\u003e104\u003c/span\u003e). Significant enrichment was detected in terms related to epigenetic and post-transcription regulation, as well as terms related to cytoskeleton and cell structure in OCD. Broader disturbed functions were detected in SCZ. Noticeably, while a broad class of inhibitory neurons in SCZ exhibited trends of enrichment in this stage compared to OCD, Lamp5 and Pvalb emerged as the most prominent. Pvalb, the largest class of inhibitory neurons, provide feedforward and feedback synaptic inhibition to a variety of neurons(\u003cspan citationid=\"CR105\" class=\"CitationRef\"\u003e105\u003c/span\u003e), which allows the elaboration on a broader, more coordinated, and more delicate firing patterns(\u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e106\u003c/span\u003e). A strong body of evidence has linked Pvalb to SCZ (\u003cspan citationid=\"CR106\" class=\"CitationRef\"\u003e106\u003c/span\u003e). Meta-analysis of post-mortem studies suggested a significant reduction of Pvalb cell density in the prefrontal cortical regions of SCZ(\u003cspan citationid=\"CR107\" class=\"CitationRef\"\u003e107\u003c/span\u003e). Animal experiments demonstrated that disruption of Pvalb development, particularly before puberty, may exacerbate the neuropathological consequence of adversity e.g. social isolation, in subsequent developmental stage(\u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e108\u003c/span\u003e), and thus leading to various pathophysiological manifestations in SCZ patients(\u003cspan citationid=\"CR108\" class=\"CitationRef\"\u003e108\u003c/span\u003e, \u003cspan citationid=\"CR109\" class=\"CitationRef\"\u003e109\u003c/span\u003e). The trajectory in our data implied the idea that this disruption may root in the genomic alterations, with primary effects occurring as early as in the mid-pregnancy, leading to delayed but prolonged synaptic effects on these late-maturing GABAergic neurons(\u003cspan citationid=\"CR110\" class=\"CitationRef\"\u003e110\u003c/span\u003e). In contrast to Pvalb, Lamp5 neurons are less well-studied. They are thought to play a role in GABAergic neurotransmission within specific neuronal subpopulations, affecting short-term synaptic plasticity(\u003cspan citationid=\"CR111\" class=\"CitationRef\"\u003e111\u003c/span\u003e). Recent evidence from snRNA-seq(\u003cspan citationid=\"CR112\" class=\"CitationRef\"\u003e112\u003c/span\u003e) and ALLEN BRAIN MAP(\u003cspan citationid=\"CR113\" class=\"CitationRef\"\u003e113\u003c/span\u003e) suggested Lamp5 and Pvalb neurons among the most altered neurons types in SCZ post-mortem samples, although information about Lamp5 remains limited.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e\n\u003ch3\u003eAngiogenesis in late-pregnancy was the core affected functional pathway shared by diseases\u003c/h3\u003e\n\u003cp\u003eThe late-pregnancy and early postnatal stage are considered as the onset of astrogliogenesis, oligodendrogenesis and synaptogenesis(\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e). Our results indicated this stage as the core period, with co-segregation coding variants significantly enriched in both the OCD and SCZ groups. The enrichment terms strongly supported the alteration of angiogenesis and blood-brain-barrier (BBB) integrity as a shared genetic etiology of OCD and SCZ. Many studies supported the brain microvascular endothelial cell and BBB dysfunction in SCZ. However, it remained unsolved whether or not the deficits were primary or compensatory(\u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e114\u003c/span\u003e, \u003cspan citationid=\"CR115\" class=\"CitationRef\"\u003e115\u003c/span\u003e). Alterations in adhesion and angiogenesis molecules in OCD have been reported in a few studies, but the evidence remains limited(\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e, \u003cspan additionalcitationids=\"CR117 CR118\" citationid=\"CR116\" class=\"CitationRef\"\u003e116\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR119\" class=\"CitationRef\"\u003e119\u003c/span\u003e). The brain angiogenesis and BBB formation are thought to start at the embryonic stage. BBB is formed on the basis of neurovascular units (NVU) (\u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e120\u003c/span\u003e). We detected the enrichment of several components in an NVU in this stage. The neurovascular system plays a critical role selectively and dynamically regulating the molecular transport between the bloodstream and brain parenchyma (\u003cspan citationid=\"CR115\" class=\"CitationRef\"\u003e115\u003c/span\u003e). This regulation is essential to maintain the integrity of the brain parenchyma and preserve subtle neurochemical homeostasis from external environment(\u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e121\u003c/span\u003e). Deficits in NVU may lead to disturbed neurotransmitter transport, disrupted neuronal signaling, or impaired neurotoxin elimination e.g. cytokines, which may ultimately contribute to the clinical manifestations of SCZ and OCD(\u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e121\u003c/span\u003e). Recent literature reported the abnormalities in gene expression, microstructure and paracellular permeability in endothelial cells of SCZ patient-derived 3D cerebral organoids(\u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e120\u003c/span\u003e). These findings support the early origins of endothelial alteration and BBB dysfunction in SCZ(\u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e120\u003c/span\u003e). It also provided explanation for the disturbance of immune-inflammatory system observed in SCZ(\u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e120\u003c/span\u003e). Impaired BBB in OCD was most noticed in the context of pediatric acute-onset neuropsychiatric syndrome (PANS), and pediatric autoimmune neuropsychiatric disorders associated with streptococcal infection (PANDAS). PANS/PANDAS sufferers usually present with OCS, and the prevailing etiology model includes the breaching of BBB by the infection-triggered antibodies(\u003cspan citationid=\"CR122\" class=\"CitationRef\"\u003e122\u003c/span\u003e, \u003cspan citationid=\"CR123\" class=\"CitationRef\"\u003e123\u003c/span\u003e). It was reported that the first-degree relatives of PANS/PANDAS sufferers have up to 10-fold risks of developing OCD, which suggested significant genetic overlap(\u003cspan citationid=\"CR122\" class=\"CitationRef\"\u003e122\u003c/span\u003e). Autoantibodies directed against basal ganglia was also detected in the cerebrospinal fluid (CSF) of drug-na\u0026iuml;ve OCD patients (\u003cspan citationid=\"CR124\" class=\"CitationRef\"\u003e124\u003c/span\u003e). Multifaceted evidence in our study indicated the NVU-based BBB formation as primary pathologies shared by SCZ and OCD. In fact, there existed a further aspect underscoring the close structural and developmental link between blood vessels and nerves due to the tight coupling between neuronal activity and blood flow(\u003cspan citationid=\"CR125\" class=\"CitationRef\"\u003e125\u003c/span\u003e). Abnormalities in vascular network and blood dynamics may exert more prolonged impact on synaptic plasticity and neuronal metabolism in SCZ and OCD pathology(\u003cspan citationid=\"CR114\" class=\"CitationRef\"\u003e114\u003c/span\u003e, \u003cspan citationid=\"CR125\" class=\"CitationRef\"\u003e125\u003c/span\u003e). Further elucidation of neurovascular system may offer new insights into the abnormal brain development and functioning underlying OCD and SCZ.\u003c/p\u003e \u003cp\u003eIn schizophrenia (SCZ), various human studies have indicated neurovascular system abnormalities. Postmortem analyses have shown structural changes in brain capillaries, such as reduced density, smaller diameters, endothelial degeneration, and extracellular matrix buildup, particularly in the prefrontal and visual cortices (\u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e126\u003c/span\u003e). Imaging studies, including 7T-MRI, have revealed altered volumes in small cerebral arteries, potentially explaining SCZ-related gray matter loss (\u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e126\u003c/span\u003e). Diffusion-prepared arterial spin labeling (DP-ASL) MRI also showed reduced neurovascular water exchange in SCZ-spectrum patients (\u003cspan citationid=\"CR127\" class=\"CitationRef\"\u003e127\u003c/span\u003e). Cerebrospinal fluid studies found elevated blood-brain barrier (BBB) permeability markers like Qalb and S100β in SCZ, suggesting compromised BBB integrity(\u003cspan citationid=\"CR126\" class=\"CitationRef\"\u003e126\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eIn contrast, only few studies have identified altered levels of related molecules, such as angiopoietin (ANG), cadherin-5 (CDH5), and ICAM-3, in peripheral blood of OCD patients (\u003cspan citationid=\"CR86\" class=\"CitationRef\"\u003e86\u003c/span\u003e, \u003cspan citationid=\"CR119\" class=\"CitationRef\"\u003e119\u003c/span\u003e), hinting at possible neurovascular dysregulation. Our study found significantly more OCD-associated genes than SCZ-associated genes in the relevant module, highlighting this pathway's potential importance in OCD pathophysiology.\u003c/p\u003e\n\u003ch3\u003eSCZ-specific module highlighted the critical role of astrocytes and calcium signaling\u003c/h3\u003e\n\u003cp\u003eCompared with OCD, there also existed another black module enriched with SCZ-specific coding variants. Cell types including astrocytes and pericytes, as well as terms related to ion calcium signaling pathways, RHOa GTPase cycle, clathrin-mediated endocytosis and transport of small molecules were enriched. It reflected the affected pathways detected in the SCZ-specific PPI network. These results marked it as another astrocyte-centric module, with sustained high expression during the whole postanal stage. Emerging genetic, pathological and serological evidence increasingly highlighted the critical role of astrocytes in SCZ (\u003cspan citationid=\"CR128\" class=\"CitationRef\"\u003e128\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAstrocytes are not only structural components of BBB, but also play key roles in glutamate metabolism, maintenance of synaptic networks, and regulation of synaptic plasticity (\u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e121\u003c/span\u003e, \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e129\u003c/span\u003e). Growing genetic, pathological, and serological evidence highlights their critical involvement in SCZ (\u003cspan citationid=\"CR128\" class=\"CitationRef\"\u003e128\u003c/span\u003e). Postmortem studies from SCZ patients have revealed altered astrocyte density and morphology, along with dysregulated expression of markers like aquaporin-4 (AQP-4). Animal models have shown that similar changes in the prefrontal cortex can lead to SCZ-related cognitive impairments (\u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e130\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTranscriptomic analyses of developmental data show that proliferation-regulating genes are downregulated during astrocyte maturation, while genes involved in neurotransmitter transport (glutamate, GABA), connexins (Cx30, Cx43), and potassium channels (Kir4.1) are upregulated (\u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e130\u003c/span\u003e). These upregulated genes support synaptic homeostasis and astrocyte-synapse interactions, suggesting that astrocyte maturation parallels increased synaptic activity. In contrast, SCZ astrocytes may exhibit immature-like phenotypes with reduced expression of Kir channels and Cx30, which align with SCZ features in transgenic mouse models (\u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e130\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTherefore, the astrocyte and small molecule transport, especially calcium-related, abnormalities observed in this module may reflect SCZ-specific functional changes, distinct from OCD, likely involving disrupted astrocyte development and related synaptic functions. Since astrocytes mediate interactions between the BBB and neurons, disturbances in the blood\u0026ndash;astrocyte\u0026ndash;neuron axis may also occur. Further investigation into astrocyte development and their role in neurotransmission could help explain core SCZ symptoms and guide new treatment strategies (\u003cspan citationid=\"CR130\" class=\"CitationRef\"\u003e130\u003c/span\u003e, \u003cspan citationid=\"CR131\" class=\"CitationRef\"\u003e131\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAstrocytes serve as the interface between BBB and the neurons, and mediate interactions between vascular dynamics and neurotransmitter system. They are not only a structural component of BBB, but also a key partitioner in a variety of neurobiological functions including glutamate metabolism, synaptic network structuring and maintenance, as well as synaptic plasticity regulation(\u003cspan citationid=\"CR121\" class=\"CitationRef\"\u003e121\u003c/span\u003e, \u003cspan citationid=\"CR129\" class=\"CitationRef\"\u003e129\u003c/span\u003e). Our results implied a more prominent alteration in molecular transportation and potential disruption in blood-glia-neuro crosstalk via astrocytes and pericytes as featured functional impairments in SCZ compared with OCD.\u003c/p\u003e \u003cdiv id=\"Sec37\" class=\"Section2\"\u003e \u003ch2\u003eShared and distinct neuronal types were found disturbed in modules about synaptic plasticity\u003c/h2\u003e \u003cp\u003eIn addition to the variant-type-enriched modules, we also identified the brown modules in OCD and SCZ to contain shared and distinct components. The brown modules in both groups were characterized by the increasing expression starting from synaptogenesis time in late-pregnancy(\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e)and maintained a consistently high expression since late infancy. Functional enrichment revealed significantly enriched terms related to synaptic plasticity. These findings are highly consistent with previous GWAS and are further supported by prior evidence implicating specific allele sets\u0026mdash;particularly genes expressed in neurons within defined brain regions and gene sets intolerant to PTVs or involved in synaptic function\u0026mdash;in shared genetic and biological mechanisms underlying both common and rare variants in SCZ (\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e). Additionally, enrichment in terms related to metabolism and membrane localization points to processes involved in maintaining neuronal system homeostasis. While not central to earlier GWAS results, these processes align with emerging insights from metabolomics studies and transcriptome-wide association studies (TWAS) in SCZ(\u003cspan additionalcitationids=\"CR133\" citationid=\"CR132\" class=\"CitationRef\"\u003e132\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR134\" class=\"CitationRef\"\u003e134\u003c/span\u003e), suggesting they may serve as phenotypic modifiers or contribute to disease expression. In the broader context of cross-disorder research, our findings provide new evidence supporting shared physiological pathways across neuropsychiatric conditions, offering further insight into the overlapping molecular underpinnings of these complex disorders.\u003c/p\u003e \u003cp\u003eWhereas genes in both OCD and SCZ groups were enriched in types of neurons, of particular interest, some genes in OCD group were specifically enriched in Pvalb inhibitory neurons, while some genes in SCZ group were specifically enriched in oligodendrocytes. Pvalb was scarcely studied in OCD. As we have detected enrichment of Pvalb in SCZ during mid-pregnancy, the enrichment of Pvalb in OCD suggested that the potential Pvalb deficits in OCD may be more closely related to the process of neurotransmitter and synaptic function. Pvalb was reported in OCD in striatum, where continuous optogenetic stimulation of Pvalb in the striatal areas connecting to lateral orbital-frontal-cortex can reduce compulsive-like grooming behaviors in Sapap3-KO mice(\u003cspan citationid=\"CR135\" class=\"CitationRef\"\u003e135\u003c/span\u003e). Knocking down Pvalb in pre-frontal cortex was found to reduce cognitive flexibility(\u003cspan citationid=\"CR136\" class=\"CitationRef\"\u003e136\u003c/span\u003e, \u003cspan citationid=\"CR137\" class=\"CitationRef\"\u003e137\u003c/span\u003e), one of the core cognitive impairments in both SCZ(\u003cspan citationid=\"CR136\" class=\"CitationRef\"\u003e136\u003c/span\u003e) and OCD(\u003cspan citationid=\"CR138\" class=\"CitationRef\"\u003e138\u003c/span\u003e, \u003cspan citationid=\"CR139\" class=\"CitationRef\"\u003e139\u003c/span\u003e). The neuropathological effects of Pvalb alteration may vary across regions(\u003cspan citationid=\"CR140\" class=\"CitationRef\"\u003e140\u003c/span\u003e) and developmental stages(\u003cspan citationid=\"CR141\" class=\"CitationRef\"\u003e141\u003c/span\u003e). As evidence was limited, more research investigating the role of Pvalb in the whole classical cortico-striato-thalamo-cortical circuits in different developmental stages was needed. Besides Pvalb in OCD, enrichment of oligodendrocytes was found in SCZ in our study. Recent research has increasingly supported the role of oligodendrocytes in SCZ, with accumulating evidence from imaging, histopathological, and molecular studies(\u003cspan citationid=\"CR142\" class=\"CitationRef\"\u003e142\u003c/span\u003e). Abundant evidence came from functional neuroimaging studies suggested SCZ as a disorder of dysconnectivity in brain(\u003cspan citationid=\"CR143\" class=\"CitationRef\"\u003e143\u003c/span\u003e). Oligodendrocytes are cells that form myelin sheaths around multiple axons, facilitating rapid conduction of electrical signals and maintaining axonal integrity at the cellular level. Oligodendrogenesis was thought to begin at late-pregnancy and followed by myelination throughout the whole childhood(\u003cspan citationid=\"CR103\" class=\"CitationRef\"\u003e103\u003c/span\u003e). However, few studies have investigated the specific impact of oligodendrocytes in SCZ-related induced pluripotent stem cell (iPSC) models (\u003cspan citationid=\"CR142\" class=\"CitationRef\"\u003e142\u003c/span\u003e). The exact mechanism by which oligodendrocytes affect SCZ remains uncertain. Possible impacts include altered synaptic plasticity, axonal degeneration, conduction velocity, neuronal circuitry or neuronal signaling(\u003cspan citationid=\"CR144\" class=\"CitationRef\"\u003e144\u003c/span\u003e). Our results revealed the enrichment of oligodendrocytes as featured impaired cell type of SCZ, highlighting the need for further investigation into the complex intercellular interactions of oligodendrocytes with neurons.\u003c/p\u003e \u003cdiv id=\"Sec38\" class=\"Section3\"\u003e \u003ch2\u003eEnlightenment brought by integrative genomic structure and developmental trajectory\u003c/h2\u003e \u003cp\u003eThe phenotypic divergence observed within complex families, ranging from unaffected individuals to those diagnosed with OCD or SCZ, may be explained by the specific configuration of rare inherited variants, their regulatory loci, and the timing of their activity during neurodevelopment. Our study indicates that while some family members inherit high-impact coding or noncoding variants active in disorder-specific pathways, others either lack such variants or possess variants that fall outside critical spatiotemporal windows. This may account for the absence of clinical symptoms in certain individuals.\u003c/p\u003e \u003cp\u003eThe \"two-hit\" or even \u0026ldquo;multi-hit\u0026rdquo; hypothesis in SCZ posits that multiple genetic and environmental factors must converge during key periods of neurodevelopment to precipitate the disorder in genetically susceptible individuals(\u003cspan citationid=\"CR145\" class=\"CitationRef\"\u003e145\u003c/span\u003e, \u003cspan citationid=\"CR146\" class=\"CitationRef\"\u003e146\u003c/span\u003e). A similarly complex architecture may underlie OCD etiology. While blood\u0026ndash;brain barrier (BBB) dysfunction and neurovascular developmental alteration has been linked to several pathogenic mechanisms, including oxidative stress, inflammation and synaptic development(\u003cspan citationid=\"CR120\" class=\"CitationRef\"\u003e120\u003c/span\u003e), our findings suggest that they may represent a core primary defect tied to genetic susceptibility in both SCZ and OCD, particularly in OCD, where environmental modulation appears less influential. Notably, deficits related to calcium or ion channels were more prominently associated with SCZ, indicating a higher pathogenic burden in this disorder. Furthermore,, the strong enrichment of noncoding variants in early-to-mid pregnancy modules, alongside moderate involvement in postnatal modules, supports the hypothesis that cell proliferation and synaptic elimination represent key vulnerability windows in SCZ (\u003cspan citationid=\"CR146\" class=\"CitationRef\"\u003e146\u003c/span\u003e, \u003cspan citationid=\"CR147\" class=\"CitationRef\"\u003e147\u003c/span\u003e). These critical developmental stages may therefore constitute promising targets for preventive strategies and epigenomic interventions.\u003c/p\u003e \u003cp\u003eEnvironmental risk factors, such as prenatal vitamin D deficiency, viral infections, poor nutrition, postnatal social adversity, and childhood trauma, may exert cumulative and interactive effects, influencing disease onset by acting during different critical windows of neurodevelopment(\u003cspan citationid=\"CR148\" class=\"CitationRef\"\u003e148\u003c/span\u003e). While prior studies highlight the transdiagnostic impact of such exposures across mental disorders (\u003cspan citationid=\"CR148\" class=\"CitationRef\"\u003e148\u003c/span\u003e), our prediction model points out that the trajectory toward a specific disorder is strongly rooted in genomic architecture. Given the reversibility of epigenetic modifications, future research should focus on clarifying how these risk factors interact with specific genes, pathways, or cell types at defined developmental stages. Such insights may ultimately facilitate targeted interventions aimed at correcting noncoding regulatory disruptions or epigenetic defects(\u003cspan citationid=\"CR148\" class=\"CitationRef\"\u003e148\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTogether, our findings offer a mechanistic framework for understanding phenotypic divergence within genetically similar individuals. Despite a shared genetic background, the presence of distinct rare variants, whether in coding regions or regulatory elements, appears to direct individuals along divergent neurodevelopmental trajectories, resulting in variable clinical outcomes.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec39\" class=\"Section2\"\u003e \u003ch2\u003eToward Disorder-specific Stratification within High-burden Families\u003c/h2\u003e \u003cp\u003eOur results suggest that mixed-pedigree designs can move cross-disorder genetics beyond population-level overlap toward within-family divergence, where individuals share substantial background yet exhibit distinct diagnoses. In this setting, combinations of rare coding and regulatory variants may bias spatiotemporal neurodevelopmental programs toward OCD, SCZ, or resilience. While the current cohort size limits definitive modeling of intra-family outcomes, the framework illustrates how family-grounded burden profiles and developmental context can be integrated to generate interpretable, disorder-weighted signatures. Future studies in larger mixed-pedigree cohorts and complementary functional systems will be required to validate specific mechanisms and assess clinical generalizability.\u003c/p\u003e \u003cdiv id=\"Sec40\" class=\"Section3\"\u003e \u003ch2\u003eLimitations\u003c/h2\u003e \u003cp\u003eSeveral limitations should be noted. First, the strict ascertainment criteria required for mixed OCD\u0026ndash;SCZ pedigrees constrained sample size, which reduces power for gene-level inference and limits the stability of disorder-specific estimates. Second, regulatory annotations for non-coding variants remain imperfect, and the interpretability of non-coding burden is therefore conditional on current epigenomic resources and priors. Third, our conclusions are based on genomic and in silico network/developmental analyses; experimental validation in cellular or animal systems will be necessary to test causal mechanisms and pinpoint functional consequences of prioritized variants and pathways.\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn stage I of the CoFFiR project, we used deep whole-genome sequencing of mixed OCD\u0026ndash;SCZ pedigrees to dissect convergent and divergent rare-variant architectures under a largely shared familial background. Integrating coding and regulatory variation, we derived family-aware burden signatures that separated affected individuals from controls and provided interpretable disorder-weighted patterns consistent with partial genetic sharing between OCD and SCZ. Developmental network mapping further nominated temporally stratified modules that may underlie shared susceptibility as well as disorder-biased programs. Together, these findings support mixed-pedigree WGS as a complementary strategy for cross-disorder psychiatric genetics and provide a framework for mechanistic prioritization and future replication.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe thanked for the substantial support from all the patients and healthy volunteers. We thanked every member in our group exploring the mechanism of OCD and better strategy of OCD treatment.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by Ethics Committee of Shanghai Mental Health Center, School of Medicine, Shanghai Jiao Tong University. Written informed consent was obtained from each participant.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData available statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMD is responsible for original draft, visualization and data curation, YW is responsible for resources collecting and project administration, WW is responsible for methodology and project administration, YB and ZL are responsible for methodology, QF, LW, SS and ZW are responsible or resources collecting, XL and GNL are responsible for project administration, review, editing, and funding acquisition, ZX is responsible for conceptualization, supervision and funding acquisition.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by National Natural Science Foundation of China (81971261, 82071518, 32200924, 82571771), the Brain Science and Brain-like Intelligence Technology-National Science and Technology Major Project (No. 2022ZD 0209100), Shanghai Science and Technology Committee (22YF1439000), Natural Science Foundation of Shanghai (no: 25ZR1401167), Hospital Project of Shanghai Mental Health Center (2019-YJ15) in the sample collection and genotyping.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eP. H. Lee, Y. A. Feng, J. W. Smoller. Pleiotropy and Cross-Disorder Genetics Among Psychiatric Disorders. Biol Psychiatry. 2021;89(1):20\u0026ndash;31.\u003c/li\u003e\n \u003cli\u003ep. American Psychiatric Association, D. S. M. T. F. a. American Psychiatric Association. Diagnostic and statistical manual of mental disorders : DSM-5\u0026trade;. 5th edition. ed. Washington, DC ;: American Psychiatric Publishing, a division of American Psychiatric Association; 2013.\u003c/li\u003e\n \u003cli\u003eWorld Health Organization. ICD-11 revision [cited 2026 24th January]. Available from: https://icd.who.int/en.\u003c/li\u003e\n \u003cli\u003eC. Wander. Schizophrenia: opportunities to improve outcomes and reduce economic burden through managed care. Am J Manag Care. 2020;26(3 Suppl):S62\u0026ndash;s8.\u003c/li\u003e\n \u003cli\u003eD. J. Stein, D. L. C. Costa, C. Lochner, E. C. Miguel, Y. C. J. Reddy, R. G. Shavitt, et al. Obsessive-compulsive disorder. Nat Rev Dis Primers. 2019;5(1):52.\u003c/li\u003e\n \u003cli\u003eP. H. Lee, V. Anttila, H. Won, Y.-C. A. Feng, J. Rosenthal, Z. Zhu, et al. Genomic Relationships, Novel Loci, and Pleiotropic Mechanisms across Eight Psychiatric Disorders. Cell. 2019;179(7):1469\u0026ndash;82.e11.\u003c/li\u003e\n \u003cli\u003eA. M. Achim, M. Maziade, E. Raymond, D. Olivier, C. Merette, M. A. Roy. How prevalent are anxiety disorders in schizophrenia? A meta-analysis and critical review on a significant association. Schizophr Bull. 2011;37(4):811\u0026ndash;21.\u003c/li\u003e\n \u003cli\u003eM. Swets, J. Dekker, K. van Emmerik-van Oortmerssen, G. E. Smid, F. Smit, L. de Haan, et al. The obsessive compulsive spectrum in schizophrenia, a meta-analysis and meta-regression exploring prevalence rates. Schizophr Res. 2014;152(2-3):458\u0026ndash;68.\u003c/li\u003e\n \u003cli\u003eM. J. Owen, A. Sawa, P. B. Mortensen. Schizophrenia. Lancet. 2016;388(10039):86\u0026ndash;97.\u003c/li\u003e\n \u003cli\u003eJ. W. Smoller, O. A. Andreassen, H. J. Edenberg, S. V. Faraone, S. J. Glatt, K. S. Kendler. Psychiatric genetics and the structure of psychopathology. Mol Psychiatry. 2019;24(3):409\u0026ndash;20.\u003c/li\u003e\n \u003cli\u003eS. M. Meier, L. Petersen, M. G. Pedersen, M. C. Arendt, P. R. Nielsen, M. Mattheisen, et al. Obsessive-compulsive disorder as a risk factor for schizophrenia: a nationwide study. Jama Psychiatry. 2014;71(11):1215\u0026ndash;21.\u003c/li\u003e\n \u003cli\u003eM. H. Chen, S. J. Tsai, C. S. Liang, C. M. Cheng, T. P. Su, T. J. Chen, et al. Diagnostic progression to schizophrenia in 35,255 patients with obsessive-compulsive disorder: a longitudinal follow-up study. Eur Arch Psychiatry Clin Neurosci. 2023;273(3):541\u0026ndash;51.\u003c/li\u003e\n \u003cli\u003eM. Cederlof, P. Lichtenstein, H. Larsson, M. Boman, C. Ruck, M. Landen, et al. Obsessive-Compulsive Disorder, Psychosis, and Bipolarity: A Longitudinal Cohort and Multigenerational Family Study. Schizophr Bull. 2015;41(5):1076\u0026ndash;83.\u003c/li\u003e\n \u003cli\u003eC. Cappi, M. E. Oliphant, Z. Peter, G. Zai, M. Conceicao do Rosario, C. A. W. Sullivan, et al. De Novo Damaging DNA Coding Mutations Are Associated With Obsessive-Compulsive Disorder and Overlap With Tourette\u0026apos;s Disorder and Autism. Biol Psychiatry. 2020;87(12):1035\u0026ndash;44.\u003c/li\u003e\n \u003cli\u003eG. N. Lin, W. Song, W. Wang, P. Wang, H. Yu, W. Cai, et al. De novo mutations identified by whole-genome sequencing implicate chromatin modifications in obsessive-compulsive disorder. Sci Adv. 2022;8(2):eabi6180.\u003c/li\u003e\n \u003cli\u003eM. Fromer, A. J. Pocklington, D. H. Kavanagh, H. J. Williams, S. Dwyer, P. Gormley, et al. De novo mutations in schizophrenia implicate synaptic networks. Nature. 2014;506(7487):179\u0026ndash;84.\u003c/li\u003e\n \u003cli\u003eS. E. McCarthy, J. Gillis, M. Kramer, J. Lihm, S. Yoon, Y. Berstein, et al. De novo mutations in schizophrenia implicate chromatin remodeling and support a genetic overlap with autism and intellectual disability. Mol Psychiatry. 2014;19(6):652\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eA. Takata, B. Xu, I. Ionita-Laza, J. L. Roos, J. A. Gogos, M. Karayiorgou. Loss-of-function variants in schizophrenia risk and SETD1A as a candidate susceptibility gene. Neuron. 2014;82(4):773\u0026ndash;80.\u003c/li\u003e\n \u003cli\u003eM. Halvorsen, R. Huh, N. Oskolkov, J. Wen, S. Netotea, P. Giusti-Rodriguez, et al. Increased burden of ultra-rare structural variants localizing to boundaries of topologically associated domains in schizophrenia. Nat Commun. 2020;11(1):1842.\u003c/li\u003e\n \u003cli\u003eE. Rees, H. D. J. Creeth, H. G. Hwu, W. J. Chen, M. Tsuang, S. J. Glatt, et al. Schizophrenia, autism spectrum disorders and developmental disorders share specific disruptive coding mutations. Nat Commun. 2021;12(1):5353.\u003c/li\u003e\n \u003cli\u003eD. Liu, D. Meyer, B. Fennessy, C. Feng, E. Cheng, J. S. Johnson, et al. Schizophrenia risk conferred by rare protein-truncating variants is conserved across diverse human populations. Nat Genet. 2023;55(3):369\u0026ndash;76.\u003c/li\u003e\n \u003cli\u003eW. A. Akingbuwa, A. R. Hammerschlag, M. Bartels, M. G. Nivard, C. M. Middeldorp. Ultra-rare and common genetic variant analysis converge to implicate negative selection and neuronal processes in the aetiology of schizophrenia. Mol Psychiatry. 2022;27(9):3699\u0026ndash;707.\u003c/li\u003e\n \u003cli\u003eP. F. Sullivan, S. Yao, J. Hjerling-Leffler. Schizophrenia genomics: genetic complexity and functional insights. Arch Gen Psychiatry. 2024.\u003c/li\u003e\n \u003cli\u003eJ. Ott, Y. Kamatani, M. Lathrop. Family-based designs for genome-wide association studies. Nat Rev Genet. 2011;12(7):465\u0026ndash;74.\u003c/li\u003e\n \u003cli\u003eW. K. Goodman, L. H. Price, S. A. Rasmussen, C. Mazure, R. L. Fleischmann, C. L. Hill, et al. The Yale-Brown Obsessive Compulsive Scale: I. Development, Use, and Reliability. Archives of General Psychiatry. 1989;46(11):1006\u0026ndash;11.\u003c/li\u003e\n \u003cli\u003eS. R. Kay, A. Fiszbein, L. A. Opler. The Positive and Negative Syndrome Scale (PANSS) for Schizophrenia. Schizophr Bull. 1987;13(2):261\u0026ndash;76.\u003c/li\u003e\n \u003cli\u003eM. Hamilton. The assessment of anxiety states by rating. Br J Med Psychol. 1959;32(1):50\u0026ndash;5.\u003c/li\u003e\n \u003cli\u003eM. Hamilton. Development of a rating scale for primary depressive illness. Br J Soc Clin Psychol. 1967;6(4):278\u0026ndash;96.\u003c/li\u003e\n \u003cli\u003eY. Chen, Y. Chen, C. Shi, Z. Huang, Y. Zhang, S. Li, et al. SOAPnuke: a MapReduce acceleration-supported software for integrated quality control and preprocessing of high-throughput sequencing data. Gigascience. 2018;7(1):1\u0026ndash;6.\u003c/li\u003e\n \u003cli\u003eH. Li, R. Durbin. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics. 2009;25(14):1754\u0026ndash;60.\u003c/li\u003e\n \u003cli\u003eP. Danecek, J. K. Bonfield, J. Liddle, J. Marshall, V. Ohan, M. O. Pollard, et al. Twelve years of SAMtools and BCFtools. Gigascience. 2021;10(2):giab008.\u003c/li\u003e\n \u003cli\u003eK. Wang, M. Li, H. Hakonarson. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010;38(16):e164\u0026ndash;e.\u003c/li\u003e\n \u003cli\u003eM. Lek, K. J. Karczewski, E. V. Minikel, K. E. Samocha, E. Banks, T. Fennell, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536(7616):285\u0026ndash;91.\u003c/li\u003e\n \u003cli\u003eM. Schubach, T. Maass, L. Nazaretyan, S. R\u0026ouml;ner, M. Kircher. CADD v1.7: using protein language models, regulatory CNNs and other nucleotide-level scores to improve genome-wide variant predictions. Nucleic Acids Res. 2024;52(D1):D1143\u0026ndash;D54.\u003c/li\u003e\n \u003cli\u003eK. E. Samocha, J. A. Kosmicki, K. J. Karczewski, A. H. O\u0026rsquo;Donnell-Luria, E. Pierce-Hoffman, D. G. MacArthur, et al. Regional missense constraint improves variant deleteriousness prediction. bioRxiv. 2017:148353.\u003c/li\u003e\n \u003cli\u003eR. Steinhaus, S. Proft, M. Schuelke, D. N. Cooper, Jana M. Schwarz, D. Seelow. MutationTaster2021. Nucleic Acids Res. 2021;49(W1):W446\u0026ndash;W51.\u003c/li\u003e\n \u003cli\u003eS. Chun, J. C. Fay. Identification of deleterious mutations within three human genomes. Genome Res. 2009;19(9):1553\u0026ndash;61.\u003c/li\u003e\n \u003cli\u003eA. Siepel, G. Bejerano, J. S. Pedersen, A. S. Hinrichs, M. Hou, K. Rosenbloom, et al. Evolutionarily conserved elements in vertebrate, insect, worm, and yeast genomes. Genome Res. 2005;15(8):1034\u0026ndash;50.\u003c/li\u003e\n \u003cli\u003eX. Liu, X. Jian, E. Boerwinkle. dbNSFP: A lightweight database of human nonsynonymous SNPs and their functional predictions. Hum Mutat. 2011;32(8):894\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eX. Liu, C. Li, C. Mou, Y. Dong, Y. Tu. dbNSFP v4: a comprehensive database of transcript-specific functional predictions and annotations for human nonsynonymous and splice-site SNVs. Genome Med. 2020;12(1):103.\u003c/li\u003e\n \u003cli\u003eJ. T. Robinson, H. Thorvaldsd\u0026oacute;ttir, W. Winckler, M. Guttman, E. S. Lander, G. Getz, et al. Integrative genomics viewer. Nat Biotechnol. 2011;29(1):24\u0026ndash;6.\u003c/li\u003e\n \u003cli\u003eM. J. Gandal, P. Zhang, E. Hadjimichael, R. L. Walker, C. Chen, S. Liu, et al. Transcriptome-wide isoform-level dysregulation in ASD, schizophrenia, and bipolar disorder. Science. 2018;362(6420).\u003c/li\u003e\n \u003cli\u003eC. Roadmap Epigenomics, A. Kundaje, W. Meuleman, J. Ernst, M. Bilenky, A. Yen, et al. Integrative analysis of 111 reference human epigenomes. Nature. 2015;518(7539):317\u0026ndash;30.\u003c/li\u003e\n \u003cli\u003eK. M. Chen, A. K. Wong, O. G. Troyanskaya, J. Zhou. A sequence-based global map of regulatory activity for deciphering human genetics. Nat Genet. 2022;54(7):940\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eK. Chen, J. W. Wallis, M. D. McLellan, D. E. Larson, J. M. Kalicki, C. S. Pohl, et al. BreakDancer: an algorithm for high-resolution mapping of genomic structural variation. Nat Methods. 2009;6(9):677\u0026ndash;81.\u003c/li\u003e\n \u003cli\u003eA. Abyzov, A. E. Urban, M. Snyder, M. Gerstein. CNVnator: an approach to discover, genotype, and characterize typical and atypical CNVs from family and population genome sequencing. Genome Res. 2011;21(6):974\u0026ndash;84.\u003c/li\u003e\n \u003cli\u003eW. McLaren, L. Gil, S. E. Hunt, H. S. Riat, G. R. S. Ritchie, A. Thormann, et al. The Ensembl Variant Effect Predictor. Genome Biol. 2016;17(1):122.\u003c/li\u003e\n \u003cli\u003eV. Geoffroy, J.-B. Lamouche, T. Guignard, S. Nicaise, A. Kress, S. Scheidecker, et al. The AnnotSV webserver in 2023: updated visualization and ranking. Nucleic Acids Res. 2023;51(W1):W39\u0026ndash;W45.\u003c/li\u003e\n \u003cli\u003eV. Geoffroy, Y. Herenger, A. Kress, C. Stoetzel, A. Piton, H. Dollfus, et al. AnnotSV: an integrated tool for structural variations annotation. Bioinformatics. 2018;34(20):3572\u0026ndash;4.\u003c/li\u003e\n \u003cli\u003eE. Sj\u0026ouml;stedt, W. Zhong, L. Fagerberg, M. Karlsson, N. Mitsios, C. Adori, et al. An atlas of the protein-coding genes in the human, pig, and mouse brain. Science. 2020;367(6482).\u003c/li\u003e\n \u003cli\u003eS. M. Weyn-Vanhentenryck, A. Mele, Q. Yan, S. Sun, N. Farny, Z. Zhang, et al. HITS-CLIP and integrative modeling define the Rbfox splicing-regulatory network linked to brain development and autism. Cell Rep. 2014;6(6):1139\u0026ndash;52.\u003c/li\u003e\n \u003cli\u003eY. Lv, L. Wen, W. J. Hu, C. Deng, H. W. Ren, Y. N. Bao, et al. Schizophrenia in the genetic era: a review from development history, clinical features and genomic research approaches to insights of susceptibility genes. Metab Brain Dis. 2024;39(1):147\u0026ndash;71.\u003c/li\u003e\n \u003cli\u003eK. J. Karczewski, L. C. Francioli, G. Tiao, B. B. Cummings, J. Alf\u0026ouml;ldi, Q. Wang, et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020;581(7809):434\u0026ndash;43.\u003c/li\u003e\n \u003cli\u003eA. Bayes, L. N. van de Lagemaat, M. O. Collins, M. D. Croning, I. R. Whittle, J. S. Choudhary, et al. Characterization of the proteome, diseases and evolution of the human postsynaptic density. Nat Neurosci. 2011;14(1):19\u0026ndash;21.\u003c/li\u003e\n \u003cli\u003eJ. C. Darnell, S. J. Van Driesche, C. Zhang, K. Y. Hung, A. Mele, C. E. Fraser, et al. FMRP stalls ribosomal translocation on mRNAs linked to synaptic function and autism. Cell. 2011;146(2):247\u0026ndash;61.\u003c/li\u003e\n \u003cli\u003eS. Chen, R. Fragoza, L. Klei, Y. Liu, J. Wang, K. Roeder, et al. An interactome perturbation framework prioritizes damaging missense mutations for developmental disorders. Nat Genet. 2018;50(7):1032\u0026ndash;40.\u003c/li\u003e\n \u003cli\u003eA. Sugathan, M. Biagioli, C. Golzio, S. Erdin, I. Blumenthal, P. Manavalan, et al. CHD8 regulates neurodevelopmental pathways associated with autism spectrum disorder in neural progenitors. Proc Natl Acad Sci U S A. 2014;111(42):E4468\u0026ndash;77.\u003c/li\u003e\n \u003cli\u003eM. Abdi, E. Aliyev, B. Trost, M. Kohailan, W. Aamer, N. Syed, et al. Genomic architecture of autism spectrum disorder in Qatar: The BARAKA-Qatar Study. Genome Med. 2023;15(1):81.\u003c/li\u003e\n \u003cli\u003eC. International Obsessive Compulsive Disorder Foundation Genetics, O. C. D. C. G. A. Studies. Revealing the complex genetic architecture of obsessive-compulsive disorder using meta-analysis. Mol Psychiatry. 2018;23(5):1181\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eS. Ripke, C. O\u0026apos;Dushlaine, K. Chambert, J. L. Moran, A. K. Kahler, S. Akterin, et al. Genome-wide association analysis identifies 13 new risk loci for schizophrenia. Nat Genet. 2013;45(10):1150\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eC. Schizophrenia Working Group of the Psychiatric Genomics. Biological insights from 108 schizophrenia-associated genetic loci. Nature. 2014;511(7510):421\u0026ndash;7.\u003c/li\u003e\n \u003cli\u003eA. F. Pardinas, P. Holmans, A. J. Pocklington, V. Escott-Price, S. Ripke, N. Carrera, et al. Common schizophrenia alleles are enriched in mutation-intolerant genes and in regions under strong background selection. Nat Genet. 2018;50(3):381\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eM. Lam, C. Y. Chen, Z. Li, A. R. Martin, J. Bryois, X. Ma, et al. Comparative genetic architectures of schizophrenia in East Asian and European populations. Nat Genet. 2019;51(12):1670\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eV. Trubetskoy, A. F. Pardinas, T. Qi, G. Panagiotaropoulou, S. Awasthi, T. B. Bigdeli, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. 2022;604(7906):502\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eL. D. Ward, M. Kellis. HaploReg: a resource for exploring chromatin states, conservation, and regulatory motif alterations within sets of genetically linked variants. Nucleic Acids Res. 2012;40(D1):D930\u0026ndash;D4.\u003c/li\u003e\n \u003cli\u003eT. Wu, E. Hu, S. Xu, M. Chen, P. Guo, Z. Dai, et al. clusterProfiler 4.0: A universal enrichment tool for interpreting omics data. The Innovation. 2021;2(3).\u003c/li\u003e\n \u003cli\u003eG. Yu, L.-G. Wang, Y. Han, Q.-Y. He. clusterProfiler: an R Package for Comparing Biological Themes Among Gene Clusters. OMICS: A Journal of Integrative Biology. 2012;16(5):284\u0026ndash;7.\u003c/li\u003e\n \u003cli\u003eM. V. Fern\u0026aacute;ndez, J. Budde, J. L. Del-Aguila, L. Iba\u0026ntilde;ez, Y. Deming, O. Harari, et al. Evaluation of Gene-Based Family-Based Methods to Detect Novel Genes Associated With Familial Late Onset Alzheimer Disease. Front Neurosci. 2018;12.\u003c/li\u003e\n \u003cli\u003eJ. Hecker, F. W. Townes, P. Kachroo, C. Laurie, J. Lasky-Su, J. Ziniti, et al. A unifying framework for rare variant association testing in family-based designs, including higher criticism approaches, SKATs, and burden tests. Bioinformatics. 2021;36(22-23):5432\u0026ndash;8.\u003c/li\u003e\n \u003cli\u003eS. Lee, M. J. Emond, M. J. Bamshad, K. C. Barnes, M. J. Rieder, D. A. Nickerson, et al. Optimal unified approach for rare-variant association testing with application to small-sample case-control whole-exome sequencing studies. Am J Hum Genet. 2012;91(2):224\u0026ndash;37.\u003c/li\u003e\n \u003cli\u003eY. Liu, S. Chen, Z. Li, A. C. Morrison, E. Boerwinkle, X. Lin. ACAT: A Fast and Powerful p Value Combination Method for Rare-Variant Analysis in Sequencing Studies. Am J Hum Genet. 2019;104(3):410\u0026ndash;21.\u003c/li\u003e\n \u003cli\u003eZ. Liu, W. Qian, W. Cai, W. Song, W. Wang, D. T. Maharjan, et al. Inferring the Effects of Protein Variants on Protein\u0026ndash;Protein Interactions with Interpretable Transformer Representations. Research.6:0219.\u003c/li\u003e\n \u003cli\u003eR. Oughtred, J. Rust, C. Chang, B. J. Breitkreutz, C. Stark, A. Willems, et al. The BioGRID database: A comprehensive biomedical resource of curated protein, genetic, and chemical interactions. Protein Sci. 2021;30(1):187\u0026ndash;200.\u003c/li\u003e\n \u003cli\u003eP. Shannon, A. Markiel, O. Ozier, N. S. Baliga, J. T. Wang, D. Ramage, et al. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003;13(11):2498\u0026ndash;504.\u003c/li\u003e\n \u003cli\u003eY. Zhou, B. Zhou, L. Pache, M. Chang, A. H. Khodabakhshi, O. Tanaseichuk, et al. Metascape provides a biologist-oriented resource for the analysis of systems-level datasets. Nat Commun. 2019;10(1):1523.\u003c/li\u003e\n \u003cli\u003eK. Siletti, R. Hodge, A. Mossi Albiach, K. W. Lee, S. L. Ding, L. Hu, et al. Transcriptomic diversity of cell types across the adult human brain. Science. 2023;382(6667):eadd7046.\u003c/li\u003e\n \u003cli\u003eN. G. Skene, S. G. Grant. Identification of Vulnerable Cell Types in Major Brain Disorders Using Single Cell Transcriptomes and Expression Weighted Cell Type Enrichment. Front Neurosci. 2016;10:16.\u003c/li\u003e\n \u003cli\u003eP. S. Emani, J. J. Liu, D. Clarke, M. Jensen, J. Warrell, C. Gupta, et al. Single-cell genomics and regulatory networks for 388 human brains. Science. 2024;384(6698):eadi5199.\u003c/li\u003e\n \u003cli\u003eS. E. Ahmari, S. L. Rauch. The prefrontal cortex and OCD. Neuropsychopharmacology. 2022;47(1):211\u0026ndash;24.\u003c/li\u003e\n \u003cli\u003eP. Langfelder, S. Horvath. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9(1):559.\u003c/li\u003e\n \u003cli\u003eS. Petrovski, A. B. Gussow, Q. Wang, M. Halvorsen, Y. Han, W. H. Weir, et al. The Intolerance of Regulatory Sequence to Genetic Variation Predicts Gene Dosage Sensitivity. PLoS Genet. 2015;11(9):e1005492.\u003c/li\u003e\n \u003cli\u003eN. Delphin, C. Aust, L. Griffiths, F. Fernandez. Epigenetic Regulation in Schizophrenia: Focus on Methylation and Histone Modifications in Human Studies. Genes (Basel). 2024;15(3).\u003c/li\u003e\n \u003cli\u003eM. G. Butler, A. B. McGuire, H. Masoud, A. M. Manzardo. Currently recognized genes for schizophrenia: High-resolution chromosome ideogram representation. Am J Med Genet B Neuropsychiatr Genet. 2016;171B(2):181\u0026ndash;202.\u003c/li\u003e\n \u003cli\u003eN. E. Clifton, L. Collado-Torres, E. E. Burke, A. F. Pardinas, J. C. Harwood, A. Di Florio, et al. Developmental Profile of Psychiatric Risk Associated With Voltage-Gated Cation Channel Activity. Biol Psychiatry. 2021;90(6):399\u0026ndash;408.\u003c/li\u003e\n \u003cli\u003eR. Zhai, G. Tong, Z. Li, W. Song, Y. Hu, S. Xu, et al. Rhesus monkeys exhibiting spontaneous ritualistic behaviors resembling obsessive-compulsive disorder. Natl Sci Rev. 2023;10(11):nwad312.\u003c/li\u003e\n \u003cli\u003eM. Deng, X. Li, D. Shi, Q. Fan, H. Zhang, Z. Wang, et al. iTRAQ-Based Serum Proteomic Analysis Reveals Multifactorial Cellular Function Impairment and Aggravated Systematic Inflammation in Drug-free Obsessive-Compulsive Disorders. ACS Chem Neurosci. 2024.\u003c/li\u003e\n \u003cli\u003eK. C. de Oliveira, C. Camilo, V. D. Gastaldi, A. Sant\u0026apos;Anna Feltrin, B. C. G. Lisboa, V. de Jesus Rodrigues de Paula, et al. Brain areas involved with obsessive-compulsive disorder present different DNA methylation modulation. BMC Genom Data. 2021;22(1):45.\u003c/li\u003e\n \u003cli\u003eJ. J. Moffat, M. Ka, E. M. Jung, A. L. Smith, W. Y. Kim. The role of MACF1 in nervous system development and maintenance. Semin Cell Dev Biol. 2017;69:9\u0026ndash;17.\u003c/li\u003e\n \u003cli\u003eQ. Wang, M. Li, Z. Yang, X. Hu, H. M. Wu, P. Ni, et al. Increased co-expression of genes harboring the damaging de novo mutations in Chinese schizophrenic patients during prenatal development. Sci Rep. 2015;5:18209.\u003c/li\u003e\n \u003cli\u003eP. J. Harrison, D. R. Weinberger. Schizophrenia genes, gene expression, and neuropathology: on the matter of their convergence. Mol Psychiatry. 2005;10(1):40\u0026ndash;68.\u003c/li\u003e\n \u003cli\u003eD. Salem, R. J. Fecek. Role of microtubule actin crosslinking factor 1 (MACF1) in bipolar disorder pathophysiology and potential in lithium therapeutic mechanism. Transl Psychiatry. 2023;13(1):221.\u003c/li\u003e\n \u003cli\u003eJ. Pol-Fuster, F. Canellas, L. Ruiz-Guerra, A. Medina-Dols, B. Bisbal-Carrio, V. Asensio, et al. Familial Psychosis Associated With a Missense Mutation at MACF1 Gene Combined With the Rare Duplications DUP3p26.3 and DUP16q23.3, Affecting the CNTN6 and CDH13 Genes. Front Genet. 2021;12:622886.\u003c/li\u003e\n \u003cli\u003eH. H. Jensen, M. Brohus, J. W. Hussey, A.-O. Busuioc, E. D. Iversen, F. Darki, et al. Calmodulin variants in schizophrenia patients display gain-of-function or loss-of-function effects. medRxiv. 2024:2024.05.22.24307674.\u003c/li\u003e\n \u003cli\u003eF. Vidal-Domenech, G. Riquelme, R. Pinacho, R. Rodriguez-Mias, A. Vera, A. Monje, et al. Calcium-binding proteins are altered in the cerebellum in schizophrenia. PLoS One. 2020;15(7):e0230400.\u003c/li\u003e\n \u003cli\u003eE. Parnell, L. Culotta, M. P. Forrest, H. A. Jalloul, B. L. Eckman, D. D. Loizzo, et al. Excitatory Dysfunction Drives Network and Calcium Handling Deficits in 16p11.2 Duplication Schizophrenia Induced Pluripotent Stem Cell-Derived Neurons. Biol Psychiatry. 2023;94(2):153\u0026ndash;63.\u003c/li\u003e\n \u003cli\u003eK.-C. Huang, K.-C. Yang, H. Lin, T. T.-H. Tsao, S.-A. Lee. Transcriptome alterations of mitochondrial and coagulation function in schizophrenia by cortical sequencing analysis. BMC Genomics. 2014;15(9):S6.\u003c/li\u003e\n \u003cli\u003eX. Weng, D. Zhou, F. Liu, H. Zhang, J. Ye, Z. Zhang, et al. DNA methylation profiling in the thalamus and hippocampus of postnatal malnourished mice, including effects related to long-term potentiation. BMC Neurosci. 2014;15(1):31.\u003c/li\u003e\n \u003cli\u003eK. H. Lim, S. Yang, S. H. Kim, E. Ko, M. Kang, J. Y. Joo. Cryptic mutations of PLC family members in brain disorders: recent discoveries and a deep-learning-based approach. Brain. 2023;146(4):1267\u0026ndash;80.\u003c/li\u003e\n \u003cli\u003eD. Malhotra, S. McCarthy, J. J. Michaelson, V. Vacic, K. E. Burdick, S. Yoon, et al. High frequencies of de novo CNVs in bipolar disorder and schizophrenia. Neuron. 2011;72(6):951\u0026ndash;63.\u003c/li\u003e\n \u003cli\u003eR. Trifiletti, H. M. Lachman, O. Manusama, D. Zheng, A. Spalice, P. Chiurazzi, et al. Identification of ultra-rare genetic variants in pediatric acute onset neuropsychiatric syndrome (PANS) by exome and whole genome sequencing. Sci Rep. 2022;12(1):11106.\u003c/li\u003e\n \u003cli\u003eK. A. Demin, N. A. Krotova, N. P. Ilyin, D. S. Galstyan, T. O. Kolesnikova, T. Strekalova, et al. Evolutionarily conserved gene expression patterns for affective disorders revealed using cross-species brain transcriptomic analyses in humans, rats and zebrafish. Sci Rep. 2022;12(1):20836.\u003c/li\u003e\n \u003cli\u003eA. Patowary, P. Zhang, C. Jops, C. K. Vuong, X. Ge, K. Hou, et al. Developmental isoform diversity in the human neocortex informs neuropsychiatric risk mechanisms. Science. 2024;384(6698):eadh7688.\u003c/li\u003e\n \u003cli\u003eJ. C. Silbereis, S. Pochareddy, Y. Zhu, M. Li, N. Sestan. The Cellular and Molecular Landscapes of the Developing Human Central Nervous System. Neuron. 2016;89(2):248\u0026ndash;68.\u003c/li\u003e\n \u003cli\u003eR. Birnbaum, D. R. Weinberger. The Genesis of Schizophrenia: An Origin Story. Am J Psychiatry. 2024;181(6):482\u0026ndash;92.\u003c/li\u003e\n \u003cli\u003eX. Mao, J. F. Staiger. Multimodal cortical neuronal cell type classification. Pflugers Arch. 2024;476(5):721\u0026ndash;33.\u003c/li\u003e\n \u003cli\u003eL. Nahar, B. M. Delacroix, H. W. Nam. The Role of Parvalbumin Interneurons in Neurotransmitter Balance and Neurological Disease. Front Psychiatry. 2021;12:679960.\u003c/li\u003e\n \u003cli\u003eS. J. Kaar, I. Angelescu, T. R. Marques, O. D. Howes. Pre-frontal parvalbumin interneurons in schizophrenia: a meta-analysis of post-mortem studies. J Neural Transm (Vienna). 2019;126(12):1637\u0026ndash;51.\u003c/li\u003e\n \u003cli\u003eJ. A. Cale, E. J. Chauhan, J. J. Cleaver, A. R. Fusciardi, S. McCann, H. C. Waters, et al. GABAergic and inflammatory changes in the frontal cortex following neonatal PCP plus isolation rearing, as a dual-hit neurodevelopmental model for schizophrenia. Mol Neurobiol. 2024.\u003c/li\u003e\n \u003cli\u003eO. Marin. Parvalbumin interneuron deficits in schizophrenia. Eur Neuropsychopharmacol. 2024;82:44\u0026ndash;52.\u003c/li\u003e\n \u003cli\u003eA. J. Price, A. E. Jaffe, D. R. Weinberger. Cortical cellular diversity and development in schizophrenia. Mol Psychiatry. 2021;26(1):203\u0026ndash;17.\u003c/li\u003e\n \u003cli\u003eM. C. Tiveron, C. Beurrier, C. Ceni, N. Andriambao, A. Combes, M. Koehl, et al. LAMP5 Fine-Tunes GABAergic Synaptic Transmission in Defined Circuits of the Mouse Brain. PLoS One. 2016;11(6):e0157052.\u003c/li\u003e\n \u003cli\u003eM. Y. Batiuk, T. Tyler, K. Dragicevic, S. Mei, R. Rydbirk, V. Petukhov, et al. Upper cortical layer\u0026ndash;driven network impairment in schizophrenia. Science Advances.8(41):eabn8367.\u003c/li\u003e\n \u003cli\u003eW. Cai, W. Song, Z. Liu, D. T. Maharjan, J. Liang, G. N. Lin. An Integrative Analysis of Identified Schizophrenia-Associated Brain Cell Types and Gene Expression Changes. Int J Mol Sci. 2022;23(19).\u003c/li\u003e\n \u003cli\u003eA. Rampino, T. Annese, S. Torretta, R. Tamma, R. M. Falcone, D. Ribatti. Involvement of vascular endothelial growth factor in schizophrenia. Neurosci Lett. 2021;760:136093.\u003c/li\u003e\n \u003cli\u003eS. Pong, R. Karmacharya, M. Sofman, J. R. Bishop, P. Lizano. The Role of Brain Microvascular Endothelial Cell and Blood-Brain Barrier Dysfunction in Schizophrenia. Complex Psychiatry. 2020;6(1-2):30\u0026ndash;46.\u003c/li\u003e\n \u003cli\u003eS. Zengil, E. Laloglu. Evaluation of serum zonulin and occludin levels in obsessive-compulsive disorder and the effect of major depressive disorder comorbidity. Front Psychiatry. 2024;15:1395235.\u003c/li\u003e\n \u003cli\u003eN. W. McGregor, C. Lochner, D. J. Stein, S. M. Hemmings. Polymorphisms within the neuronal cadherin (CDH2) gene are associated with obsessive-compulsive disorder (OCD) in a South African cohort. Metab Brain Dis. 2016;31(1):191\u0026ndash;6.\u003c/li\u003e\n \u003cli\u003eM. Zamanian-Azodi, M. Rezaei-Tavirani, M. Rezaei Tavirani. Obsessive-Compulsive Disorder Interactome Profile Analysis: A Perspective From Molecular Mechanism. Basic Clin Neurosci. 2021;12(2):187\u0026ndash;98.\u003c/li\u003e\n \u003cli\u003eA. Bilgi\u0026ccedil;, B. Akbaş İleri, İ. Kılın\u0026ccedil;, \u0026Ouml;. F. Ak\u0026ccedil;a. Decreased plasma intercellular adhesion molecule-3 levels are associated with obsessive compulsive disorder in medication-free children. J. Obsessive Compuls. Relat. Disord. 2020;27.\u003c/li\u003e\n \u003cli\u003eI. Stankovic, M. Notaras, P. Wolujewicz, T. Lu, R. Lis, M. E. Ross, et al. Schizophrenia endothelial cells exhibit higher permeability and altered angiogenesis patterns in patient-derived organoids. Transl Psychiatry. 2024;14(1):53.\u003c/li\u003e\n \u003cli\u003eS. Stanca, M. Rossetti, L. Bokulic Panichi, P. Bongioanni. The Cellular Dysfunction of the Brain-Blood Barrier from Endothelial Cells to Astrocytes: The Pathway towards Neurotransmitter Impairment in Schizophrenia. Int J Mol Sci. 2024;25(2).\u003c/li\u003e\n \u003cli\u003eA. Vreeland, D. Calaprice, N. Or-Geva, R. E. Frye, D. Agalliu, H. M. Lachman, et al. Postinfectious Inflammation, Autoimmunity, and Obsessive-Compulsive Disorder: Sydenham Chorea, Pediatric Autoimmune Neuropsychiatric Disorder Associated with Streptococcal Infection, and Pediatric Acute-Onset Neuropsychiatric Disorder. Dev Neurosci. 2023;45(6):361\u0026ndash;74.\u003c/li\u003e\n \u003cli\u003eD. Endres, T. A. Pollak, K. Bechter, D. Denzel, K. Pitsch, K. Nickel, et al. Immunological causes of obsessive-compulsive disorder: is it time for the concept of an \u0026quot;autoimmune OCD\u0026quot; subtype? Transl Psychiatry. 2022;12(1):5.\u003c/li\u003e\n \u003cli\u003eS. Bhattacharyya, S. Khanna, K. Chakrabarty, A. Mahadevan, R. Christopher, S. K. Shankar. Anti-brain autoantibodies and altered excitatory neurotransmitters in obsessive-compulsive disorder. Neuropsychopharmacology. 2009;34(12):2489\u0026ndash;96.\u003c/li\u003e\n \u003cli\u003eD. Ribatti, D. Guidolin. Morphogenesis of vascular and neuronal networks and the relationships between their remodeling processes. Brain Res Bull. 2022;186:62\u0026ndash;9.\u003c/li\u003e\n \u003cli\u003eJ. Kealy, C. Greene, M. Campbell. Blood-brain barrier regulation in psychiatric disorders. Neurosci Lett. 2020;726:133664.\u003c/li\u003e\n \u003cli\u003eE. L. Goldwaser, D. J. J. Wang, B. M. Adhikari, J. Chiappelli, X. Shao, J. Yu, et al. Evidence of Neurovascular Water Exchange and Endothelial Vascular Dysfunction in Schizophrenia: An Exploratory Study. Schizophr Bull. 2023;49(5):1325\u0026ndash;35.\u003c/li\u003e\n \u003cli\u003eC. Y. Chang, D. Z. Luo, J. C. Pei, M. C. Kuo, Y. C. Hsieh, W. S. Lai. Not Just a Bystander: The Emerging Role of Astrocytes and Research Tools in Studying Cognitive Dysfunctions in Schizophrenia. Int J Mol Sci. 2021;22(10).\u003c/li\u003e\n \u003cli\u003eS. Kiarash, B. Mohammad, A. Banafshe, P. Zeynab, E. Zahra, A. Shiva, et al. Astrocytic Abnormalities in Schizophrenia. In: Thomas H, editor. Neurophysiology. Rijeka: IntechOpen; 2022. p. Ch. 12.\u003c/li\u003e\n \u003cli\u003eE. C. de Oliveira Figueiredo, C. Cali, F. Petrelli, P. Bezzi. Emerging evidence for astrocyte dysfunction in schizophrenia. Glia. 2022;70(9):1585\u0026ndash;604.\u003c/li\u003e\n \u003cli\u003eV. V. Tarasov, A. A. Svistunov, V. N. Chubarev, S. S. Sologova, P. Mukhortova, D. Levushkin, et al. Alterations of Astrocytes in the Context of Schizophrenic Dementia. Front Pharmacol. 2019;10:1612.\u003c/li\u003e\n \u003cli\u003eD. Senko, O. Efimova, M. Osetrova, N. Anikanov, M. Boyko, M. Sharaev, et al. White matter lipidome alterations in the schizophrenia brain. Schizophrenia (Heidelb). 2024;10(1):123.\u003c/li\u003e\n \u003cli\u003eM. Vald\u0026eacute;s-Tovar, A. M. Rodr\u0026iacute;guez-Ram\u0026iacute;rez, L. Rodr\u0026iacute;guez-C\u0026aacute;rdenas, C. E. Sotelo-Ram\u0026iacute;rez, B. Camarena, M. A. Sanabrais-Jim\u0026eacute;nez, et al. Insights into myelin dysfunction in schizophrenia and bipolar disorder. World J Psychiatry. 2022;12(2):264\u0026ndash;85.\u003c/li\u003e\n \u003cli\u003eL. Bast, S. Yao, J. A. Martinez-Lopez, F. Memic, H. French, M. Valiukonyte, et al. Transcriptomic and genetic analysis suggests a role for mitochondrial dysregulation in schizophrenia. medRxiv. 2025.\u003c/li\u003e\n \u003cli\u003eS. L. Mondragon-Gonzalez, C. Schreiweis, E. Burguiere. Closed-loop recruitment of striatal interneurons prevents compulsive-like grooming behaviors. Nat Neurosci. 2024;27(6):1148\u0026ndash;56.\u003c/li\u003e\n \u003cli\u003eS. M. Perez, A. Boley, D. J. Lodge. Region specific knockdown of Parvalbumin or Somatostatin produces neuronal and behavioral deficits consistent with those observed in schizophrenia. Transl Psychiatry. 2019;9(1):264.\u003c/li\u003e\n \u003cli\u003eA. J. Murray, M. U. Woloszynowska-Fraser, L. Ansel-Bollepalli, K. L. Cole, A. Foggetti, B. Crouch, et al. Parvalbumin-positive interneurons of the prefrontal cortex support working memory and cognitive flexibility. Sci Rep. 2015;5:16778.\u003c/li\u003e\n \u003cli\u003eA. M. Frota Lisboa Pereira de Souza, L. Pellegrini, N. A. Fineberg. Cognitive inflexibility, obsessive-compulsive symptoms and traits and poor post-pandemic adjustment. Neuroscience Applied. 2024;3.\u003c/li\u003e\n \u003cli\u003eP. Gruner, C. Pittenger. Cognitive inflexibility in Obsessive-Compulsive Disorder. Neuroscience. 2017;345:243\u0026ndash;55.\u003c/li\u003e\n \u003cli\u003eP. Monteiro, B. Barak, Y. Zhou, R. McRae, D. Rodrigues, I. R. Wickersham, et al. Dichotomous parvalbumin interneuron populations in dorsolateral and dorsomedial striatum. J Physiol. 2018;596(16):3695\u0026ndash;707.\u003c/li\u003e\n \u003cli\u003eA. Caballero, E. Flores-Barrera, D. R. Thomases, K. Y. Tseng. Downregulation of parvalbumin expression in the prefrontal cortex during adolescence causes enduring prefrontal disinhibition in adulthood. Neuropsychopharmacology. 2020;45(9):1527\u0026ndash;35.\u003c/li\u003e\n \u003cli\u003eF. J. Raabe, L. Slapakova, M. J. Rossner, L. Cantuti-Castelvetri, M. Simons, P. G. Falkai, et al. Oligodendrocytes as A New Therapeutic Target in Schizophrenia: From Histopathological Findings to Neuron-Oligodendrocyte Interaction. Cells. 2019;8(12).\u003c/li\u003e\n \u003cli\u003eP. Falkai, M. J. Rossner, F. J. Raabe, E. Wagner, D. Keeser, I. Maurus, et al. Disturbed Oligodendroglial Maturation Causes Cognitive Dysfunction in Schizophrenia: A New Hypothesis. Schizophr Bull. 2023;49(6):1614\u0026ndash;24.\u003c/li\u003e\n \u003cli\u003eN. Takahashi, T. Sakurai, K. L. Davis, J. D. Buxbaum. Linking oligodendrocyte and myelin dysfunction to neurocircuitry abnormalities in schizophrenia. Prog Neurobiol. 2011;93(1):13\u0026ndash;24.\u003c/li\u003e\n \u003cli\u003eJ. Davis, H. Eyre, F. N. Jacka, S. Dodd, O. Dean, S. McEwen, et al. A review of vulnerability and risks for schizophrenia: Beyond the two hit hypothesis. Neurosci Biobehav Rev. 2016;65:185\u0026ndash;94.\u003c/li\u003e\n \u003cli\u003eO. D. Howes, E. C. Onwordi. The synaptic hypothesis of schizophrenia version III: a master mechanism. Mol Psychiatry. 2023;28(5):1843\u0026ndash;56.\u003c/li\u003e\n \u003cli\u003eL. D. Selemon, N. Zecevic. Schizophrenia: a tale of two critical periods for prefrontal cortical development. Transl Psychiatry. 2015;5(8):e623.\u003c/li\u003e\n \u003cli\u003eE. J. Nestler, C. J. Pena, M. Kundakovic, A. Mitchell, S. Akbarian. Epigenetic Basis of Mental Illness. Neuroscientist. 2016;22(5):447\u0026ndash;63.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 5 are available in the supplementary files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-8943594/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8943594/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePsychiatric disorders often co-occur and share liability, yet why distinct diagnoses arise under a largely shared familial genetic background remains poorly understood. We assembled 21 Han Chinese mixed pedigrees co-ascertained for obsessive\u0026ndash;compulsive disorder (OCD) and schizophrenia (SCZ) (21 OCD, 21 SCZ, and 38 unaffected first-degree relatives) and performed deep (60\u0026times;) whole-genome sequencing. Leveraging within-family structure as an internal control, we integrated rare coding and regulatory variation across co-segregating and disorder-biased loci and evaluated diagnostic separation using stratified group cross-validation to prevent relatedness-driven leakage. Mutation-burden models distinguished OCD from controls (AUC 0.839), SCZ from controls (AUC 0.749), and affected individuals from controls (AUC 0.792), whereas OCD\u0026ndash;SCZ discrimination remained modest (AUC 0.631), consistent with partial genetic sharing. Burden decomposition suggested that coding-region signals accounted for most of the discriminative performance, while non-coding burden provided limited incremental contribution under current annotations. Developmental network mapping nominated temporally stratified prenatal-to-postnatal modules, including a late-pregnancy angiogenesis-related module shared across disorders and SCZ-biased astrocyte/calcium-related programs. Together, these results illustrate how mixed-pedigree WGS can help disentangle convergent versus divergent rare-variant architectures across OCD and SCZ and provide family-grounded, interpretable signatures for future disorder-specific stratification.\u003c/p\u003e","manuscriptTitle":"Whole-genome sequencing of mixed OCD–schizophrenia pedigrees characterizes shared and divergent rare-variant architectures","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-03-18 04:56:07","doi":"10.21203/rs.3.rs-8943594/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a68528a6-1f64-4de6-8909-75d97eef232b","owner":[],"postedDate":"March 18th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Reject after peer review","date":"2026-05-11T14:45:43+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":64611094,"name":"Biological sciences/Genetics"},{"id":64611095,"name":"Biological sciences/Molecular biology"},{"id":64611096,"name":"Biological sciences/Neuroscience"},{"id":64611097,"name":"Health sciences/Diseases/Psychiatric disorders/Schizophrenia"}],"tags":[],"updatedAt":"2026-05-11T14:52:47+00:00","versionOfRecord":[],"versionCreatedAt":"2026-03-18 04:56:07","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8943594","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8943594","identity":"rs-8943594","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}