Cross-disorder genomic structural equation modeling reveals common genetic basis of neuroimmune diseases

Cross-disorder genomic structural equation modeling reveals common genetic basis of neuroimmune diseases | 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 Cross-disorder genomic structural equation modeling reveals common genetic basis of neuroimmune diseases Haifeng Chen¹, Yuxiong Liao¹, Luejun Tang¹, Xiaoyun Wei¹, Tongshun Li¹, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7328766/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 The genetic architecture underlying neuroimmune syndrome-related traits remains poorly characterized. We employed genomic structural equation modeling (genomic SEM) alongside comprehensive post-GWAS analytical frameworks to identify causal single nucleotide polymorphisms (SNPs) associated with phenotypic variance independent of measured traits, revealing 33 genome-wide significant loci. Multi-tissue transcriptome-wide association analyses were conducted across tissue, cellular, and genomic regulatory elements to characterize susceptibility gene signals and regulatory components with high relevance to neuroimmune syndrome GWAS. We subsequently leveraged extensive human disease datasets to determine neuroimmune syndrome-associated risk factors and explored potential therapeutic targets through plasma proteomics and drug target prioritization analyses. Additionally, summary statistics-based polygenic scoring assessed chromosomal contributions to neuroimmune syndrome risk. Our investigation represents the first comprehensive genetic architecture mapping of neuroimmune syndromes through GWAS of a latent, unmeasured phenotype, providing unprecedented insights into shared genetic mechanisms underlying these complex disorders. Biological sciences/Computational biology and bioinformatics Health sciences/Diseases Biological sciences/Genetics Health sciences/Neurology Biological sciences/Neuroscience genomic structural equation modeling neuroimmune syndromes genome-wide association study transcriptome-wide association study Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Introduction A century ago, upon discovering the blood-brain barrier, scientists considered the brain an "immune-privileged organ". However, we now recognize that this "privilege" masks the most complex interactions between the nervous and immune systems[1,2]. Neuroimmune syndromes (NIS) represent not merely a disease spectrum concept, but rather a complex, multidimensional pathobiological process, characterized by blood-brain barrier dysfunction and neuroimmune homeostatic imbalance, profoundly influenced by genetic susceptibility, environmental triggers, and immune system dysregulation[3-5]. With the dramatic global increase in autoimmune disease incidence, the prevalence of neuroimmune syndromes has risen rapidly, emerging as a major challenge in neuroscience, immunology, and medical genetics[6,7]. Despite significant advances in understanding molecular mechanisms and targeted therapies for individual neuroimmune diseases, our comprehension of the specific shared genetic and cross-disease biological mechanisms underlying neuroimmune syndromes remains limited. Studies indicate that neuroinflammatory responses, autoimmune attacks, and myelin damage dysfunction may constitute important drivers of neuroimmune syndromes, yet these findings remain insufficient to fully explain the substantial inter-individual variation in neuroimmune disease progression and susceptibility[8,9]. To address these challenges, this investigation aims to integrate multiple genetic analytical tools and robust association methods to unveil potential common molecular mechanisms and expand the genetic connections between neuroimmune syndromes and various related disorders. Specifically, we focus on pleiotropic genomic loci and critical chromosomal regions associated with neuroimmune syndromes to reveal potential therapeutic targets. This study not only expands our understanding of neuroimmune syndromes but also provides theoretical and practical support for precision medicine intervention strategies in global neuroimmune diseases[10]. To address the current lack of precise measurement of common mechanisms in neuroimmune syndromes, we designed an innovative GWAS approach targeting latent, unmeasured neuroimmune syndrome phenotypes. We employed genomic structural equation modeling (Genomic SEM) applied to published GWAS summary statistics of neuroimmune-related diseases and immune biomarkers[11]. Through these statistics, we obtained SNP association strengths with latent neuroimmune syndrome phenotypes, thereby establishing a GWAS study of previously unmeasured latent neuroimmune syndrome phenotypes. We further adapted comprehensive analytical methods from systems biology, defining genetic variants in our neuroimmune syndrome structural equation model that remain unexplained by known immune biomarkers as potential cross-disease shared genetic markers. These underwent extensive GWAS-related functional annotation and pathway enrichment analyses. While this approach may not perfectly capture the true relationships between neuroimmune disease-related pathways and multifactorial interactions, as neuroimmune syndromes represent complex processes driven jointly by genetic, environmental, and stochastic immune dysregulation factors, this analysis effectively excludes confounding influences based on single neuroimmune disease markers, enabling precise analysis of previously difficult-to-study cross-disease common mechanisms. From a direct clinical application perspective, we conducted tens of thousands of association analyses to construct a concise, practical genetic risk factor atlas for non-biostatistical professionals (neurologists, rheumatologists), enabling direct application of relevant risk factor maps for developing potential individualized prevention and precision intervention strategies for patients. Our research aims to establish a simple yet efficient translational pathway from genomic statistics to neuroimmune basic research and clinical precision medicine strategy development. Methods A flowchart overview is presented in Fig.1. Study Design and Data Sources Our genomic structural equation modeling GWAS for neuroimmune syndromes utilized summary statistics from five independent GWAS studies encompassing neuroimmune-related disorders: Guillain-Barré syndrome (GBS), myasthenia gravis (MG), multiple sclerosis (MS), systemic lupus erythematosus (SLE), and systemic connective tissue diseases (SCTD). All constituent GWAS studies received institutional review board approval, with informed consent obtained from all participants, and underwent rigorous quality control procedures. Guillain-Barré Syndrome (GBS): GWAS data were obtained from the FinnGen Research Program Release 12 (n = 492,134; 551 cases, 491,583 controls)[12]. The FinnGen initiative represents a large-scale Finnish biobank study designed to provide comprehensive insights into genetic foundations and risk factors for rare neuroimmune disorders through integration of genome-wide sequencing and health registry data. This dataset leverages the unique genetic architecture of the Finnish population, enabling enhanced detection of rare acute inflammatory demyelinating polyneuropathies. Myasthenia Gravis (MG): Summary statistics were derived from FinnGen R12 (n = 496,227; 560 cases, 495,667 controls). This investigation utilized comprehensive diagnostic data from the Finnish healthcare registry system combined with large-scale genotyping, providing valuable genetic information for neuromuscular junction autoimmune disorders. Multiple Sclerosis (MS): GWAS data originated from FinnGen R12 (n = 498,857; 2,926 cases, 495,931 controls). This study employed stringent phenotype definitions and quality control procedures, establishing a crucial foundation for central nervous system demyelinating disease genetics research. As the most prevalent phenotype within our neuroimmune syndrome framework, MS provided substantial statistical power. Systemic Lupus Erythematosus (SLE): Data were obtained from FinnGen R12 (n = 499,333; 291 cases, 499,042 controls). As a prototypical systemic autoimmune disease, this study provided foundational insights into genetic mechanisms underlying systemic immune dysregulation components of neuroimmune syndromes. Systemic Connective Tissue Diseases (SCTD): GWAS data were sourced from FinnGen R12 (n = 500,348; 16,088 cases, 484,260 controls). SCTD, representing mixed connective tissue disorders frequently involving neurological manifestations, contributed the largest case sample size, providing substantial genetic evidence and statistical power for understanding systemic pathological mechanisms in neuroimmune syndromes. All GWAS datasets underwent comprehensive quality control procedures, including sample quality filtering, single nucleotide polymorphism (SNP) quality screening (minor allele frequency [MAF] > 0.01, imputation quality INFO > 0.8), population stratification correction, and relatedness adjustment. To ensure data consistency, all analyses utilized the GRCh37/hg19 reference genome with the 1000 Genomes Project European population as the linkage disequilibrium (LD) reference panel. Detailed GWAS specifications are provided in Supplementary Table 1. Quality Control Procedures Sample-level filtering: Samples with missing call rates exceeding 5% were excluded from analyses. MHC region handling: Given the genetic diversity and structural complexity of the major histocompatibility complex (MHC) region (chromosome 6: 25,000,000-35,000,000 bp), particularly immune-related gene polymorphisms, specialized processing was applied to this genomic region[13]. SNP-level quality control: For constructing neuroimmune syndrome summary statistics, we employed recommended default quality control parameters, retaining all autosomal SNPs from the five constituent neuroimmune-related GWAS studies after filtering to the 1000 Genomes Phase 3 European reference panel. SNPs with MAF < 0.01 were excluded due to increased error rates from small genotype clusters and typically elevated LD score regression standard errors. Additionally, SNPs with zero effect estimates were removed to prevent matrix singularity issues essential for genomic structural equation modeling. SNPs inconsistent with the reference panel and those with allelic mismatches were also excluded. Sample overlap assessment: Given that constituent single-trait GWAS originated from different genomic repositories with distinct participant cohorts, we carefully considered potential sample intersections across different cohorts to ensure result accuracy and generalizability while accounting for statistical implications of potential sample overlap. Genomic Structural Equation Modeling We implemented genomic structural equation modeling (Genomic SEM) using the GenomicSEM R package (v.0.0.5)[14] to conduct genomic structural equation GWAS analysis across GBS, MG, MS, SLE, and SCTD, investigating broad shared genetic susceptibility underlying these neuroimmune-related phenotypes. Genomic SEM represents a novel multivariate methodology enabling exploration of multiple latent multivariate models to investigate potential common genetic architecture across the neuroimmune disease spectrum. Detailed analytical standards are provided in Table 1. Genomic SEM demonstrates robustness against sample overlap bias (e.g., FinnGen participants appearing across multiple input GWAS disease phenotypes) and sample size imbalances[15]. Additionally, this approach facilitates identification of genetic variants affecting only subsets rather than all neuroimmune phenotypes, thus distinguishing variants representing broad cross-disease susceptibility from those reflecting disease-specific genetic mechanisms. Genomic SEM analysis proceeded through two distinct phases. Phase I estimated empirical genetic covariance matrices and corresponding sampling covariance matrices. We prepared neuroimmune-related disease GWAS summary statistics for Phase I analysis, employing multivariate extensions of cross-trait LD score regression to generate empirical genetic covariance matrices among the five neuroimmune phenotypes as input for SEM common factor modeling (Supplementary File 3). Phase II specified SEM models minimizing discrepancies between hypothesized covariance structures and empirically calculated covariance matrices from Phase I. Given our primary objective of identifying shared genetic architecture underlying the five neuroimmune-related phenotypes, we tested a single-factor model to characterize neuroimmune syndromes. Model fit was evaluated using standardized root mean square residual (SRMR), model χ², Akaike information criterion (AIC), and comparative fit index (CFI) (Supplementary Table S4a-b). Through application of appropriate common factor SEM specifications, individual autosomal SNP associations were incorporated into genetic and corresponding sample covariance matrices, generating neuroimmune syndrome structural equation GWAS results for 7,170,714 SNPs representing shared covariance across the five constituent neuroimmune-related GWAS studies. SNP Heterogeneity Assessment To evaluate whether SNP associations in our neuroimmune syndrome structural equation GWAS were appropriately modeled within the multivariate SEM framework, we calculated SNP heterogeneity statistics (Q_SNP). The null hypothesis posited that SNP associations from individual phenotype GWAS could be completely statistically accounted for by our neuroimmune syndrome structural equation model. Consequently, significant Q_SNP values ( P < 0.05) in our neuroimmune syndrome structural equation GWAS indicated that specific SNPs exerted effects through pathways beyond the established shared genetic mechanisms among neuroimmune-related diseases in our model. This heterogeneity analysis facilitated identification of genetic variants potentially exerting unique effects on specific neuroimmune phenotypes rather than operating through common genetic factors. Multi-level Model Evaluation We implemented multi-tiered strategic adjustments for our genomic structural equation model, including different significance thresholds ( P < 5×10 -12 and P < 5×10 -12 ) to identify novel SNP loci across varying confidence levels, balancing statistical power with false positive control. Simultaneously, we employed genomic control strategies based on two-step LD Score regression methodology. Quality control parameters retained all SNPs with missing values, INFO scores < 0.9, MAF < 0.01, P-values outside conventional ranges, and non-standard or ambiguous strand orientations, exclusively removing partitioned LD scores with zero variance and utilizing two-step estimators for analysis (cutoff threshold = 30). Genomic Loci Definition and Novel Variant Identification We utilized the "Functional Mapping and Annotation of Genetic Associations" methodology implemented in FUMA to identify genomic loci and determine lead SNPs associated with our neuroimmune syndrome structural equation GWAS[16]. These SNPs exhibited low LD correlation with other SNPs (r² < 0.1) while achieving genome-wide significance ( P < 5×10 -8 ). Initially, we input neuroimmune syndrome structural equation SNP summary statistics to assess association strengths. We compared lead SNPs and loci with original univariate GWAS relationships, defining loci as novel when located >1 Mb from previously identified loci in constituent univariate GWAS data. To determine whether the 36 lead SNPs from our neuroimmune syndrome structural equation GWAS exhibited pleiotropic associations, we consulted published significant associations ( P < 5×10 -8 ) in the GWAS Catalog[17]. Additionally, we conducted risk loci analysis using FUMA software functionality with significance thresholds of P < 5×10 -8 , analyzing output files through MAGMA (Multi-marker Analysis of GenoMic Annotation)[18]. MAGMA serves as a post-GWAS processing tool designed to evaluate gene-phenotype associations by aggregating multiple genetic markers into gene-level signals and calculating gene-phenotype association strengths. This approach extracts gene function-related information from genome-wide SNP data for gene-level genetic signal analysis, employing significance thresholds of FDR P < 0.05. Fine-mapping Analysis To identify most probable causal variants associated with our neuroimmune syndrome structural equation GWAS, we employed SuSIE and FINEMAP methodologies implemented in the echolocatoR R package (v.2.0.3). We established posterior probability thresholds of 0.95 for defining credible sets of potential causal variants. Causal variant identification: Both SuSIE (Sum of Single Effects) and FINEMAP represent fine-mapping analytical tools designed to determine most likely causal variants associated with specific phenotypes. We utilized 250 kb windows encompassing regions associated with each lead SNP, calculating causal inference probabilities for each SNP within these regions. Credible sets: We established 0.95 posterior probability thresholds; variants exceeding this threshold were considered potential causal variants. Consensus SNPs: echolocatoR defined 'consensus SNPs' as variants appearing in both SuSIE and FINEMAP results, calculating average posterior probabilities and determining average credible sets based on probability outcomes (credibility = 1 when both SuSIE and FINEMAP SNP posterior probabilities exceeded 0.95, otherwise 0). Transcriptome-wide Association Study Following potential causal variant localization, we conducted transcriptome-wide association study (TWAS) to prioritize genes associated with our neuroimmune syndrome structural equation GWAS based on gene expression-phenotype relationships[19]. We employed the FUSION methodology using 37,920 pre-computed expression quantitative trait loci (eQTL) features (gene/tissue pairs) from GTEx v.8 data. These features facilitated calculation of expression associations across different genes and tissues[20]. TWAS results analysis: Our neuroimmune syndrome structural equation GWAS data contained sufficient variation to analyze 36,149 features (from 37,920 eQTL features), indicating high data quality. Genes with P < 0.05 (significantly associated with our neuroimmune syndrome structural equation GWAS) were included in subsequent analyses. For TWAS-significant genes, we implemented FOCUS methodology (fine-mapping approach specifically designed for TWAS studies)[21]. FOCUS evaluates potential causal relationships between genes and phenotypes based on posterior inclusion probabilities. We considered TWAS-significant genes demonstrating both TWAS significance and consistency with additional evidence (e.g., FOCUS), suggesting potential causality. Gene Set and Disease Ontology Enrichment Analysis We conducted gene enrichment and pathway analyses using MAGMA and FUMA (GSEA) to investigate potential relationships between our neuroimmune syndrome structural equation GWAS and Mendelian disease genes with associated pathways[22]. Additionally, we performed gene enrichment analysis using MendelVar (https://mendelvar.mrcieu.ac.uk/submit/). Cell-type Annotation Analysis To identify etiological cell types associated with our neuroimmune syndrome structural equation GWAS, we employed CELLECT (cell-type expression specificity integration for complex traits using single-cell RNA sequencing data). We utilized the Tabula Muris dataset containing transcriptomic data from 100,000 mouse (Mus musculus) cells across 20 organs and tissues. We preprocessed and normalized Tabula Muris single-cell RNA sequencing data using CELLEX, calculating expression specificity scores for each gene. Cell-type-specific analysis was performed using LDSC software with cell type classification and false discovery rate (FDR) thresholds of 0.05[23]. Genomic Regional Heritability Partitioning We employed LDSC tools to calculate partitioned heritability across genomic regions[24]. This approach assigns phenotypic genetic information to different genomic regions (e.g., genes, enhancers, silencers) to evaluate each region's contribution to phenotypic heritability. Specifically, LDSC utilizes weighted LD matrices, genotype frequency files, and summary statistics for calculations, ultimately estimating genetic contributions from each region. Biomarker and Risk Factor Association Analysis For candidate drug target validation, we utilized expression quantitative trait loci (eQTL) data from the eQTLGen Consortium[25], comprising gene expression measurements from 31,864 individuals of European ancestry. For proteome-wide drug target discovery, we analyzed 4,907 plasma proteins from the deCODE genetics consortium[26]. Additionally, we incorporated brain, cerebrospinal fluid (CSF), and plasma protein quantitative trait loci (pQTL) data from Yang et al.[27], which provided complementary neurologically-relevant proteomic measurements across multiple tissue types. For comprehensive biomarker screening, we conducted phenome-wide association analysis using 50,033 phenotypes from the IEU OpenGWAS database[28]. Summary Statistics-based Polygenic Score Construction We calculated polygenic risk scores (PRS) based on genome-wide summary statistics and evaluated genetic contributions from different chromosomal regions to disease susceptibility[23]. Specifically, we utilized PRS-CS (Polygenic Risk Score with Continuous Shrinkage) software to estimate SNP posterior effect sizes through GWAS data and external LD reference panels, subsequently calculating polygenic risk scores[24]. This methodology employs Bayesian regression models to estimate effect sizes by integrating LD reference panels based on GWAS summary statistics, ultimately computing PRS values. Results Structural Equation Model Statistical Framework Construction LD score regression analysis revealed differential heritability contributions across the five constituent univariate input GWAS studies: GBS (h 2 = 0.0013, Z = 1.56), MG (h 2 = 0.0019, Z = 1.95), MS (h² = 0.0084, Z = 6.61), SLE (h 2 = 0.0011, Z = 1.08), and SCTD (h 2 = 0.0085, Z = 4.90). Pairwise genetic covariance estimates demonstrated substantial shared genetic architecture: GBS-MG (0.5016), GBS-MS (0.1685), GBS-SLE (0.293), GBS-SCTD (0.2469), MG-MS (0.033), MG-SLE (0.2303), MG-SCTD (0.5397), MS-SLE (0.235), MS-SCTD (0.1989), and SLE-SCTD (0.5854) (detailed single-factor genetic parameters in Supplementary Table 2, Figure 2). Structural equation modeling analysis preceding model construction demonstrated excellent fit between the genetic covariance matrix from five input GWAS studies and the empirical covariance matrix under a common factor model (comparative fit index [CFI] = 1, standardized root mean square residual [SRMR] = 0.095) (detailed model stability assessment in Supplementary Table 4a; latent factor [F1] and univariate structural equation model parameters in Supplementary Table 4b). Exploratory factor modeling (Supplementary File 3) provided compelling evidence for shared genetic factors underlying these neuroimmune phenotypes. Genomic Structural Equation Model GWAS Stratified Evaluation Using genomic structural equation modeling (genomic SEM) to incorporate individual genetic variation across multiple neuroimmune disorders, we conducted an indirectly measured genome-wide association study (GWAS) that estimated associations between 7,170,712 single nucleotide polymorphisms (SNPs) and a latent neuroimmune syndrome factor. The complete results are presented in Supplementary Table 5. Under the P < 5×10 -12 threshold, we identified 31 lead SNPs across 41 genomic loci (Supplementary Table 5a), while the more stringent P < 5×10 -16 threshold yielded 29 lead SNPs across 39 genomic loci (Supplementary Table 5b). Among the 7,170,712 SNPs analyzed across different P-value thresholds, 34 represented novel discoveries distinct from loci identified in the five constituent single-trait GWAS studies, highlighting the enhanced discovery power of genomic SEM. These novel neuroimmune syndrome lead SNPs demonstrated enrichment in pathways related to immune system regulation, neurodevelopment, and inflammatory responses. At the P < 5×10 -12 stratification level, 31 of 41 lead SNPs had been previously identified in the literature (though not in the context of neuroimmune structural equation modeling) (Supplementary Table 5c). Similarly, under the P < 5×10 -16 stratification, 29 of 39 lead SNPs showed prior literature identification (Supplementary Table 5d). Key lead SNPs identified include rs148729815, rs76210604, rs17098406, and rs73195472, among others, providing crucial novel insights into neuroimmune syndrome genetic architecture. This investigation, through implementation of genomic structural equation modeling, revealed SNPs associated with neuroimmune syndromes across different association thresholds, offering important insights into shared genetic foundations and potential therapeutic targets. Genomic Control Assessment Based on LD Score Regression Our systematic quality control procedures resulted in exclusion of 6,032,680 SNPs while retaining 1,138,032 effective SNPs following regression coefficient preservation criteria. Comprehensive genomic inflation assessment revealed: mean χ² = 0.523, genomic control lambda (λGC) = 0.908, maximum χ² = 913.763, genome-wide significant hits = 7, heterogeneity testing = passed ( P > 0.05), observed-scale heritability (h 2 ) = 0.0139 (SE = 0.0021), genetic-environmental contribution ratio < 0, regression intercept = 0.4979, and regression intercept standard error = 0.0025. These multiple estimation parameters collectively demonstrate that potential inflation in our structural equation framework resulted from polygenic heritability signals rather than population stratification bias or pleiotropy parameter effects. FUMA-based Neuroimmune Syndrome Structural Equation Model Assessment FUMA software evaluation of our genomic structural equation model identified 36 risk gene loci (Supplementary Table 5e, Figure 3), with 2 potential neuroimmune syndrome-associated genes achieving genome-wide significance control (significance threshold = 5×10 -8 , FDR < 0.05)(Figure 4).Through FUMA annotation, we mapped 34 lead SNP loci, with the majority located in intergenic regions (Supplementary Table 6). Our novel GWAS subtraction analysis did not identify genetic variants significantly associated with neuroimmune syndromes beyond those detected through standard approaches.No GWAS subtraction loci were identified (Supplementary Table 7). Fine-mapping Analysis Fine-mapping analysis identified high-confidence causal variants (mean posterior probability > 0.95) across four major association regions: chromosome 1 association cluster (encompassing 6 variants including rs56332939 and rs76692181), chromosome 4 association cluster (7 variants including rs10029041 and rs7664257), chromosome 8 association cluster (4 variants including rs35642230 and rs72678550), and chromosome 11 association cluster (4 variants including rs34851197 and rs71482185). Regional association plots demonstrated pronounced association peaks at these loci (Figure 5 and Supplementary Table 8). Transcriptomic Prediction Analysis Subsequently, we conducted transcriptome-wide association study (TWAS) using FUSION to identify gene-level associations with neuroimmune syndromes. Following multiple comparison correction, no genes achieved significance thresholds for association (Extended Data Table 1). However, subsequent FOCUS fine-mapping analysis of our genomic structural equation data identified 78 genes potentially representing pathogenic signals for neuroimmune syndromes. To further characterize these high-confidence gene-level associations, we performed intersection analysis. Seven genes (ISCA2, NPC2, PRSS8, ENDOD1, PNKD, GSAP, and ATG101) exhibited positive TWAS Z-scores, indicating that predicted gene expression positively correlates with neuroimmune syndrome risk, suggesting that upregulation of these genes may associate with increased neuroimmune syndrome susceptibility. Conversely, eight genes (PGBD1, DNAJC24, LINC02980, TUBGCP5, LSM12P1, ATF6B, PFN1P2, and MDC1) showed negative TWAS Z-scores, indicating that their downregulation associates with increased neuroimmune syndrome risk (TWAS and FOCUS intersection results in Supplementary Table 9). Pathway, Cell Type, and Mendelian Disease Gene Enrichment Multi-marker genomic annotation analysis (MAGMA) identified 2 genes through genomic mapping (Supplementary Table 10). Gene set analysis utilizing these genes demonstrated enrichment in Gene Set Enrichment Analysis (GSEA) categories (Supplementary Table 11,Figure 6). Additionally, biological processes mapped through gene enrichment were validated in Gene Ontology (GO) terms, including "regulation of mast cell cytokine production" and "positive regulation of mast cell cytokine production." Disease enrichment analysis revealed multiple neuroimmune-related disorders exceeding significance standards following multiple comparison correction (Supplementary Table 12). The most significant disease categories included ectodermal dysplasia 10A and neurodegenerative disease. Neuroimmune syndrome enrichment proved most significant in immune regulatory processes, with identified biological processes extensively involving mast cell and T cell regulation (regulation of mast cell cytokine production, regulation of T cell differentiation). Nervous system disease enrichment was similarly significant (including neurodegenerative disease, motor neuron disease, and nervous system disease; P < 0.05). EEG abnormality phenotypes (EEG with polyspike wave complexes) provided additional support for neurophysiological alterations. Cell type enrichment analysis revealed that while no cell types achieved significance following multiple comparison correction, 19 cell types demonstrated enrichment at nominal significance levels ( P < 0.05). The two most significant cell types were kidney macrophages ( P = 0.0023) and tracheal blood cells ( P = 0.0025). Neuroimmune syndrome enrichment patterns in immune cells were pronounced, with 18 of 19 nominally significant cell types representing immune cell populations (including various macrophage, B cell, myeloid cell, and natural killer cell subtypes), underscoring the critical role of immune systems in disease mechanisms. Genomic Regional Heritability Contribution Results Genomic regional heritability contribution analysis identified 25 genomic functional regions achieving significance following multiple comparison correction (FDR < 0.05) (Supplementary Table 13). Genetic contributions concentrated primarily in chromosomal regulatory regions, including transcription start sites (TSS), promoter regions, enhancers, and super-enhancers as key regulatory elements. The most significant regions included transcription start sites (TSS_Hoffman; enrichment = 21.77, FDR = 0.0007) and H3K4me3-modified transcriptionally active regions (enrichment = 13.30, FDR = 0.016). Histone modification-associated regions demonstrated significant enrichment, particularly H3K27ac, H3K4me1, and H3K4me3 modification regions representing important markers of transcriptional activity and enhancer function. Promoter regions (Promoter_UCSC) and various enhancer regions (including super-enhancers SuperEnhancer_Hnisz) all displayed significant heritability enrichment, indicating central roles of gene expression regulatory mechanisms in disease susceptibility. These findings suggest that genetic variants primarily exert effects through influencing gene transcriptional regulatory networks rather than altering protein-coding sequences. Multi-Platform Drug Target and Biomarker Discovery Analysis of 4,907 plasma proteins from the deCODE genetics consortium identified eight significant associations ( P < 0.05). BDH2 showed the strongest risk association (OR = 1.14, 95% CI: 1.02-1.26, P = 0.018), followed by WFDC2 (OR = 1.13, 95% CI: 1.00-1.26, P = 0.047), ALPL (OR = 1.07, 95% CI: 1.00-1.14, P = 0.048), LILRA2 (OR = 1.04, 95% CI: 1.00-1.09, P = 0.046), PCDHA4 (OR = 1.04, 95% CI: 1.00-1.07, P = 0.026), GRID2 (OR = 1.03, 95% CI: 1.01-1.06, P = 0.015), and RNASE6 (OR = 1.03, 95% CI: 1.00-1.07, P = 0.039). EPB41 was the only protective protein (OR = 0.85, 95% CI: 0.73-1.00, P = 0.046)(Supplementary Table 14, Fig.7). Analysis of candidate therapeutic targets identified five significant associations. Protective effects were observed for CTLA4 (OR = 0.91, 95% CI: 0.86-0.96, P = 0.0005), TYK2 (OR = 0.97, 95% CI: 0.96-0.99, P = 0.008), and CCR5 (OR = 0.97, 95% CI: 0.95-0.99, P = 0.003). Risk effects were found for LILRA2 (OR = 1.01, 95% CI: 1.00-1.02, P = 0.031) and BDH2 (OR = 1.01, 95% CI: 1.00-1.02, P = 0.0075)(Supplementary Table 15,Fig.8). Analysis using Yang et al. brain, cerebrospinal fluid, and plasma pQTL data identified 529 proteins with significant associations (FDR < 0.05). Key findings included complement components (C2, C5, C9), inflammatory mediators (TNF, IL1B), glial activation markers (GFAP), neuroprotective factors (GDNF, NGF), immune checkpoint molecules (CD274), and vascular markers (VCAM1). EPB41 showed consistent associations across both deCODE and Yang et al. Analyses(Supplementary Table 16,Figure 9). Analysis of 50,033 phenotypes identified 17 traits with significant causal associations (FDR < 0.05). The strongest association was thyroid problems (not cancer) (OR = 1.826, 95% CI: 1.494-2.232, P = 4.11×10 -9 , FDR = 9.43×10 -5 ), followed by self-reported hypothyroidism (OR = 1.743, 95% CI: 1.393-2.180, P = 1.14×10 -6 ) and rheumatoid arthritis (OR = 1.035, 95% CI: 1.022-1.047, P = 3.24×10 -8 ). Sixteen traits showed risk-increasing effects, supporting shared autoimmune and endocrine mechanisms(Supplementary Table 17, Figure 10). The analyses identified novel metabolic targets (BDH2), established drug targets (CTLA4, TYK2, CCR5), validated inflammatory pathways (complement, TNF), neuroprotective networks (GDNF, NGF), and clinical disease associations (thyroid disorders, autoimmune diseases). EPB41 demonstrated cross-platform consistency, while BDH2 and LILRA2 showed convergent evidence across multiple approaches. Chromosomal-level Results Polygenic risk score analysis using 1,077,341 effective variants revealed a highly non-uniform distribution of genetic contributions across chromosomes, with a dominant two-chromosome architecture. Chromosome 4 demonstrated the highest genetic contribution (39.1%), followed by chromosome 14 (23.8%), together accounting for 62.9% of the total genetic variance in neuroimmune syndrome susceptibility. Beyond the two dominant chromosomes, genetic contributions were substantially lower, with chromosomes 1 (3.5%) and 2 (3.2%) showing modest contributions above the 5% threshold. Notably, chromosome 6, containing the HLA region, contributed 2.4%, indicating that while HLA variants play a role, the genetic architecture is predominantly non-HLA driven. All remaining chromosomes contributed less than 3% each to the overall genetic risk. Correlation analysis between variant count and genetic contribution revealed a weak linear relationship (Pearson r = 0.182, P = 0.418) but a strong rank correlation (Spearman ρ = 0.874, P = 2.14×10 -6 ), demonstrating that effect size rather than variant quantity drives chromosomal contributions. This suggests the presence of high-impact variants concentrated on chromosomes 4 and 14, warranting detailed investigation for major effect loci and therapeutic target identification(Figure S11). Discussion This investigation represents the first comprehensive genetic architecture mapping of neuroimmune syndromes through genomic structural equation modeling of latent, unmeasured phenotypes. By integrating five neuroimmune-related disorders—Guillain-Barré syndrome, myasthenia gravis, multiple sclerosis, systemic lupus erythematosus, and systemic connective tissue diseases—we established a novel GWAS framework that identified 33 genome-wide significant loci, 78 candidate pathogenic genes, and critical therapeutic targets. Our findings reveal that neuroimmune syndrome susceptibility operates through complex multi-system mechanisms spanning metabolic dysregulation, immune checkpoint dysfunction, and transcriptional regulatory networks, providing unprecedented insights for precision medicine strategies[31,32,33]. Our genomic structural equation modeling analysis demonstrated substantial genetic covariances among the five constituent phenotypes, with the strongest correlations observed between systemic lupus erythematosus and systemic connective tissue diseases , myasthenia gravis and systemic connective tissue diseases , and Guillain-Barré syndrome and myasthenia gravis . These findings provide quantitative evidence for shared genetic susceptibility foundations in autoimmune diseases[34,35,36], supporting the hypothesis that common immune dysregulation mechanisms underlie diverse neuroimmune manifestations. Multiple sclerosis and systemic connective tissue diseases exhibited relatively higher heritability estimates within this disease spectrum, suggesting more pronounced genetic contributions to these conditions. The excellent model fit confirms that these diseases represent interconnected spectra rather than isolated entities, operating through shared genetic mechanisms that transcend traditional diagnostic boundaries[37]. Functional annotation analysis revealed that the 33 identified loci are predominantly enriched in gene regulatory regions rather than protein-coding sequences, with transcription start sites showing the strongest enrichment followed by H3K4me3-modified transcriptionally active regions. Twenty-five genomic functional regions achieved significance after multiple comparison correction, encompassing enhancers, super-enhancers, and histone modification sites (H3K27ac, H3K4me1, H3K4me3). This regulatory enrichment pattern aligns with established frameworks demonstrating that autoimmune disease variants primarily function through epigenetic modifications and chromatin architecture alterations[38,39,40,41]. These findings support the omnigenic model[42], where disease susceptibility emerges from coordinated dysregulation of transcriptional networks rather than individual gene effects, indicating that therapeutic interventions targeting regulatory mechanisms may prove more effective than approaches focused on single protein targets[43]. Fine-mapping analysis identified 21 high-confidence causal variants distributed across four chromosomal clusters (chromosomes 1, 4, 8, and 11), providing precise molecular targets for functional validation. Transcriptomic analysis through FOCUS methodology revealed 78 candidate pathogenic genes exhibiting complex bidirectional regulatory patterns—upregulation of genes including ISCA2, NPC2, PRSS8, ENDOD1, PNKD, GSAP, and ATG101, and downregulation of genes including PGBD1, DNAJC24, LINC02980, TUBGCP5, LSM12P1, ATF6B, PFN1P2, and MDC1 both associated with increased disease risk[44]. This bidirectional regulation reflects the intricate homeostatic balance required for neuroimmune system stability, where both excessive activation and insufficient suppression contribute to pathogenesis[45]. Pathway enrichment analysis revealed significant involvement of mast cell cytokine production regulation and T cell differentiation processes, connecting our genetic discoveries to established immunological mechanisms in neuroinflammation[46,47,48]. Polygenic risk score analysis using PRS-CS methodology with 1,077,341 effective variants revealed a remarkably concentrated genetic architecture dominated by two chromosomes. Chromosome 4 contributed 39.1% of total genetic variance, followed by chromosome 14 (23.8%), together accounting for 62.9% of neuroimmune syndrome susceptibility. This highly non-uniform distribution contrasts sharply with traditional polygenic models and suggests the presence of major effect loci warranting intensive investigation[49,50]. Importantly, chromosome 6 containing the HLA region contributed only 2.4%, indicating that neuroimmune syndrome genetic susceptibility operates predominantly through non-HLA mechanisms—a finding that challenges conventional autoimmune disease models emphasizing HLA dominance. Correlation analysis demonstrated that effect size rather than variant quantity drives chromosomal contributions, suggesting high-impact variants concentrated on chromosomes 4 and 14 represent priority targets for therapeutic development and mechanistic investigation[51]. Our systematic multi-platform Mendelian randomization approach identified therapeutic opportunities across multiple biological pathways, providing genetic validation for both novel and established drug targets[52,53,54]. Novel discoveries included BDH2, establishing the first causal connection between ketone body metabolism and neuroimmune diseases, suggesting that metabolic interventions targeting mitochondrial function may provide therapeutic benefits. Established drug target validation identified three proteins with significant protective associations: CTLA4, TYK2, and CCR5, providing genetic evidence for immune checkpoint inhibition, JAK inhibition, and chemokine receptor antagonism as therapeutic strategies[55,56]. Cross-platform validation through Yang et al. multi-tissue analysis of 529 significant proteins (FDR < 0.05) extensively confirmed our findings, revealing complement system activation (C2, C5, C9)[27], neuroinflammation markers (TNF, GFAP)[58,59], and neuroprotective factors (GDNF, NGF)[60], establishing robust biological validation for our analytical framework. Large-scale phenome-wide association analysis of 50,033 traits identified 17 phenotypes with significant causal associations[61,62], significantly expanding our understanding of neuroimmune syndrome risk factors. Thyroid disorders emerged as the strongest risk factor, representing nearly two-fold increased disease risk, followed by self-reported hypothyroidismand rheumatoid arthritis. The predominantly risk-increasing pattern (16 of 17 significant associations) supports shared pathophysiological mechanisms across autoimmune and endocrine systems, suggesting that neuroimmune syndromes represent manifestations of broader systemic dysregulation[63]. These validated associations provide crucial insights for clinical risk stratification and therapeutic target identification, establishing robust causal evidence for precision medicine strategies[64]. The convergence of our multi-tiered findings establishes a comprehensive framework for understanding neuroimmune syndrome pathophysiology. The identification of concentrated chromosomal architecture (Chr4/Chr14), regulatory enrichment patterns, metabolic factors (ketone body metabolism), immune checkpoints (CTLA4), signaling pathways (TYK2, JAK), and chemokine networks (CCR5) alongside validated disease associations (thyroid-autoimmune axis) provides multiple therapeutic entry points spanning immune modulation, metabolic intervention, and transcriptional regulation. This integrated approach reveals that effective neuroimmune syndrome treatment likely requires combination strategies addressing the complex interplay between genetic predisposition, metabolic dysfunction, and immune dysregulation rather than targeting individual pathways in isolation[65,66]. The progression from unbiased discovery through targeted validation to biological confirmation and clinical relevance assessment provides robust evidence for precision medicine strategies, with particular strength in complement-targeted therapies, immune checkpoint modulation, and metabolic interventions[67]. Several limitations require consideration in interpreting these findings. First, our analysis primarily utilized European ancestry populations from FinnGen, limiting generalizability across diverse populations. Future studies should expand to Asian, African, and Native American populations to ensure cross-ancestry validity[68,69,70]. Second, while we identified concentrated genetic architecture on chromosomes 4 and 14, the specific biological mechanisms underlying this concentration require detailed functional investigation through experimental validation. Third, although our transcriptomic analysis revealed 78 candidate genes with bidirectional regulatory patterns, connecting these discoveries to specific molecular pathways necessitates laboratory-based functional studies. Additionally, environmental factors likely interact with genetic susceptibility in neuroimmune syndrome development, necessitating future gene-environment interaction studies to fully understand disease etiology[71]. The complex bidirectional regulatory patterns observed in our candidate genes suggest that therapeutic interventions must carefully balance activation and suppression mechanisms to avoid unintended consequences[72]. Our findings provide several immediate translational opportunities that warrant clinical investigation. The identification of CTLA4, TYK2, and CCR5 as protective targets offers genetic validation for existing therapeutic approaches including immune checkpoint inhibitors, JAK inhibitors, and chemokine receptor antagonists, supporting drug repurposing strategies in neuroimmune diseases. The strong thyroid-neuroimmune association suggests that screening protocols for thyroid dysfunction in patients with neuroimmune manifestations may improve clinical outcomes and enable earlier intervention. The concentrated chromosomal architecture indicates that genetic risk scores incorporating chromosome 4 and 14 variants may provide more accurate risk prediction than traditional genome-wide approaches, potentially enabling personalized prevention strategies. Finally, the regulatory enrichment patterns suggest that future therapeutic development should prioritize epigenetic modulators and transcriptional interventions rather than traditional protein-targeting approaches, potentially opening new avenues for precision medicine in neuroimmune diseases. The metabolic component revealed through BDH2 associations suggests that nutritional interventions targeting ketone body metabolism warrant investigation as adjunctive therapies in neuroimmune syndrome management. Conclusions This investigation represents the first comprehensive genetic architecture mapping of neuroimmune syndromes through genomic structural equation modeling, revealing 33 genome-wide significant loci and 78 candidate pathogenic genes with shared genetic susceptibility across five neuroimmune-related disorders. Our findings demonstrate a remarkably concentrated genetic architecture dominated by chromosomes 4 and 14, which together account for 62.9% of total genetic variance, challenging traditional polygenic models and indicating non-HLA-driven mechanisms as primary contributors to disease susceptibility. The identification of validated therapeutic targets including CTLA4, TYK2, and CCR5, along with novel metabolic pathways involving BDH2, provides genetic evidence for precision medicine strategies encompassing immune checkpoint modulation, JAK inhibition, and metabolic interventions. The strong causal association with thyroid disorders establishes important clinical risk stratification opportunities, while the regulatory enrichment patterns suggest that epigenetic modulators may prove more effective than traditional protein-targeting approaches. Future research should expand to diverse populations, conduct functional validation of the concentrated chromosomal architecture, and investigate gene-environment interactions to fully understand neuroimmune syndrome etiology and optimize personalized therapeutic interventions. Declarations Data Availability The data that support the findings of this study are available in FinnGen Research Program at https://www.finngen.fi/en, reference number Release 12. These data were derived from the following resources available in the public domain: - FinnGen GWAS summary statistics for neuroimmune disorders, https://www.finngen.fi/en/access_results - eQTLGen Consortium, https://www.eqtlgen.org/ - IEU OpenGWAS database, https://gwas.mrcieu.ac.uk/ - Yang et al. brain, CSF, and plasma pQTL, https://doi.org/10.1038/s41593-021-00886-6 - deCODE genetics consortium, https://doi.org/10.1038/s41588-021-00978-w. The neuroimmune syndrome structural equation modeling summary statistics generated in this study are available from the corresponding author upon reasonable request. Conflicts of Interest The authors declare that there are no conflicts of interest regarding the publication of this article. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Acknowledgments The authors thank the FinnGen Research Program and all participants who contributed to the FinnGen study. We acknowledge the eQTLGen Consortium for providing eQTL data, the deCODE genetics consortium for plasma protein data, and the IEU OpenGWAS database for phenome-wide association data. We also thank the research teams who generated and made publicly available the GWAS summary statistics used in this study. The authors appreciate the technical support and computational resources that made this analysis possible. Declaration of Contributions and AI Usage: All authors contributed significantly to this work and agree to be accountable for all aspects of the research content and conclusions. No third-party services or individuals not listed as authors were involved in the research or manuscript preparation. No artificial intelligence software was used in any aspect of manuscript preparation, including data collection, analysis, writing, or editing. Author Contributions Haifeng Chen¹† and Yuxiong Liao¹† contributed equally to this work as co-first authors and were responsible for conceptualization, methodology, formal analysis, investigation, data curation, and writing the original draft. Luejun Tang¹ contributed to data curation, formal analysis, and validation. Xiaoyun Wei¹ contributed to methodology, software implementation, and validation. Tongshun Li¹ provided resources, assisted with data curation, and validation. Wei Chen¹* (corresponding author) was responsible for conceptualization, methodology, supervision, project administration, and final manuscript approval. All authors participated in writing review and editing and agree to be accountable for all aspects of the work content and conclusions. Ethical Approval This study utilized publicly available genome-wide association study (GWAS) summary statistics from the FinnGen Research Program Release 12 and other established genomic consortia. All constituent GWAS studies included in this analysis had received institutional review board approval from their respective institutions, with informed consent obtained from all participants. As this investigation employed only summary-level genetic data without access to individual participant information, additional ethical approval was not required for this secondary analysis. Funding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Availability of data and materials The data that support the findings of this study are available from publicly accessible genomic databases. FinnGen GWAS summary statistics are available at https://www.finngen.fi/en/access_results (Release 12). Additional datasets used include: eQTLGen Consortium data (https://www.eqtlgen.org/), IEU OpenGWAS database (https://gwas.mrcieu.ac.uk/), Yang et al. brain, cerebrospinal fluid, and plasma protein quantitative trait loci data (https://doi.org/10.1038/s41593-021-00886-6), and deCODE genetics consortium plasma protein data (https://doi.org/10.1038/s41588-021-00978-w). The neuroimmune syndrome structural equation modeling summary statistics generated in this study are available from the corresponding author upon reasonable request. References Engelhardt B, Vajkoczy P, Weller RO. The movers and shapers in immune privilege of the CNS. Nat Immunol. 2017 Feb;18(2):123-131. Prinz M, Jung S, Priller J. Microglia biology: one century of evolving concepts. Cell. 2019;179(2):292-311. Kipnis J. Multifaceted interactions between adaptive immunity and the central nervous system. Science. 2016;353(6301):766-771. Absinta M, Lassmann H, Trapp BD. Mechanisms underlying progression in multiple sclerosis. Nat Rev Neurol. 2020;16(11):657‑668. Kuchroo VK, Ohashi PS, Sartor RB, Vinuesa CG. Dysregulation of immune homeostasis in autoimmune diseases. Nat Med. 2012;18(1):42‑47. Bach JF. The hygiene hypothesis in autoimmunity: the role of pathogens and commensals. Nat Rev Immunol. 2018;18(2):105‑120. Kim A, Xie F, Abed OA, Moon JJ. Vaccines for immune tolerance against autoimmune disease. Adv Drug Deliv Rev. 2023 Dec;203:115-140. Reich DS, Lucchinetti CF, Calabresi PA. Multiple sclerosis. N Engl J Med. 2018;378(2):169‑180. Goverman JM. Autoimmune T cell responses in the central nervous system. Nat Rev Immunol. 2021;21(9):560‑573. Ashley EA. Towards precision medicine. Nat Rev Genet. 2016;17(9):507-522. Grotzinger AD, Rhemtulla M, de Vlaming R, et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. Nat Hum Behav. 2019;3(5):513-525. Kurki MI, Karjalainen J, Palta P, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613(7944):508-518. Tian C, Gregersen PK, Seldin MF. Accounting for ancestry: population substructure and genome-wide association studies. Hum Mol Genet. 2008;17(R2):R143-R150. Turley P, Walters RK, Maghzian O, et al. Multi-trait analysis of genome-wide association summary statistics using MTAG. Nat Genet. 2018;50(2):229-237. Bulik-Sullivan B, Finucane HK, Anttila V, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet. 2015;47(11):1236-1241. Watanabe K, Taskesen E, van Bochoven A, Posthuma D. Functional mapping and annotation of genetic associations with FUMA. Nat Commun. 2017;8(1):1826. Sollis E, Mosaku A, Abid A, et al. The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. Nucleic Acids Res. 2023;51(D1):D977-D985. de Leeuw CA, Mooij JM, Heskes T, Posthuma D. MAGMA: generalized gene-set analysis of GWAS data. PLoS Comput Biol. 2015;11(4):e1004219. Gusev A, Ko A, Shi H, et al. Integrative approaches for large-scale transcriptome-wide association studies. Nat Genet. 2016;48(3):245-252. GTEx Consortium. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. 2020 Sep 11;369(6509):1318-1330. Mancuso N, Freund MK, Johnson R, et al. Probabilistic fine-mapping of transcriptome-wide association studies. Nat Genet. 2019;51(4):675-682. Liberzon A, Birger C, Thorvaldsdóttir H, Ghandi M, Mesirov JP, Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015;1(6):417-425. Tabula Muris Consortium, Overall coordination, Logistical coordination.Single-cell transcriptomics of 20 mouse organs creates a Tabula Muris. Nature. 2018;562(7727):367-372. Finucane HK, Bulik-Sullivan B, Gusev A, et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat Genet. 2015;47(11):1228-1235. Võsa U, Claringbould A, Westra HJ, et al. Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet. 2021;53(9):1300-1310. Ferkingstad E, Sulem P, Atlason BA, et al. Large-scale integration of the plasma proteome with genetics and disease. Nat Genet. 2021;53(12):1712-1721. Yang C, Farias FHG, Ibanez L, et al. Genomic atlas of the proteome from brain, CSF and plasma prioritizes proteins implicated in neurological disorders. Nat Neurosci. 2021;24(9):1302-1312. Hemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. eLife. 2018;7:e34408. Lambert SA, Gil L, Jupp S, et al. The Polygenic Score Catalog as an open database for reproducibility and systematic evaluation. Nat Genet. 2021;53(4):420-425. Ge T, Chen CY, Ni Y, Feng YCA, Smoller JW. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nat Commun. 2019;10(1):1776. Ransohoff RM, Engelhardt B. The anatomical and cellular basis of immune surveillance in the central nervous system. Nat Rev Immunol. 2012;12(9):623-635. Mucida D, Husain MM, Muroi S, et al. Transcriptional reprogramming of mature CD4+ helper T cells generates distinct MHC class II-restricted cytotoxic T lymphocytes. Nat Immunol. 2013;14(3):281-289. Dendrou CA, Fugger L, Friese MA. Immunopathology of multiple sclerosis. Nat Rev Immunol. 2015;15(9):545-558. Okada Y, Wu D, Trynka G, et al. Genetics of rheumatoid arthritis contributes to biology and drug discovery. Nature. 2014;506(7488):376-381. Ellinghaus D, Jostins L, Spain SL, et al. Analysis of five chronic inflammatory diseases identifies 27 new associations and highlights disease-specific patterns. Nat Genet. 2016;48(5):510-518. Cross-Disorder Group of the Psychiatric Genomics Consortium. Identification of risk loci with shared effects on five major psychiatric disorders. Lancet. 2013;381(9875):1371-1379. Cotsapas C, Voight BF, Rossin E, et al. Pervasive sharing of genetic effects in autoimmune disease. PLoS Genet. 2011;7(8):e1002254. Javierre BM, Burren OS, Wilder SP, et al. Lineage-specific genome organization links enhancers and non-coding disease variants to target gene promoters. Cell. 2016;167(5):1369-1384. Farh KK, Marson A, Zhu J, et al. Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature. 2015;518(7539):337-343. Roadmap Epigenomics Consortium. Integrative analysis of 111 reference human epigenomes. Nature. 2015;518(7539):317-330. Ernst J, Kellis M. ChromHMM: automating chromatin-state discovery and characterization. Nat Methods. 2012;9(3):215-216. Boyle EA, Li YI, Pritchard JK. An expanded view of complex traits: from polygenic to omnigenic. Cell. 2017;169(7):1177-1186. ENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature. 2012;489(7414):57-74. Soskic B, Cano-Gamez E, Smyth DJ, et al. Chromatin activity at GWAS loci identifies T cell states driving complex immune diseases. Nat Genet. 2019;51(10):1486-1493. Lyons JJ, Yu X, Hughes JD, et al. Elevated basal serum tryptase identifies a multisystem disorder associated with increased TPSAB1 copy number. Nat Genet. 2016;48(12):1564-1569. Hormozdiari F, van de Bunt M, Segrè AV, et al. Colocalization of GWAS and eQTL signals detects target genes. Am J Hum Genet. 2016;99(6):1245-1260. Wallace C. Eliciting priors and relaxing the single causal variant assumption in colocalisation analyses. PLoS Genet. 2020;16(4):e1008720. Giambartolomei C, Vukcevic D, Schadt EE, et al. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10(5):e1004383. Choi SW, Mak TS, O'Reilly PF. Tutorial: a guide to performing polygenic risk score analyses. Nat Protoc. 2020;15(9):2759-2772. Mars N, Kerminen S, Feng YCA, et al. Polygenic and clinical risk scores and their impact on age at onset and prediction of cardiometabolic diseases and common cancers. Nat Med. 2020;26(4):549-557. Vilhjálmsson BJ, Yang J, Finucane HK, et al. Modeling linkage disequilibrium increases accuracy of polygenic risk scores. Am J Hum Genet. 2015;97(4):576-592. Schmidt AF, Finan C, Gordillo-Marañón M, et al. Genetic drug target validation using Mendelian randomisation. Nat Commun. 2020;11(1):3255. Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658-665. Bowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304-314. Verbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693-698. Nelson MR, Tipney H, Painter JL, et al. The support of human genetic evidence for approved drug indications. Nat Genet. 2015;47(8):856-860. Ricklin D, Hajishengallis G, Yang K, Lambris JD. Complement: a key system for immune surveillance and homeostasis. Nat Immunol. 2010;11(9):785-797. Kalliolias GD, Ivashkiv LB. TNF biology, pathogenic mechanisms and emerging therapeutic strategies. Nat Rev Rheumatol. 2016;12(1):49-62. Abdelhak A, Huss A, Kassubek J, et al. Serum GFAP as a biomarker for disease severity in multiple sclerosis. Sci Rep. 2018;8(1):14798. Kerschensteiner M, Gallmeier E, Behrens L, et al. Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor. Ann Neurol. 1999;46(1):90-100. Pendergrass SA, Browning BL, Dudek SM, et al. Phenome-wide association study (PheWAS) for detection of pleiotropy within the Population Architecture using Genomics and Epidemiology (PAGE) Network. PLoS Genet. 2013;9(1):e1003087. Denny JC, Ritchie MD, Basford MA, et al. PheWAS: demonstrating the feasibility of a phenome-wide scan to discover gene-disease associations. Bioinformatics. 2010;26(9):1205-1210. Medici M, Visser WE, Visser TJ, Peeters RP. Genetic determination of the hypothalamic-pituitary-thyroid axis: where do we stand? Endocr Rev. 2015;36(2):214-244. Gusev A, Mancuso N, Won H, et al. Transcriptome-wide association study of schizophrenia and chromatin activity yields mechanistic disease insights. Nat Genet. 2018;50(4):538-548. Zhu Z, Zhang F, Hu H, et al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. 2016;48(5):481-487. Hauberg ME, Zhang W, Giambartolomei C, et al. Large-scale identification of common trait and disease variants affecting gene expression. Am J Hum Genet. 2017;101(1):157-173. Wu Y, Zeng J, Zhang F, et al. Integrative analysis of omics summary data reveals putative mechanisms underlying complex traits. Nat Commun. 2018;9(1):918. Popejoy AB, Fullerton SM. Genomics is failing on diversity. Nature. 2016;538(7624):161-164. Martin AR, Kanai M, Kamatani Y, et al. Clinical use of current polygenic risk scores may exacerbate health disparities. Nat Genet. 2019;51(4):584-591. Peterson RE, Kuchenbaecker K, Walters RK, et al. Genome-wide association studies in ancestrally diverse populations: opportunities, methods, pitfalls, and recommendations. Cell. 2019;179(3):589-603. Duncan L, Shen H, Gelaye B, et al. Analysis of polygenic risk score usage and performance in diverse human populations. Nat Commun. 2019;10(1):3328. Hunter DJ. Gene-environment interactions in human diseases. Nat Rev Genet. 2005;6(4):287-298. Tables Table 1. Parameters and Quality Metrics for Genomic Structural Equation Modeling. Phenotype NSNPs h2_se λGC Mean_ChiSquare Intercept_se Ratio_se GBS 1159723 0.0013 (0.0013) 1.0132 1.0055 0.9927 (0.0067) 1.3374 (1.2246) MG 1159735 0.0019 (0.001) 1.0216 1.0157 0.9975 (0.0068) 0.1612 (0.435) MS 1159738 0.0084 (0.0013) 1.089 1.113 1.0287 (0.0084) 0.254 (0.0745) SLE 1159730 0.0011 (0.001) 1.0072 1.0036 0.9927 (0.0077) 2.0275 (2.1232) SCTD 1159740 0.0085 (0.0017) 1.1112 1.1285 1.0385 (0.0095) 0.2994 (0.0737) Notes: h2_se represents SNP-based heritability estimates with standard errors in parentheses. λGC indicates genomic control lambda for population stratification assessment. LDSC intercept values assess confounding due to population stratification and cryptic relatedness. Additional Declarations No competing interests reported. Supplementary Files STable17.txt STable2.txt STable4a.txt STable5a.xlsx ExtendedDataTable1.txt STable4b.txt STable6.txt STable5aleadsnp.xlsx STable5EGenomicRiskLoci.xlsx STable5b.xlsx STable7.txt STable10.txt STable5bleadsnp.xlsx STable12.txt STable13.txt STable11.txt STable15.xlsx STable14.csv STable16.csv SupplementaryFile3.txt.rdata STable8.txt SupplementaryInformation.docx STable9.txt STable17.txt STable5.zip 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-7328766","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":501594137,"identity":"a235205b-f642-4516-8de7-7d38d6eea02d","order_by":0,"name":"Haifeng Chen¹","email":"","orcid":"","institution":"Nanning Hospital of Traditional Chinese Medicine, Affiliated to Guangxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Haifeng","middleName":"","lastName":"Chen¹","suffix":""},{"id":501594138,"identity":"4091d7a7-d979-4f25-9a4b-9a08d6a33c85","order_by":1,"name":"Yuxiong Liao¹","email":"","orcid":"","institution":"Nanning Hospital of Traditional Chinese Medicine, Affiliated to Guangxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Yuxiong","middleName":"","lastName":"Liao¹","suffix":""},{"id":501594139,"identity":"fd4bf1ec-2391-4172-9de1-ba8e9e8cc6d7","order_by":2,"name":"Luejun Tang¹","email":"","orcid":"","institution":"Nanning Hospital of Traditional Chinese Medicine, Affiliated to Guangxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Luejun","middleName":"","lastName":"Tang¹","suffix":""},{"id":501594140,"identity":"99c9f761-8ec1-4441-b318-32719df08c60","order_by":3,"name":"Xiaoyun Wei¹","email":"","orcid":"","institution":"Nanning Hospital of Traditional Chinese Medicine, Affiliated to Guangxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Xiaoyun","middleName":"","lastName":"Wei¹","suffix":""},{"id":501594141,"identity":"5b4faa10-48c4-449c-a2e2-9e2c7389b07b","order_by":4,"name":"Tongshun Li¹","email":"","orcid":"","institution":"Nanning Hospital of Traditional Chinese Medicine, Affiliated to Guangxi University of Chinese Medicine","correspondingAuthor":false,"prefix":"","firstName":"Tongshun","middleName":"","lastName":"Li¹","suffix":""},{"id":501594142,"identity":"d41e0a07-374a-4a20-99a3-9a00c17b2854","order_by":5,"name":"Wei Chen¹","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6klEQVRIiWNgGAWjYDACZiBmbGBgMGBgYGNIqJCQkydRyxkLY8MGYmyCa2Fsq0hkOEBAtcFx5ocPfu6wyTNnP/7swcN5EgmMDcwPH93Ao0Wymc3YsPdMWrFlT465QeI2iTx2BjZj4xw8WviZGcykGdsOJ244kMMmAdRSzNjAwyaNTwsbM/s3oJb/iRvOP38mkThHIrHhAAEt/Mw8IFsOJG64kWAGVE+EFslmnmLD3rbkxJ0z3phJJByTMDZsJuAXg/PHNz742WaXuJ0//Znkj5o6OXn25oeP8WnBAphJUz4KRsEoGAWjAAsAAJKjSTsCAF3ZAAAAAElFTkSuQmCC","orcid":"","institution":"Nanning Hospital of Traditional Chinese Medicine, Affiliated to Guangxi University of Chinese Medicine","correspondingAuthor":true,"prefix":"","firstName":"Wei","middleName":"","lastName":"Chen¹","suffix":""}],"badges":[],"createdAt":"2025-08-08 15:38:56","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7328766/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7328766/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":89664393,"identity":"9113f608-1104-4552-b861-62b30015170e","added_by":"auto","created_at":"2025-08-22 11:40:27","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":3134592,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWorkflow schematic of genomic structural equation modeling and multi-omics integration for neuroimmune syndrome analysis.\u003c/strong\u003e\u003c/p\u003e","description":"","filename":"Fig.1.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/1afdef96d7326c2cd4dddef1.png"},{"id":89664392,"identity":"085e9052-de3c-48f5-a250-2698d879e0ff","added_by":"auto","created_at":"2025-08-22 11:40:27","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1363169,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eGenetic correlations and structural equation modeling of neuroimmune syndrome.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNote：\u003c/strong\u003e(A) Genetic correlation matrix between five neuroimmune diseases showing shared genetic architecture. (B) Genomic structural equation model with neuroimmune syndrome (NIS) as a latent factor. Factor loadings (standard errors): SCTD 0.84 (0.01), SLE 0.67 (0.05), MG 0.61 (0.03), GBS 0.39 (0.13), MS 0.23 (0.11). Residual variances (μ) represent disease-specific components. Model fit: CFI = 1.0, SRMR = 0.095.\u003c/p\u003e","description":"","filename":"Fig.2.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/34be9607712a7ad76b675a61.png"},{"id":89663575,"identity":"0306fb70-6d97-44e1-bca4-6ac088bf7a4f","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1727664,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRegional association plots of genome-wide significant neuroimmune syndrome GWAS loci.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNote：\u003c/strong\u003eLocusZoom plots for four top lead SNPs with surrounding variants color-coded by linkage disequilibrium (r²): red (0.8-1.0), orange (0.6-0.8), green (0.4-0.6), cyan (0.2-0.4), blue (0.0-0.2). (A) Chromosome 4: rs564762 near CLOCK gene . (B) Chromosome 8: rs9297354 . (C) Chromosome 1: rs55710223 in gene-rich region . (D) Chromosome 12: rs73191114 near ALDH2 gene . Blue lines represent recombination rates.\u003c/p\u003e","description":"","filename":"Fig.3.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/5c88e27c85aeab187a2576af.png"},{"id":89664395,"identity":"6e150c82-1ac1-4471-ab0c-694b4549b392","added_by":"auto","created_at":"2025-08-22 11:40:27","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":847840,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eManhattan plots of GWAS and gene-based association analyses for neuroimmune syndrome.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNotes: \u003c/strong\u003e(A) GWAS Manhattan plot displaying genome-wide association results from the neuroimmune syndrome structural equation modeling. (B) Gene-based association Manhattan plot showing MAGMA analysis results. The x-axis represents chromosomal positions, and the y-axis shows -log10(P-values). The red dashed horizontal line indicates the genome-wide significance threshold (P \u0026lt; 5×10^-8). Genes approaching significance are labeled, including BCL10, AHDC1, SLC9A1, POLL, PLA2G7, ACACB, RNASE6, UNG, DBNDD1, and TRIM4. Alternating colors (blue and gray) distinguish adjacent chromosomes. Gene-based analysis was performed using MAGMA as implemented in FUMA software.\u003c/p\u003e","description":"","filename":"Fig.4.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/0e0b729e78f1c2fbe37d2d56.png"},{"id":89663576,"identity":"8ac4035a-f984-4b86-88af-e1dbf5f5122c","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1397496,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eRegional association plots of genome-wide significant neuroimmune syndrome GWAS loci.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNotes: \u003c/strong\u003eLocusZoom plots for four genome-wide significant loci showing regional association patterns. (A) SLC9A1 locus; (B) SPCS1 locus; (C) MTHFR locus; (D) RP11-610P22.1 locus. In each panel, the x-axis represents chromosomal position, and the y-axis shows -log10(P-values). SNPs are color-coded by linkage disequilibrium (r²) with the lead SNP (purple diamond): red (0.8-1.0), orange (0.6-0.8), green (0.4-0.6), cyan (0.2-0.4), blue (0.0-0.2). Gene structures are displayed above each plot, and recombination rates are shown in blue below. These represent the most significant association signals identified in the neuroimmune syndrome GWAS.\u003c/p\u003e","description":"","filename":"Fig.5.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/ab4fdf0c284fd84a7798160d.png"},{"id":89663579,"identity":"245ad066-7905-44cb-984e-e9dcff31453d","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":3941889,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eTissue-specific expression and functional enrichment analysis of MAGMA-identified genes.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u003c/strong\u003e(A) Heatmap showing tissue-specific expression levels of BCL10 and UNG genes identified through MAGMA analysis across multiple human tissues. Expression values are color-coded from blue (low, 0) to red (high, 5.61). (B) Mendelian disease gene enrichment analysis showing associations with ectodermal dysplasia and neurodegenerative diseases. (C) Gene Ontology biological process enrichment, highlighting significant enrichment in \"regulation of mast cell cytokine production\" and \"positive regulation of mast cell cytokine production\" (empirical \u003cem\u003ep\u003c/em\u003e \u0026lt; 0.05). (D) Cellular component and molecular function enrichment analysis. (E) Human phenotype ontology enrichment showing anatomical and clinical phenotype associations. Each dot represents a gene set, with x-axis indicating the number of overlapping genes, dot size representing the ratio of gene overlap, and color indicating empirical p-values. These results validate the immune-related mechanisms underlying neuroimmune syndrome pathogenesis.\u003c/p\u003e","description":"","filename":"Fig.6.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/a88c3737ae901362b2e02c79.png"},{"id":89664397,"identity":"009764cc-6c8e-4848-ae7e-ecfb56480a68","added_by":"auto","created_at":"2025-08-22 11:40:27","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":621603,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMendelian randomization analysis of protein exposures and neuroimmune syndrome risk.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u003c/strong\u003eForest plot showing causal associations between circulating protein levels and neuroimmune syndrome risk using inverse variance weighted method. Analysis of 4,907 plasma proteins from the deCODE genetics consortium identified eight significant associations (\u003cem\u003ep \u003c/em\u003e\u0026lt; 0.05). Seven proteins showed risk effects with BDH2 demonstrating the strongest association (OR=1.14), while EPB41 was the only protective protein (OR=0.85). The x-axis represents odds ratios with 95% confidence intervals, and the red dashed line indicates null effect (OR=1).\u003c/p\u003e","description":"","filename":"Fig.7.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/4440f1bcfa913e28d8576f30.png"},{"id":89663583,"identity":"b87212da-0f1c-41cc-92cb-5502955f09d3","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":1826506,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSensitivity analysis of Mendelian randomization for candidate therapeutic targets.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u003c/strong\u003eForest plot showing sensitivity analysis of causal associations between five candidate therapeutic target proteins and neuroimmune syndrome risk using multiple Mendelian randomization methods. Each protein (CTLA4, CCR5, TYK2, LILRA2, BDH2) was analyzed using six different methods: MR Egger, Weighted median, Inverse variance weighted (IVW), Simple mode, Weighted mode, and MRConMix. Analysis identified five significant associations with consistent results across methods. Protective effects were observed for CTLA4 (OR=0.91, 95% CI: 0.86-0.96, \u003cem\u003eP\u003c/em\u003e=0.0005), TYK2 (OR=0.97, 95% CI: 0.96-0.99, \u003cem\u003eP\u003c/em\u003e=0.008), and CCR5 (OR=0.97, 95% CI: 0.95-0.99, \u003cem\u003eP\u003c/em\u003e=0.003), while risk effects were found for LILRA2 (OR=1.01, 95% CI: 1.00-1.02, P=0.031) and BDH2 (OR=1.01, 95% CI: 1.00-1.02, \u003cem\u003eP\u003c/em\u003e=0.0075). The x-axis represents odds ratios with 95% confidence intervals, with the red dashed line indicating null effect (OR=1). Results shown are from the inverse variance weighted method, with additional methods providing sensitivity analysis to strengthen causal inference.\u003c/p\u003e","description":"","filename":"Fig.8.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/474ee9958c71e8a7d12d8289.png"},{"id":89663594,"identity":"040da12c-fe16-447f-9a2a-6d9750a49616","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":6246374,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMendelian randomization analysis using Yang et al. brain, cerebrospinal fluid, and plasma pQTL data.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u003c/strong\u003eForest plot or heatmap showing causal associations between protein levels and neuroimmune syndrome risk using Yang et al. brain, cerebrospinal fluid, and plasma protein quantitative trait loci (pQTL) data. Analysis identified 529 proteins with significant associations (FDR \u0026lt; 0.05). Key protein categories include: complement system components (C2, C5, C9), inflammatory mediators (TNF, IL1B), glial activation markers (GFAP), neuroprotective factors (GDNF, NGF), immune checkpoint molecules (CD274), and vascular dysfunction markers (VCAM1). EPB41 demonstrated consistent protective associations across both deCODE genetics consortium (Figure 7) and Yang et al. analyses, strengthening evidence for its potential therapeutic relevance. The analysis provides tissue-specific insights into protein-mediated mechanisms underlying neuroimmune syndrome pathogenesis, particularly highlighting brain and cerebrospinal fluid protein involvement.\u003c/p\u003e","description":"","filename":"Fig.9.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/63d1fab431c52c51e16da52f.png"},{"id":89663580,"identity":"28a51b8a-c929-4978-b17f-aaa66e7fd580","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":5291920,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMendelian randomization analysis of phenome-wide associations with neuroimmune syndrome risk.\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u003c/strong\u003eForest plot showing causal associations between various phenotypes and neuroimmune syndrome risk using Mendelian randomization analysis of 50,033 phenotypes. Analysis identified 17 traits with significant causal associations (FDR \u0026lt; 0.05). The x-axis represents odds ratios (OR) with 95% confidence intervals, and the red dashed line indicates null effect (OR=1). The strongest association was observed for thyroid problems (not cancer) (OR=1.826, 95% CI: 1.494-2.232, P=4.11×10⁻⁹, FDR=9.43×10⁻⁵), followed by self-reported hypothyroidism (OR=1.743, 95% CI: 1.393-2.180,\u003cem\u003e P\u003c/em\u003e=1.14×10⁻⁶) and rheumatoid arthritis (OR=1.035, 95% CI: 1.022-1.047, \u003cem\u003eP\u003c/em\u003e=3.24×10⁻⁸). Sixteen of the 17 traits showed risk-increasing effects (OR \u0026gt; 1), supporting shared pathophysiological mechanisms involving autoimmune dysfunction and endocrine dysregulation in neuroimmune syndrome pathogenesis. The results provide evidence for systemic immune and metabolic perturbations underlying neuroimmune conditions.\u003c/p\u003e","description":"","filename":"Fig.10.png","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/ab47d405224f0ae43c76331e.png"},{"id":90749999,"identity":"052842d4-2f68-4ad7-99e4-642c7bc7c681","added_by":"auto","created_at":"2025-09-07 09:46:57","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":23017615,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/5cbfc444-26ec-4d93-bdf4-66377b0cba55.pdf"},{"id":89663599,"identity":"667c29d3-a4f6-4d77-8191-1c1f2d13f058","added_by":"auto","created_at":"2025-08-22 11:32:31","extension":"txt","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":93260130,"visible":true,"origin":"","legend":"","description":"","filename":"STable17.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/ab4dfed0163a1e423bf7cff1.txt"},{"id":89663566,"identity":"60409eb2-5798-4314-a1ad-928ccdcaa1c1","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":439,"visible":true,"origin":"","legend":"","description":"","filename":"STable2.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/13961f56f0b292137291b014.txt"},{"id":89663567,"identity":"d0876fd8-f2e5-4886-80a2-bab8fb4c8680","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":107,"visible":true,"origin":"","legend":"","description":"","filename":"STable4a.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/7400595dfccb6d20c4150446.txt"},{"id":89664685,"identity":"5c5239c7-3b74-4e6b-b30d-f9e25a5b3011","added_by":"auto","created_at":"2025-08-22 11:48:27","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":14057,"visible":true,"origin":"","legend":"","description":"","filename":"STable5a.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/6596d3d4f8301bbbc26eb47e.xlsx"},{"id":89663572,"identity":"c12fa716-d087-4bf4-9a59-ad114d3978cb","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":198,"visible":true,"origin":"","legend":"","description":"","filename":"ExtendedDataTable1.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/d70a862ebaa2b875cd541d7b.txt"},{"id":89663569,"identity":"ec92f99e-503d-40df-891c-6fb73371b26d","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":1126,"visible":true,"origin":"","legend":"","description":"","filename":"STable4b.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/58ece6f59c7dbfef0394b20d.txt"},{"id":89664684,"identity":"0d28ed8f-442b-4dc0-b24b-b4b04d5a97c4","added_by":"auto","created_at":"2025-08-22 11:48:27","extension":"txt","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":6013,"visible":true,"origin":"","legend":"","description":"","filename":"STable6.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/287cae97db6b9ab95d3fd62a.txt"},{"id":89663584,"identity":"ef4bfb95-0d14-42ea-b546-33f546ae5fa0","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"xlsx","order_by":8,"title":"","display":"","copyAsset":false,"role":"supplement","size":12295,"visible":true,"origin":"","legend":"","description":"","filename":"STable5aleadsnp.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/3026bae41db5c20a7fff86b0.xlsx"},{"id":89663587,"identity":"4d1b3c04-fd68-493a-a3b2-6fb9bed6886a","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"xlsx","order_by":9,"title":"","display":"","copyAsset":false,"role":"supplement","size":13493,"visible":true,"origin":"","legend":"","description":"","filename":"STable5EGenomicRiskLoci.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/d4b994e2536e2ee6c9f672e8.xlsx"},{"id":89663581,"identity":"742d8e36-76c6-4a5e-be14-2264a28207f8","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"xlsx","order_by":10,"title":"","display":"","copyAsset":false,"role":"supplement","size":13889,"visible":true,"origin":"","legend":"","description":"","filename":"STable5b.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/dac033175534c468f3d51fe7.xlsx"},{"id":89663592,"identity":"4017706b-e2c0-4c48-b95f-4ce8ac9e15f9","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":11,"title":"","display":"","copyAsset":false,"role":"supplement","size":16942,"visible":true,"origin":"","legend":"","description":"","filename":"STable7.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/69339674bec2ac485d99b80f.txt"},{"id":89663586,"identity":"d58ce8d4-6c93-4f62-8347-22ba015569a8","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":12,"title":"","display":"","copyAsset":false,"role":"supplement","size":194,"visible":true,"origin":"","legend":"","description":"","filename":"STable10.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/a573cb9cddf1d14444d17ee7.txt"},{"id":89663593,"identity":"e1d04c2b-ba96-4cca-9d9e-51b9a8154a17","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"xlsx","order_by":13,"title":"","display":"","copyAsset":false,"role":"supplement","size":12082,"visible":true,"origin":"","legend":"","description":"","filename":"STable5bleadsnp.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/3cd369aee692ee5bcea3ef85.xlsx"},{"id":89664686,"identity":"6ac16913-562c-4928-8dcc-fba3d064bc11","added_by":"auto","created_at":"2025-08-22 11:48:27","extension":"txt","order_by":14,"title":"","display":"","copyAsset":false,"role":"supplement","size":12356,"visible":true,"origin":"","legend":"","description":"","filename":"STable12.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/b04b065b8e72c45bc0f512e6.txt"},{"id":89663585,"identity":"56ae45be-9b00-48c9-be26-76028d482281","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":15,"title":"","display":"","copyAsset":false,"role":"supplement","size":11512,"visible":true,"origin":"","legend":"","description":"","filename":"STable13.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/312f362bbc2ec3ae840db419.txt"},{"id":89663588,"identity":"486368a8-9fab-45f1-9632-7804a6ed9d7c","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":16,"title":"","display":"","copyAsset":false,"role":"supplement","size":54,"visible":true,"origin":"","legend":"","description":"","filename":"STable11.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/027dcafc175850fe67e4d593.txt"},{"id":89663595,"identity":"ccee9d84-970c-4747-86a7-0e14790c3005","added_by":"auto","created_at":"2025-08-22 11:32:28","extension":"xlsx","order_by":17,"title":"","display":"","copyAsset":false,"role":"supplement","size":12297,"visible":true,"origin":"","legend":"","description":"","filename":"STable15.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/b6a8f63735ad65513c66eac7.xlsx"},{"id":89664400,"identity":"463863cd-1d5c-499b-b97b-df37de11b4eb","added_by":"auto","created_at":"2025-08-22 11:40:27","extension":"csv","order_by":18,"title":"","display":"","copyAsset":false,"role":"supplement","size":3250982,"visible":true,"origin":"","legend":"","description":"","filename":"STable14.csv","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/b4d4946df6e9774e993b3fc9.csv"},{"id":89663591,"identity":"5775a13b-cba2-403d-b4ce-5f08bde66dfa","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"csv","order_by":19,"title":"","display":"","copyAsset":false,"role":"supplement","size":69025,"visible":true,"origin":"","legend":"","description":"","filename":"STable16.csv","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/2104c68c7ba8ecdd99cee872.csv"},{"id":89664399,"identity":"5ba27686-f196-4617-9900-9fbbb62f2b27","added_by":"auto","created_at":"2025-08-22 11:40:27","extension":"rdata","order_by":20,"title":"","display":"","copyAsset":false,"role":"supplement","size":918,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryFile3.txt.rdata","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/53ebe2c4525475d4679133cb.rdata"},{"id":89663590,"identity":"153128fd-c351-4def-b2c5-e1d562033433","added_by":"auto","created_at":"2025-08-22 11:32:27","extension":"txt","order_by":21,"title":"","display":"","copyAsset":false,"role":"supplement","size":5171980,"visible":true,"origin":"","legend":"","description":"","filename":"STable8.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/6821531072ae874eb0c3b809.txt"},{"id":89663596,"identity":"4ce17ae4-2814-45e8-a69a-d93b713f2108","added_by":"auto","created_at":"2025-08-22 11:32:28","extension":"docx","order_by":22,"title":"","display":"","copyAsset":false,"role":"supplement","size":2609602,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryInformation.docx","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/b3d1c6767a1b4541f2828360.docx"},{"id":89663597,"identity":"d7f6aba5-913f-4cc3-83ce-fb4dc3937663","added_by":"auto","created_at":"2025-08-22 11:32:28","extension":"txt","order_by":23,"title":"","display":"","copyAsset":false,"role":"supplement","size":12636383,"visible":true,"origin":"","legend":"","description":"","filename":"STable9.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/5a2d8f96a291356403e5e024.txt"},{"id":89663600,"identity":"e4859555-b729-4088-a2ba-22bf2038d8cc","added_by":"auto","created_at":"2025-08-22 11:32:32","extension":"txt","order_by":24,"title":"","display":"","copyAsset":false,"role":"supplement","size":93260130,"visible":true,"origin":"","legend":"","description":"","filename":"STable17.txt","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/502da5705a478320e6ad6235.txt"},{"id":89663601,"identity":"f8aef2a6-432d-45c4-bd9f-050c3d30b476","added_by":"auto","created_at":"2025-08-22 11:32:36","extension":"zip","order_by":25,"title":"","display":"","copyAsset":false,"role":"supplement","size":237646448,"visible":true,"origin":"","legend":"","description":"","filename":"STable5.zip","url":"https://assets-eu.researchsquare.com/files/rs-7328766/v1/8d7d35d0c83fae46c0970c3a.zip"}],"financialInterests":"No competing interests reported.","formattedTitle":"Cross-disorder genomic structural equation modeling reveals common genetic basis of neuroimmune diseases","fulltext":[{"header":"Introduction","content":"\u003cp\u003eA century ago, upon discovering the blood-brain barrier, scientists considered the brain an \u0026quot;immune-privileged organ\u0026quot;. However, we now recognize that this \u0026quot;privilege\u0026quot; masks the most complex interactions between the nervous and immune systems[1,2]. Neuroimmune syndromes (NIS) represent not merely a disease spectrum concept, but rather a complex, multidimensional pathobiological process, characterized by blood-brain barrier dysfunction and neuroimmune homeostatic imbalance, profoundly influenced by genetic susceptibility, environmental triggers, and immune system dysregulation[3-5].\u003c/p\u003e\n\u003cp\u003eWith the dramatic global increase in autoimmune disease incidence, the prevalence of neuroimmune syndromes has risen rapidly, emerging as a major challenge in neuroscience, immunology, and medical genetics[6,7]. Despite significant advances in understanding molecular mechanisms and targeted therapies for individual neuroimmune diseases, our comprehension of the specific shared genetic and cross-disease biological mechanisms underlying neuroimmune syndromes remains limited. Studies indicate that neuroinflammatory responses, autoimmune attacks, and myelin damage dysfunction may constitute important drivers of neuroimmune syndromes, yet these findings remain insufficient to fully explain the substantial inter-individual variation in neuroimmune disease progression and susceptibility[8,9].\u003c/p\u003e\n\u003cp\u003eTo address these challenges, this investigation aims to integrate multiple genetic analytical tools and robust association methods to unveil potential common molecular mechanisms and expand the genetic connections between neuroimmune syndromes and various related disorders. Specifically, we focus on pleiotropic genomic loci and critical chromosomal regions associated with neuroimmune syndromes to reveal potential therapeutic targets. This study not only expands our understanding of neuroimmune syndromes but also provides theoretical and practical support for precision medicine intervention strategies in global neuroimmune diseases[10].\u003c/p\u003e\n\u003cp\u003eTo address the current lack of precise measurement of common mechanisms in neuroimmune syndromes, we designed an innovative GWAS approach targeting latent, unmeasured neuroimmune syndrome phenotypes. We employed genomic structural equation modeling (Genomic SEM) applied to published GWAS summary statistics of neuroimmune-related diseases and immune biomarkers[11]. Through these statistics, we obtained SNP association strengths with latent neuroimmune syndrome phenotypes, thereby establishing a GWAS study of previously unmeasured latent neuroimmune syndrome phenotypes.\u003c/p\u003e\n\u003cp\u003eWe further adapted comprehensive analytical methods from systems biology, defining genetic variants in our neuroimmune syndrome structural equation model that remain unexplained by known immune biomarkers as potential cross-disease shared genetic markers. These underwent extensive GWAS-related functional annotation and pathway enrichment analyses. While this approach may not perfectly capture the true relationships between neuroimmune disease-related pathways and multifactorial interactions, as neuroimmune syndromes represent complex processes driven jointly by genetic, environmental, and stochastic immune dysregulation factors, this analysis effectively excludes confounding influences based on single neuroimmune disease markers, enabling precise analysis of previously difficult-to-study cross-disease common mechanisms.\u003c/p\u003e\n\u003cp\u003eFrom a direct clinical application perspective, we conducted tens of thousands of association analyses to construct a concise, practical genetic risk factor atlas for non-biostatistical professionals (neurologists, rheumatologists), enabling direct application of relevant risk factor maps for developing potential individualized prevention and precision intervention strategies for patients. Our research aims to establish a simple yet efficient translational pathway from genomic statistics to neuroimmune basic research and clinical precision medicine strategy development.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eA flowchart overview is presented in Fig.1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eStudy Design and Data Sources\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur genomic structural equation modeling GWAS for neuroimmune syndromes utilized summary statistics from five independent GWAS studies encompassing neuroimmune-related disorders: Guillain-Barr\u0026eacute;\u0026nbsp;syndrome (GBS), myasthenia gravis (MG), multiple sclerosis (MS), systemic lupus erythematosus (SLE), and systemic connective tissue diseases (SCTD). All constituent GWAS studies received institutional review board approval, with informed consent obtained from all participants, and underwent rigorous quality control procedures.\u003c/p\u003e\n\u003cp\u003eGuillain-Barr\u0026eacute;\u0026nbsp;Syndrome (GBS): GWAS data were obtained from the FinnGen Research Program Release 12 (n = 492,134; 551 cases, 491,583 controls)[12]. The FinnGen initiative represents a large-scale Finnish biobank study designed to provide comprehensive insights into genetic foundations and risk factors for rare neuroimmune disorders through integration of genome-wide sequencing and health registry data. This dataset leverages the unique genetic architecture of the Finnish population, enabling enhanced detection of rare acute inflammatory demyelinating polyneuropathies.\u003c/p\u003e\n\u003cp\u003eMyasthenia Gravis (MG): Summary statistics were derived from FinnGen R12 (n = 496,227; 560 cases, 495,667 controls). This investigation utilized comprehensive diagnostic data from the Finnish healthcare registry system combined with large-scale genotyping, providing valuable genetic information for neuromuscular junction autoimmune disorders.\u003c/p\u003e\n\u003cp\u003eMultiple Sclerosis (MS): GWAS data originated from FinnGen R12 (n = 498,857; 2,926 cases, 495,931 controls). This study employed stringent phenotype definitions and quality control procedures, establishing a crucial foundation for central nervous system demyelinating disease genetics research. As the most prevalent phenotype within our neuroimmune syndrome framework, MS provided substantial statistical power.\u003c/p\u003e\n\u003cp\u003eSystemic Lupus Erythematosus (SLE): Data were obtained from FinnGen R12 (n = 499,333; 291 cases, 499,042 controls). As a prototypical systemic autoimmune disease, this study provided foundational insights into genetic mechanisms underlying systemic immune dysregulation components of neuroimmune syndromes.\u003c/p\u003e\n\u003cp\u003eSystemic Connective Tissue Diseases (SCTD): GWAS data were sourced from FinnGen R12 (n = 500,348; 16,088 cases, 484,260 controls). SCTD, representing mixed connective tissue disorders frequently involving neurological manifestations, contributed the largest case sample size, providing substantial genetic evidence and statistical power for understanding systemic pathological mechanisms in neuroimmune syndromes.\u003c/p\u003e\n\u003cp\u003eAll GWAS datasets underwent comprehensive quality control procedures, including sample quality filtering, single nucleotide polymorphism (SNP) quality screening (minor allele frequency [MAF] \u0026gt; 0.01, imputation quality INFO \u0026gt; 0.8), population stratification correction, and relatedness adjustment. To ensure data consistency, all analyses utilized the GRCh37/hg19 reference genome with the 1000 Genomes Project European population as the linkage disequilibrium (LD) reference panel. Detailed GWAS specifications are provided in Supplementary Table 1.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eQuality Control Procedures\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSample-level filtering: Samples with missing call rates exceeding 5% were excluded from analyses.\u003c/p\u003e\n\u003cp\u003eMHC region handling: Given the genetic diversity and structural complexity of the major histocompatibility complex (MHC) region (chromosome 6: 25,000,000-35,000,000 bp), particularly immune-related gene polymorphisms, specialized processing was applied to this genomic region[13].\u003c/p\u003e\n\u003cp\u003eSNP-level quality control: For constructing neuroimmune syndrome summary statistics, we employed recommended default quality control parameters, retaining all autosomal SNPs from the five constituent neuroimmune-related GWAS studies after filtering to the 1000 Genomes Phase 3 European reference panel. SNPs with MAF \u0026lt; 0.01 were excluded due to increased error rates from small genotype clusters and typically elevated LD score regression standard errors. Additionally, SNPs with zero effect estimates were removed to prevent matrix singularity issues essential for genomic structural equation modeling. SNPs inconsistent with the reference panel and those with allelic mismatches were also excluded.\u003c/p\u003e\n\u003cp\u003eSample overlap assessment: Given that constituent single-trait GWAS originated from different genomic repositories with distinct participant cohorts, we carefully considered potential sample intersections across different cohorts to ensure result accuracy and generalizability while accounting for statistical implications of potential sample overlap.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenomic Structural Equation Modeling\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe implemented genomic structural equation modeling (Genomic SEM) using the GenomicSEM R package (v.0.0.5)[14] to conduct genomic structural equation GWAS analysis across GBS, MG, MS, SLE, and SCTD, investigating broad shared genetic susceptibility underlying these neuroimmune-related phenotypes. Genomic SEM represents a novel multivariate methodology enabling exploration of multiple latent multivariate models to investigate potential common genetic architecture across the neuroimmune disease spectrum. Detailed analytical standards are provided in Table 1.\u003c/p\u003e\n\u003cp\u003eGenomic SEM demonstrates robustness against sample overlap bias (e.g., FinnGen participants appearing across multiple input GWAS disease phenotypes) and sample size imbalances[15]. Additionally, this approach facilitates identification of genetic variants affecting only subsets rather than all neuroimmune phenotypes, thus distinguishing variants representing broad cross-disease susceptibility from those reflecting disease-specific genetic mechanisms.\u003c/p\u003e\n\u003cp\u003eGenomic SEM analysis proceeded through two distinct phases. Phase I estimated empirical genetic covariance matrices and corresponding sampling covariance matrices. We prepared neuroimmune-related disease GWAS summary statistics for Phase I analysis, employing multivariate extensions of cross-trait LD score regression to generate empirical genetic covariance matrices among the five neuroimmune phenotypes as input for SEM common factor modeling (Supplementary File 3). Phase II specified SEM models minimizing discrepancies between hypothesized covariance structures and empirically calculated covariance matrices from Phase I. Given our primary objective of identifying shared genetic architecture underlying the five neuroimmune-related phenotypes, we tested a single-factor model to characterize neuroimmune syndromes. Model fit was evaluated using standardized root mean square residual (SRMR), model \u0026chi;\u0026sup2;, Akaike information criterion (AIC), and comparative fit index (CFI)\u0026nbsp;(Supplementary Table S4a-b). Through application of appropriate common factor SEM specifications, individual autosomal SNP associations were incorporated into genetic and corresponding sample covariance matrices, generating neuroimmune syndrome structural equation GWAS results for 7,170,714 SNPs representing shared covariance across the five constituent neuroimmune-related GWAS studies.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSNP Heterogeneity Assessment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo evaluate whether SNP associations in our neuroimmune syndrome structural equation GWAS were appropriately modeled within the multivariate SEM framework, we calculated SNP heterogeneity statistics (Q_SNP). The null hypothesis posited that SNP associations from individual phenotype GWAS could be completely statistically accounted for by our neuroimmune syndrome structural equation model. Consequently, significant Q_SNP values (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05) in our neuroimmune syndrome structural equation GWAS indicated that specific SNPs exerted effects through pathways beyond the established shared genetic mechanisms among neuroimmune-related diseases in our model. This heterogeneity analysis facilitated identification of genetic variants potentially exerting unique effects on specific neuroimmune phenotypes rather than operating through common genetic factors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMulti-level Model Evaluation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe implemented multi-tiered strategic adjustments for our genomic structural equation model, including different significance thresholds (\u003cem\u003eP\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-12\u0026nbsp;\u003c/sup\u003eand \u003cem\u003eP\u0026nbsp;\u003c/em\u003e\u0026lt; 5\u0026times;10\u003csup\u003e-12\u003c/sup\u003e) to identify novel SNP loci across varying confidence levels, balancing statistical power with false positive control. Simultaneously, we employed genomic control strategies based on two-step LD Score regression methodology. Quality control parameters retained all SNPs with missing values, INFO scores \u0026lt; 0.9, MAF \u0026lt; 0.01, P-values outside conventional ranges, and non-standard or ambiguous strand orientations, exclusively removing partitioned LD scores with zero variance and utilizing two-step estimators for analysis (cutoff threshold = 30).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenomic Loci Definition and Novel Variant Identification\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe utilized the \u0026quot;Functional Mapping and Annotation of Genetic Associations\u0026quot; methodology implemented in FUMA to identify genomic loci and determine lead SNPs associated with our neuroimmune syndrome structural equation GWAS[16]. These SNPs exhibited low LD correlation with other SNPs (r\u0026sup2;\u0026nbsp;\u0026lt; 0.1) while achieving genome-wide significance (\u003cem\u003eP\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-8\u003c/sup\u003e). Initially, we input neuroimmune syndrome structural equation SNP summary statistics to assess association strengths. We compared lead SNPs and loci with original univariate GWAS relationships, defining loci as novel when located \u0026gt;1 Mb from previously identified loci in constituent univariate GWAS data.\u003c/p\u003e\n\u003cp\u003eTo determine whether the 36 lead SNPs from our neuroimmune syndrome structural equation GWAS exhibited pleiotropic associations, we consulted published significant associations (\u003cem\u003eP\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-8\u003c/sup\u003e) in the GWAS Catalog[17]. Additionally, we conducted risk loci analysis using FUMA software functionality with significance thresholds of \u003cem\u003eP\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-8\u003c/sup\u003e, analyzing output files through MAGMA (Multi-marker Analysis of GenoMic Annotation)[18]. MAGMA serves as a post-GWAS processing tool designed to evaluate gene-phenotype associations by aggregating multiple genetic markers into gene-level signals and calculating gene-phenotype association strengths. This approach extracts gene function-related information from genome-wide SNP data for gene-level genetic signal analysis, employing significance thresholds of FDR \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFine-mapping Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify most probable causal variants associated with our neuroimmune syndrome structural equation GWAS, we employed SuSIE and FINEMAP methodologies implemented in the echolocatoR R package (v.2.0.3). We established posterior probability thresholds of 0.95 for defining credible sets of potential causal variants.\u003c/p\u003e\n\u003cp\u003eCausal variant identification: Both SuSIE (Sum of Single Effects) and FINEMAP represent fine-mapping analytical tools designed to determine most likely causal variants associated with specific phenotypes. We utilized 250 kb windows encompassing regions associated with each lead SNP, calculating causal inference probabilities for each SNP within these regions. Credible sets: We established 0.95 posterior probability thresholds; variants exceeding this threshold were considered potential causal variants. Consensus SNPs: echolocatoR defined \u0026apos;consensus SNPs\u0026apos; as variants appearing in both SuSIE and FINEMAP results, calculating average posterior probabilities and determining average credible sets based on probability outcomes (credibility = 1 when both SuSIE and FINEMAP SNP posterior probabilities exceeded 0.95, otherwise 0).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTranscriptome-wide Association Study\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFollowing potential causal variant localization, we conducted transcriptome-wide association study (TWAS) to prioritize genes associated with our neuroimmune syndrome structural equation GWAS based on gene expression-phenotype relationships[19]. We employed the FUSION methodology using 37,920 pre-computed expression quantitative trait loci (eQTL) features (gene/tissue pairs) from GTEx v.8 data. These features facilitated calculation of expression associations across different genes and tissues[20].\u003c/p\u003e\n\u003cp\u003eTWAS results analysis: Our neuroimmune syndrome structural equation GWAS data contained sufficient variation to analyze 36,149 features (from 37,920 eQTL features), indicating high data quality. Genes with \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05 (significantly associated with our neuroimmune syndrome structural equation GWAS) were included in subsequent analyses.\u003c/p\u003e\n\u003cp\u003eFor TWAS-significant genes, we implemented FOCUS methodology (fine-mapping approach specifically designed for TWAS studies)[21]. FOCUS evaluates potential causal relationships between genes and phenotypes based on posterior inclusion probabilities. We considered TWAS-significant genes demonstrating both TWAS significance and consistency with additional evidence (e.g., FOCUS), suggesting potential causality.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGene Set and Disease Ontology Enrichment Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe conducted gene enrichment and pathway analyses using MAGMA and FUMA (GSEA) to investigate potential relationships between our neuroimmune syndrome structural equation GWAS and Mendelian disease genes with associated pathways[22]. Additionally, we performed gene enrichment analysis using MendelVar (https://mendelvar.mrcieu.ac.uk/submit/).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCell-type Annotation Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eTo identify etiological cell types associated with our neuroimmune syndrome structural equation GWAS, we employed CELLECT (cell-type expression specificity integration for complex traits using single-cell RNA sequencing data). We utilized the Tabula Muris dataset containing transcriptomic data from 100,000 mouse (Mus musculus) cells across 20 organs and tissues. We preprocessed and normalized Tabula Muris single-cell RNA sequencing data using CELLEX, calculating expression specificity scores for each gene. Cell-type-specific analysis was performed using LDSC software with cell type classification and false discovery rate (FDR) thresholds of 0.05[23].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenomic Regional Heritability Partitioning\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe employed LDSC tools to calculate partitioned heritability across genomic regions[24]. This approach assigns phenotypic genetic information to different genomic regions (e.g., genes, enhancers, silencers) to evaluate each region\u0026apos;s contribution to phenotypic heritability. Specifically, LDSC utilizes weighted LD matrices, genotype frequency files, and summary statistics for calculations, ultimately estimating genetic contributions from each region.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eBiomarker and Risk Factor Association Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFor candidate drug target validation, we utilized expression quantitative trait loci (eQTL) data from the eQTLGen Consortium[25], comprising gene expression measurements from 31,864 individuals of European ancestry. For proteome-wide drug target discovery, we analyzed 4,907 plasma proteins from the deCODE genetics consortium[26]. Additionally, we incorporated brain, cerebrospinal fluid (CSF), and plasma protein quantitative trait loci (pQTL) data from Yang et al.[27], which provided complementary neurologically-relevant proteomic measurements across multiple tissue types. For comprehensive biomarker screening, we conducted phenome-wide association analysis using 50,033 phenotypes from the IEU OpenGWAS database[28].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSummary Statistics-based Polygenic Score Construction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe calculated polygenic risk scores (PRS) based on genome-wide summary statistics and evaluated genetic contributions from different chromosomal regions to disease susceptibility[23]. Specifically, we utilized PRS-CS (Polygenic Risk Score with Continuous Shrinkage) software to estimate SNP posterior effect sizes through GWAS data and external LD reference panels, subsequently calculating polygenic risk scores[24]. This methodology employs Bayesian regression models to estimate effect sizes by integrating LD reference panels based on GWAS summary statistics, ultimately computing PRS values.\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003e\u003cstrong\u003eStructural Equation Model Statistical Framework Construction\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLD score regression analysis revealed differential heritability contributions across the five constituent univariate input GWAS studies: GBS (h\u003csup\u003e2\u003c/sup\u003e = 0.0013, Z = 1.56), MG (h\u003csup\u003e2\u003c/sup\u003e = 0.0019, Z = 1.95), MS (h\u0026sup2;\u0026nbsp;= 0.0084, Z = 6.61), SLE (h\u003csup\u003e2\u0026nbsp;\u003c/sup\u003e= 0.0011, Z = 1.08), and SCTD (h\u003csup\u003e2\u0026nbsp;\u003c/sup\u003e= 0.0085, Z = 4.90). Pairwise genetic covariance estimates demonstrated substantial shared genetic architecture: GBS-MG (0.5016), GBS-MS (0.1685), GBS-SLE (0.293), GBS-SCTD (0.2469), MG-MS (0.033), MG-SLE (0.2303), MG-SCTD (0.5397), MS-SLE (0.235), MS-SCTD (0.1989), and SLE-SCTD (0.5854) (detailed single-factor genetic parameters in Supplementary Table 2, Figure 2).\u003c/p\u003e\n\u003cp\u003eStructural equation modeling analysis preceding model construction demonstrated excellent fit between the genetic covariance matrix from five input GWAS studies and the empirical covariance matrix under a common factor model (comparative fit index [CFI] = 1, standardized root mean square residual [SRMR] = 0.095) (detailed model stability assessment in Supplementary Table 4a; latent factor [F1] and univariate structural equation model parameters in Supplementary Table 4b).\u0026nbsp;Exploratory factor modeling (Supplementary File 3) provided compelling evidence for shared genetic\u0026nbsp;factors underlying these neuroimmune phenotypes.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenomic Structural Equation Model GWAS Stratified Evaluation\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eUsing genomic structural equation modeling (genomic SEM) to incorporate individual genetic variation across multiple neuroimmune disorders, we conducted an indirectly measured genome-wide association study (GWAS) that estimated associations between 7,170,712 single nucleotide polymorphisms (SNPs) and a latent neuroimmune syndrome factor. The complete results are presented in Supplementary Table 5. Under the \u003cem\u003eP\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-12\u003c/sup\u003e threshold, we identified 31 lead SNPs across 41 genomic loci (Supplementary Table 5a), while the more stringent \u003cem\u003eP\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-16\u003c/sup\u003e threshold yielded 29 lead SNPs across 39 genomic loci (Supplementary Table 5b). Among the 7,170,712 SNPs analyzed across different P-value thresholds, 34 represented novel discoveries distinct from loci identified in the five constituent single-trait GWAS studies, highlighting the enhanced discovery power of genomic SEM. These novel neuroimmune syndrome lead SNPs demonstrated enrichment in pathways related to immune system regulation, neurodevelopment, and inflammatory responses.\u003c/p\u003e\n\u003cp\u003eAt the \u003cem\u003eP\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-12\u003c/sup\u003e stratification level, 31 of 41 lead SNPs had been previously identified in the literature (though not in the context of neuroimmune structural equation modeling) (Supplementary Table 5c). Similarly, under the\u003cem\u003e\u0026nbsp;P\u003c/em\u003e \u0026lt; 5\u0026times;10\u003csup\u003e-16\u003c/sup\u003e stratification, 29 of 39 lead SNPs showed prior literature identification (Supplementary Table 5d). Key lead SNPs identified include rs148729815, rs76210604, rs17098406, and rs73195472, among others, providing crucial novel insights into neuroimmune syndrome genetic architecture. This investigation, through implementation of genomic structural equation modeling, revealed SNPs associated with neuroimmune syndromes across different association thresholds, offering important insights into shared genetic foundations and potential therapeutic targets.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenomic Control Assessment Based on LD Score Regression\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur systematic quality control procedures resulted in exclusion of 6,032,680 SNPs while retaining 1,138,032 effective SNPs following regression coefficient preservation criteria. Comprehensive genomic inflation assessment revealed: mean \u0026chi;\u0026sup2; = 0.523, genomic control lambda (\u0026lambda;GC) = 0.908, maximum \u0026chi;\u0026sup2; = 913.763, genome-wide significant hits = 7, heterogeneity testing = passed (\u003cem\u003eP\u003c/em\u003e \u0026gt; 0.05), observed-scale heritability (h\u003csup\u003e2\u003c/sup\u003e) = 0.0139 (SE = 0.0021), genetic-environmental contribution ratio \u0026lt; 0, regression intercept = 0.4979, and regression intercept standard error = 0.0025. These multiple estimation parameters collectively demonstrate that potential inflation in our structural equation framework resulted from polygenic heritability signals rather than population stratification bias or pleiotropy parameter effects.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFUMA-based Neuroimmune Syndrome Structural Equation Model Assessment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFUMA software evaluation of our genomic structural equation model identified 36 risk gene loci (Supplementary Table 5e, Figure 3), with 2 potential neuroimmune syndrome-associated genes achieving genome-wide significance control (significance threshold = 5\u0026times;10\u003csup\u003e-8\u003c/sup\u003e, FDR \u0026lt; 0.05)(Figure 4).Through FUMA annotation, we mapped 34 lead SNP loci, with the majority located in intergenic regions (Supplementary Table 6). Our novel GWAS subtraction analysis did not identify genetic variants significantly associated with neuroimmune syndromes beyond those detected through standard approaches.No GWAS subtraction loci were identified (Supplementary Table 7).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFine-mapping Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFine-mapping analysis identified high-confidence causal variants (mean posterior probability \u0026gt; 0.95) across four major association regions: chromosome 1 association cluster (encompassing 6 variants including rs56332939 and rs76692181), chromosome 4 association cluster (7 variants including rs10029041 and rs7664257), chromosome 8 association cluster (4 variants including rs35642230 and rs72678550), and chromosome 11 association cluster (4 variants including rs34851197 and rs71482185). Regional association plots demonstrated pronounced association peaks at these loci (Figure 5 and Supplementary Table 8).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTranscriptomic Prediction Analysis\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSubsequently, we conducted transcriptome-wide association study (TWAS) using FUSION to identify gene-level associations with neuroimmune syndromes. Following multiple comparison correction, no genes achieved significance thresholds for association (Extended Data Table 1).\u0026nbsp;However, subsequent FOCUS fine-mapping analysis of our genomic structural equation data identified 78 genes potentially representing pathogenic signals for neuroimmune syndromes.\u003c/p\u003e\n\u003cp\u003eTo further characterize these high-confidence gene-level associations, we performed intersection analysis. Seven genes (ISCA2, NPC2, PRSS8, ENDOD1, PNKD, GSAP, and ATG101) exhibited positive TWAS Z-scores, indicating that predicted gene expression positively correlates with neuroimmune syndrome risk, suggesting that upregulation of these genes may associate with increased neuroimmune syndrome susceptibility. Conversely, eight genes (PGBD1, DNAJC24, LINC02980, TUBGCP5, LSM12P1, ATF6B, PFN1P2, and MDC1) showed negative TWAS Z-scores, indicating that their downregulation associates with increased neuroimmune syndrome risk (TWAS and FOCUS intersection results in Supplementary Table 9).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePathway, Cell Type, and Mendelian Disease Gene Enrichment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eMulti-marker genomic annotation analysis (MAGMA) identified 2 genes through genomic mapping (Supplementary Table 10). Gene\u0026nbsp;set analysis utilizing these genes demonstrated enrichment in Gene Set Enrichment Analysis (GSEA) categories (Supplementary Table 11,Figure 6).\u0026nbsp;Additionally, biological processes mapped through gene enrichment were validated in Gene Ontology (GO) terms, including \u0026quot;regulation of mast cell cytokine production\u0026quot; and \u0026quot;positive regulation of mast cell cytokine production.\u0026quot;\u003c/p\u003e\n\u003cp\u003eDisease enrichment analysis revealed multiple neuroimmune-related disorders exceeding significance standards following multiple comparison correction (Supplementary Table 12). The most significant disease categories included\u0026nbsp;ectodermal dysplasia 10A and neurodegenerative disease. Neuroimmune syndrome enrichment proved most significant in immune regulatory processes, with identified biological processes extensively involving mast cell and T cell regulation (regulation of mast cell cytokine production, regulation of T cell differentiation). Nervous system disease enrichment was similarly significant (including neurodegenerative disease, motor neuron disease, and nervous system disease; \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05). EEG abnormality phenotypes (EEG with polyspike wave complexes) provided additional support for neurophysiological alterations.\u003c/p\u003e\n\u003cp\u003eCell type enrichment analysis revealed that while no cell types achieved significance following multiple comparison correction, 19 cell types demonstrated enrichment at nominal significance levels (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05). The two most significant cell types were kidney macrophages (\u003cem\u003eP\u003c/em\u003e = 0.0023) and tracheal blood cells (\u003cem\u003eP\u003c/em\u003e = 0.0025). Neuroimmune syndrome enrichment patterns in immune cells were pronounced, with 18 of 19 nominally significant cell types representing immune cell populations (including various macrophage, B cell, myeloid cell, and natural killer cell subtypes), underscoring the critical role of immune systems in disease mechanisms.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eGenomic Regional Heritability Contribution Results\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eGenomic regional heritability contribution analysis identified 25 genomic functional regions achieving significance following multiple comparison correction (FDR \u0026lt; 0.05) (Supplementary Table 13). Genetic contributions concentrated primarily in chromosomal regulatory regions, including transcription start sites (TSS), promoter regions, enhancers, and super-enhancers as key regulatory elements. The most significant regions included transcription start sites (TSS_Hoffman; enrichment = 21.77, FDR = 0.0007) and H3K4me3-modified transcriptionally active regions (enrichment = 13.30, FDR = 0.016).\u003c/p\u003e\n\u003cp\u003eHistone modification-associated regions demonstrated significant enrichment, particularly H3K27ac, H3K4me1, and H3K4me3 modification regions representing important markers of transcriptional activity and enhancer function. Promoter regions (Promoter_UCSC) and various enhancer regions (including super-enhancers SuperEnhancer_Hnisz) all displayed significant heritability enrichment, indicating central roles of gene expression regulatory mechanisms in disease susceptibility. These findings suggest that genetic variants primarily exert effects through influencing gene transcriptional regulatory networks rather than altering protein-coding sequences.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMulti-Platform Drug Target and Biomarker Discovery\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAnalysis of 4,907 plasma proteins from the deCODE genetics consortium identified eight significant associations (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05). BDH2 showed the strongest risk association (OR = 1.14, 95% CI: 1.02-1.26, \u003cem\u003eP\u003c/em\u003e = 0.018), followed by WFDC2 (OR = 1.13, 95% CI: 1.00-1.26, \u003cem\u003eP\u003c/em\u003e = 0.047), ALPL (OR = 1.07, 95% CI: 1.00-1.14, \u003cem\u003eP\u003c/em\u003e = 0.048), LILRA2 (OR = 1.04, 95% CI: 1.00-1.09, \u003cem\u003eP\u003c/em\u003e = 0.046), PCDHA4 (OR = 1.04, 95% CI: 1.00-1.07, \u003cem\u003eP\u003c/em\u003e = 0.026), GRID2 (OR = 1.03, 95% CI: 1.01-1.06, \u003cem\u003eP\u003c/em\u003e = 0.015), and RNASE6 (OR = 1.03, 95% CI: 1.00-1.07, \u003cem\u003eP\u003c/em\u003e = 0.039). EPB41 was the only protective protein (OR = 0.85, 95% CI: 0.73-1.00, \u003cem\u003eP\u003c/em\u003e = 0.046)(Supplementary Table 14,\u0026nbsp;Fig.7).\u003c/p\u003e\n\u003cp\u003eAnalysis of candidate therapeutic targets identified five significant associations. Protective effects were observed for CTLA4 (OR = 0.91, 95% CI: 0.86-0.96,\u003cem\u003e\u0026nbsp;P\u003c/em\u003e = 0.0005), TYK2 (OR = 0.97, 95% CI: 0.96-0.99, \u003cem\u003eP\u003c/em\u003e = 0.008), and CCR5 (OR = 0.97, 95% CI: 0.95-0.99, \u003cem\u003eP\u003c/em\u003e = 0.003). Risk effects were found for LILRA2 (OR = 1.01, 95% CI: 1.00-1.02, \u003cem\u003eP\u003c/em\u003e = 0.031) and BDH2 (OR = 1.01, 95% CI: 1.00-1.02, \u003cem\u003eP\u003c/em\u003e = 0.0075)(Supplementary Table 15,Fig.8).\u003c/p\u003e\n\u003cp\u003eAnalysis using Yang et al. brain, cerebrospinal fluid, and plasma pQTL data identified 529 proteins with significant associations (FDR \u0026lt; 0.05). Key findings included complement components (C2, C5, C9), inflammatory mediators (TNF, IL1B), glial activation markers (GFAP), neuroprotective factors (GDNF, NGF), immune checkpoint molecules (CD274), and vascular markers (VCAM1). EPB41 showed consistent associations across both deCODE and Yang et al. Analyses(Supplementary Table 16,Figure 9).\u003c/p\u003e\n\u003cp\u003eAnalysis of 50,033 phenotypes identified 17 traits with significant causal associations (FDR \u0026lt; 0.05). The strongest association was thyroid problems (not cancer) (OR = 1.826, 95% CI: 1.494-2.232, \u003cem\u003eP\u003c/em\u003e = 4.11\u0026times;10\u003csup\u003e-9\u003c/sup\u003e, FDR = 9.43\u0026times;10\u003csup\u003e-5\u003c/sup\u003e), followed by self-reported hypothyroidism (OR = 1.743, 95% CI: 1.393-2.180, \u003cem\u003eP\u003c/em\u003e = 1.14\u0026times;10\u003csup\u003e-6\u003c/sup\u003e) and rheumatoid arthritis (OR = 1.035, 95% CI: 1.022-1.047, \u003cem\u003eP\u003c/em\u003e = 3.24\u0026times;10\u003csup\u003e-8\u003c/sup\u003e). Sixteen traits showed risk-increasing effects, supporting shared autoimmune and endocrine mechanisms(Supplementary Table 17, Figure 10).\u003c/p\u003e\n\u003cp\u003eThe analyses identified novel metabolic targets (BDH2), established drug targets (CTLA4, TYK2, CCR5), validated inflammatory pathways (complement, TNF), neuroprotective networks (GDNF, NGF), and clinical disease associations (thyroid disorders, autoimmune diseases). EPB41 demonstrated cross-platform consistency, while BDH2 and LILRA2 showed convergent evidence across multiple approaches.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eChromosomal-level Results\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003ePolygenic risk score analysis using 1,077,341 effective variants revealed a highly non-uniform distribution of genetic contributions across chromosomes, with a dominant two-chromosome architecture. Chromosome 4 demonstrated the highest genetic contribution (39.1%), followed by chromosome 14 (23.8%), together accounting for 62.9% of the total genetic variance in neuroimmune syndrome susceptibility.\u003c/p\u003e\n\u003cp\u003eBeyond the two dominant chromosomes, genetic contributions were substantially lower, with chromosomes 1 (3.5%) and 2 (3.2%) showing modest contributions above the 5% threshold. Notably, chromosome 6, containing the HLA region, contributed 2.4%, indicating that while HLA variants play a role, the genetic architecture is predominantly non-HLA driven. All remaining chromosomes contributed less than 3% each to the overall genetic risk.\u003c/p\u003e\n\u003cp\u003eCorrelation analysis between variant count and genetic contribution revealed a weak linear relationship (Pearson r = 0.182, \u003cem\u003eP\u003c/em\u003e = 0.418) but a strong rank correlation (Spearman \u0026rho; = 0.874, \u003cem\u003eP\u003c/em\u003e = 2.14\u0026times;10\u003csup\u003e-6\u003c/sup\u003e), demonstrating that effect size rather than variant quantity drives chromosomal contributions. This suggests the presence of high-impact variants concentrated on chromosomes 4 and 14, warranting detailed investigation for major effect loci and therapeutic target identification(Figure S11).\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eThis investigation represents the first comprehensive genetic architecture mapping of neuroimmune syndromes through genomic structural equation modeling of latent, unmeasured phenotypes. By integrating five neuroimmune-related disorders\u0026mdash;Guillain-Barr\u0026eacute; syndrome, myasthenia gravis, multiple sclerosis, systemic lupus erythematosus, and systemic connective tissue diseases\u0026mdash;we established a novel GWAS framework that identified 33 genome-wide significant loci, 78 candidate pathogenic genes, and critical therapeutic targets. Our findings reveal that neuroimmune syndrome susceptibility operates through complex multi-system mechanisms spanning metabolic dysregulation, immune checkpoint dysfunction, and transcriptional regulatory networks, providing unprecedented insights for precision medicine strategies[31,32,33].\u003c/p\u003e\n\u003cp\u003eOur genomic structural equation modeling analysis demonstrated substantial genetic covariances among the five constituent phenotypes, with the strongest correlations observed between systemic lupus erythematosus and systemic connective tissue diseases , myasthenia gravis and systemic connective tissue diseases , and Guillain-Barr\u0026eacute; syndrome and myasthenia gravis . These findings provide quantitative evidence for shared genetic susceptibility foundations in autoimmune diseases[34,35,36], supporting the hypothesis that common immune dysregulation mechanisms underlie diverse neuroimmune manifestations. Multiple sclerosis and systemic connective tissue diseases exhibited relatively higher heritability estimates within this disease spectrum, suggesting more pronounced genetic contributions to these conditions. The excellent model fit confirms that these diseases represent interconnected spectra rather than isolated entities, operating through shared genetic mechanisms that transcend traditional diagnostic boundaries[37].\u003c/p\u003e\n\u003cp\u003eFunctional annotation analysis revealed that the 33 identified loci are predominantly enriched in gene regulatory regions rather than protein-coding sequences, with transcription start sites showing the strongest enrichment \u0026nbsp;followed by H3K4me3-modified transcriptionally active regions. Twenty-five genomic functional regions achieved significance after multiple comparison correction, encompassing enhancers, super-enhancers, and histone modification sites (H3K27ac, H3K4me1, H3K4me3). This regulatory enrichment pattern aligns with established frameworks demonstrating that autoimmune disease variants primarily function through epigenetic modifications and chromatin architecture alterations[38,39,40,41]. These findings support the omnigenic model[42], where disease susceptibility emerges from coordinated dysregulation of transcriptional networks rather than individual gene effects, indicating that therapeutic interventions targeting regulatory mechanisms may prove more effective than approaches focused on single protein targets[43].\u003c/p\u003e\n\u003cp\u003eFine-mapping analysis identified 21 high-confidence causal variants distributed across four chromosomal clusters (chromosomes 1, 4, 8, and 11), providing precise molecular targets for functional validation. Transcriptomic analysis through FOCUS methodology revealed 78 candidate pathogenic genes exhibiting complex bidirectional regulatory patterns\u0026mdash;upregulation of genes including ISCA2, NPC2, PRSS8, ENDOD1, PNKD, GSAP, and ATG101, and downregulation of genes including PGBD1, DNAJC24, LINC02980, TUBGCP5, LSM12P1, ATF6B, PFN1P2, and MDC1 both associated with increased disease risk[44]. This bidirectional regulation reflects the intricate homeostatic balance required for neuroimmune system stability, where both excessive activation and insufficient suppression contribute to pathogenesis[45]. Pathway enrichment analysis revealed significant involvement of mast cell cytokine production regulation and T cell differentiation processes, connecting our genetic discoveries to established immunological mechanisms in neuroinflammation[46,47,48].\u003c/p\u003e\n\u003cp\u003ePolygenic risk score analysis using PRS-CS methodology with 1,077,341 effective variants revealed a remarkably concentrated genetic architecture dominated by two chromosomes. Chromosome 4 contributed 39.1% of total genetic variance, followed by chromosome 14 (23.8%), together accounting for 62.9% of neuroimmune syndrome susceptibility. This highly non-uniform distribution contrasts sharply with traditional polygenic models and suggests the presence of major effect loci warranting intensive investigation[49,50]. Importantly, chromosome 6 containing the HLA region contributed only 2.4%, indicating that neuroimmune syndrome genetic susceptibility operates predominantly through non-HLA mechanisms\u0026mdash;a finding that challenges conventional autoimmune disease models emphasizing HLA dominance. Correlation analysis demonstrated that effect size rather than variant quantity drives chromosomal contributions, suggesting high-impact variants concentrated on chromosomes 4 and 14 represent priority targets for therapeutic development and mechanistic investigation[51].\u003c/p\u003e\n\u003cp\u003eOur systematic multi-platform Mendelian randomization approach identified therapeutic opportunities across multiple biological pathways, providing genetic validation for both novel and established drug targets[52,53,54]. Novel discoveries included BDH2, establishing the first causal connection between ketone body metabolism and neuroimmune diseases, suggesting that metabolic interventions targeting mitochondrial function may provide therapeutic benefits. Established drug target validation identified three proteins with significant protective associations: CTLA4, TYK2, and CCR5, providing genetic evidence for immune checkpoint inhibition, JAK inhibition, and chemokine receptor antagonism as therapeutic strategies[55,56]. Cross-platform validation through Yang et al. multi-tissue analysis of 529 significant proteins (FDR \u0026lt; 0.05) extensively confirmed our findings, revealing complement system activation (C2, C5, C9)[27], neuroinflammation markers (TNF, GFAP)[58,59], and neuroprotective factors (GDNF, NGF)[60], establishing robust biological validation for our analytical framework.\u003c/p\u003e\n\u003cp\u003eLarge-scale phenome-wide association analysis of 50,033 traits identified 17 phenotypes with significant causal associations[61,62], significantly expanding our understanding of neuroimmune syndrome risk factors. Thyroid disorders emerged as the strongest risk factor, representing nearly two-fold increased disease risk, followed by self-reported hypothyroidismand rheumatoid arthritis. The predominantly risk-increasing pattern (16 of 17 significant associations) supports shared pathophysiological mechanisms across autoimmune and endocrine systems, suggesting that neuroimmune syndromes represent manifestations of broader systemic dysregulation[63]. These validated associations provide crucial insights for clinical risk stratification and therapeutic target identification, establishing robust causal evidence for precision medicine strategies[64].\u003c/p\u003e\n\u003cp\u003eThe convergence of our multi-tiered findings establishes a comprehensive framework for understanding neuroimmune syndrome pathophysiology. The identification of concentrated chromosomal architecture (Chr4/Chr14), regulatory enrichment patterns, metabolic factors (ketone body metabolism), immune checkpoints (CTLA4), signaling pathways (TYK2, JAK), and chemokine networks (CCR5) alongside validated disease associations (thyroid-autoimmune axis) provides multiple therapeutic entry points spanning immune modulation, metabolic intervention, and transcriptional regulation. This integrated approach reveals that effective neuroimmune syndrome treatment likely requires combination strategies addressing the complex interplay between genetic predisposition, metabolic dysfunction, and immune dysregulation rather than targeting individual pathways in isolation[65,66]. The progression from unbiased discovery through targeted validation to biological confirmation and clinical relevance assessment provides robust evidence for precision medicine strategies, with particular strength in complement-targeted therapies, immune checkpoint modulation, and metabolic interventions[67].\u003c/p\u003e\n\u003cp\u003eSeveral limitations require consideration in interpreting these findings. First, our analysis primarily utilized European ancestry populations from FinnGen, limiting generalizability across diverse populations. Future studies should expand to Asian, African, and Native American populations to ensure cross-ancestry validity[68,69,70]. Second, while we identified concentrated genetic architecture on chromosomes 4 and 14, the specific biological mechanisms underlying this concentration require detailed functional investigation through experimental validation. Third, although our transcriptomic analysis revealed 78 candidate genes with bidirectional regulatory patterns, connecting these discoveries to specific molecular pathways necessitates laboratory-based functional studies. Additionally, environmental factors likely interact with genetic susceptibility in neuroimmune syndrome development, necessitating future gene-environment interaction studies to fully understand disease etiology[71]. The complex bidirectional regulatory patterns observed in our candidate genes suggest that therapeutic interventions must carefully balance activation and suppression mechanisms to avoid unintended consequences[72].\u003c/p\u003e\n\u003cp\u003eOur findings provide several immediate translational opportunities that warrant clinical investigation. The identification of CTLA4, TYK2, and CCR5 as protective targets offers genetic validation for existing therapeutic approaches including immune checkpoint inhibitors, JAK inhibitors, and chemokine receptor antagonists, supporting drug repurposing strategies in neuroimmune diseases. The strong thyroid-neuroimmune association suggests that screening protocols for thyroid dysfunction in patients with neuroimmune manifestations may improve clinical outcomes and enable earlier intervention. The concentrated chromosomal architecture indicates that genetic risk scores incorporating chromosome 4 and 14 variants may provide more accurate risk prediction than traditional genome-wide approaches, potentially enabling personalized prevention strategies. Finally, the regulatory enrichment patterns suggest that future therapeutic development should prioritize epigenetic modulators and transcriptional interventions rather than traditional protein-targeting approaches, potentially opening new avenues for precision medicine in neuroimmune diseases. The metabolic component revealed through BDH2 associations suggests that nutritional interventions targeting ketone body metabolism warrant investigation as adjunctive therapies in neuroimmune syndrome management.\u003c/p\u003e"},{"header":"Conclusions","content":"\u003cp\u003eThis investigation represents the first comprehensive genetic architecture mapping of neuroimmune syndromes through genomic structural equation modeling, revealing 33 genome-wide significant loci and 78 candidate pathogenic genes with shared genetic susceptibility across five neuroimmune-related disorders. Our findings demonstrate a remarkably concentrated genetic architecture dominated by chromosomes 4 and 14, which together account for 62.9% of total genetic variance, challenging traditional polygenic models and indicating non-HLA-driven mechanisms as primary contributors to disease susceptibility.\u003c/p\u003e\n\u003cp\u003eThe identification of validated therapeutic targets including CTLA4, TYK2, and CCR5, along with novel metabolic pathways involving BDH2, provides genetic evidence for precision medicine strategies encompassing immune checkpoint modulation, JAK inhibition, and metabolic interventions. The strong causal association with thyroid disorders establishes important clinical risk stratification opportunities, while the regulatory enrichment patterns suggest that epigenetic modulators may prove more effective than traditional protein-targeting approaches.\u003c/p\u003e\n\u003cp\u003eFuture research should expand to diverse populations, conduct functional validation of the concentrated chromosomal architecture, and investigate gene-environment interactions to fully understand neuroimmune syndrome etiology and optimize personalized therapeutic interventions.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe data that support the findings of this study are available in FinnGen Research Program at https://www.finngen.fi/en, reference number Release 12. These data were derived from the following resources available in the public domain: - FinnGen GWAS summary statistics for neuroimmune disorders, https://www.finngen.fi/en/access_results - eQTLGen Consortium, https://www.eqtlgen.org/ - IEU OpenGWAS database, https://gwas.mrcieu.ac.uk/ - Yang et al. brain, CSF, and plasma pQTL, https://doi.org/10.1038/s41593-021-00886-6 - deCODE genetics consortium, https://doi.org/10.1038/s41588-021-00978-w. The neuroimmune syndrome structural equation modeling summary statistics generated in this study are available from the corresponding author upon reasonable request.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflicts of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that there are no conflicts of interest regarding the publication of this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgments\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors thank the FinnGen Research Program and all participants who contributed to the FinnGen study. We acknowledge the eQTLGen Consortium for providing eQTL data, the deCODE genetics consortium for plasma protein data, and the IEU OpenGWAS database for phenome-wide association data. We also thank the research teams who generated and made publicly available the GWAS summary statistics used in this study. The authors appreciate the technical support and computational resources that made this analysis possible.\u003c/p\u003e\n\u003cp\u003eDeclaration of Contributions and AI Usage: All authors contributed significantly to this work and agree to be accountable for all aspects of the research content and conclusions. No third-party services or individuals not listed as authors were involved in the research or manuscript preparation. No artificial intelligence software was used in any aspect of manuscript preparation, including data collection, analysis, writing, or editing.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eHaifeng Chen\u0026sup1;\u0026dagger; and Yuxiong Liao\u0026sup1;\u0026dagger; contributed equally to this work as co-first authors and were responsible for conceptualization, methodology, formal analysis, investigation, data curation, and writing the original draft. Luejun Tang\u0026sup1; contributed to data curation, formal analysis, and validation. Xiaoyun Wei\u0026sup1; contributed to methodology, software implementation, and validation. Tongshun Li\u0026sup1; provided resources, assisted with data curation, and validation. Wei Chen\u0026sup1;* (corresponding author) was responsible for conceptualization, methodology, supervision, project administration, and final manuscript approval. All authors participated in writing review and editing and agree to be accountable for all aspects of the work content and conclusions.\u003c/p\u003e\n\u003cp\u003eEthical Approval This study utilized publicly available genome-wide association study (GWAS) summary statistics from the FinnGen Research Program Release 12 and other established genomic consortia. All constituent GWAS studies included in this analysis had received institutional review board approval from their respective institutions, with informed consent obtained from all participants. As this investigation employed only summary-level genetic data without access to individual participant information, additional ethical approval was not required for this secondary analysis.\u003c/p\u003e\n\u003cp\u003eFunding This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003eAvailability of data and materials The data that support the findings of this study are available from publicly accessible genomic databases. FinnGen GWAS summary statistics are available at https://www.finngen.fi/en/access_results (Release 12). Additional datasets used include: eQTLGen Consortium data (https://www.eqtlgen.org/), IEU OpenGWAS database (https://gwas.mrcieu.ac.uk/), Yang et al. brain, cerebrospinal fluid, and plasma protein quantitative trait loci data (https://doi.org/10.1038/s41593-021-00886-6), and deCODE genetics consortium plasma protein data (https://doi.org/10.1038/s41588-021-00978-w). The neuroimmune syndrome structural equation modeling summary statistics generated in this study are available from the corresponding author upon reasonable request.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eEngelhardt B, Vajkoczy P, Weller RO. The movers and shapers in immune privilege of the CNS. Nat Immunol. 2017 Feb;18(2):123-131.\u003c/li\u003e\n\u003cli\u003ePrinz M, Jung S, Priller J. Microglia biology: one century of evolving concepts. Cell. 2019;179(2):292-311.\u003c/li\u003e\n\u003cli\u003eKipnis J. Multifaceted interactions between adaptive immunity and the central nervous system. Science. 2016;353(6301):766-771.\u003c/li\u003e\n\u003cli\u003eAbsinta M, Lassmann H, Trapp BD. Mechanisms underlying progression in multiple sclerosis. Nat Rev Neurol. 2020;16(11):657‑668.\u003c/li\u003e\n\u003cli\u003eKuchroo VK, Ohashi PS, Sartor RB, Vinuesa CG. Dysregulation of immune homeostasis in autoimmune diseases. Nat Med. 2012;18(1):42‑47. \u003c/li\u003e\n\u003cli\u003eBach JF. The hygiene hypothesis in autoimmunity: the role of pathogens and commensals. Nat Rev Immunol. 2018;18(2):105‑120.\u003c/li\u003e\n\u003cli\u003eKim A, Xie F, Abed OA, Moon JJ. Vaccines for immune tolerance against autoimmune disease. Adv Drug Deliv Rev. 2023 Dec;203:115-140. \u003c/li\u003e\n\u003cli\u003eReich DS, Lucchinetti CF, Calabresi PA. Multiple sclerosis. N Engl J Med. 2018;378(2):169‑180.\u003c/li\u003e\n\u003cli\u003eGoverman JM. Autoimmune T cell responses in the central nervous system. Nat Rev Immunol. 2021;21(9):560‑573.\u003c/li\u003e\n\u003cli\u003eAshley EA. Towards precision medicine. Nat Rev Genet. 2016;17(9):507-522.\u003c/li\u003e\n\u003cli\u003eGrotzinger AD, Rhemtulla M, de Vlaming R, et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. Nat Hum Behav. 2019;3(5):513-525. \u003c/li\u003e\n\u003cli\u003eKurki MI, Karjalainen J, Palta P, et al. FinnGen provides genetic insights from a well-phenotyped isolated population. Nature. 2023;613(7944):508-518.\u003c/li\u003e\n\u003cli\u003eTian C, Gregersen PK, Seldin MF. Accounting for ancestry: population substructure and genome-wide association studies. Hum Mol Genet. 2008;17(R2):R143-R150.\u003c/li\u003e\n\u003cli\u003eTurley P, Walters RK, Maghzian O, et al. Multi-trait analysis of genome-wide association summary statistics using MTAG. Nat Genet. 2018;50(2):229-237.\u003c/li\u003e\n\u003cli\u003eBulik-Sullivan B, Finucane HK, Anttila V, et al. An atlas of genetic correlations across human diseases and traits. Nat Genet. 2015;47(11):1236-1241.\u003c/li\u003e\n\u003cli\u003eWatanabe K, Taskesen E, van Bochoven A, Posthuma D. Functional mapping and annotation of genetic associations with FUMA. Nat Commun. 2017;8(1):1826.\u003c/li\u003e\n\u003cli\u003eSollis E, Mosaku A, Abid A, et al. The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. Nucleic Acids Res. 2023;51(D1):D977-D985.\u003c/li\u003e\n\u003cli\u003ede Leeuw CA, Mooij JM, Heskes T, Posthuma D. MAGMA: generalized gene-set analysis of GWAS data. PLoS Comput Biol. 2015;11(4):e1004219.\u003c/li\u003e\n\u003cli\u003eGusev A, Ko A, Shi H, et al. Integrative approaches for large-scale transcriptome-wide association studies. Nat Genet. 2016;48(3):245-252.\u003c/li\u003e\n\u003cli\u003eGTEx Consortium. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. 2020 Sep 11;369(6509):1318-1330.\u003c/li\u003e\n\u003cli\u003eMancuso N, Freund MK, Johnson R, et al. Probabilistic fine-mapping of transcriptome-wide association studies. Nat Genet. 2019;51(4):675-682.\u003c/li\u003e\n\u003cli\u003eLiberzon A, Birger C, Thorvaldsd\u0026oacute;ttir H, Ghandi M, Mesirov JP, Tamayo P. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 2015;1(6):417-425.\u003c/li\u003e\n\u003cli\u003eTabula Muris Consortium, Overall coordination, Logistical coordination.Single-cell transcriptomics of 20 mouse organs creates a Tabula Muris. Nature. 2018;562(7727):367-372.\u003c/li\u003e\n\u003cli\u003eFinucane HK, Bulik-Sullivan B, Gusev A, et al. Partitioning heritability by functional annotation using genome-wide association summary statistics. Nat Genet. 2015;47(11):1228-1235.\u003c/li\u003e\n\u003cli\u003eV\u0026otilde;sa U, Claringbould A, Westra HJ, et al. Large-scale cis- and trans-eQTL analyses identify thousands of genetic loci and polygenic scores that regulate blood gene expression. Nat Genet. 2021;53(9):1300-1310.\u003c/li\u003e\n\u003cli\u003eFerkingstad E, Sulem P, Atlason BA, et al. Large-scale integration of the plasma proteome with genetics and disease. Nat Genet. 2021;53(12):1712-1721.\u003c/li\u003e\n\u003cli\u003eYang C, Farias FHG, Ibanez L, et al. Genomic atlas of the proteome from brain, CSF and plasma prioritizes proteins implicated in neurological disorders. Nat Neurosci. 2021;24(9):1302-1312.\u003c/li\u003e\n\u003cli\u003eHemani G, Zheng J, Elsworth B, et al. The MR-Base platform supports systematic causal inference across the human phenome. eLife. 2018;7:e34408.\u003c/li\u003e\n\u003cli\u003eLambert SA, Gil L, Jupp S, et al. The Polygenic Score Catalog as an open database for reproducibility and systematic evaluation. Nat Genet. 2021;53(4):420-425.\u003c/li\u003e\n\u003cli\u003eGe T, Chen CY, Ni Y, Feng YCA, Smoller JW. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nat Commun. 2019;10(1):1776.\u003c/li\u003e\n\u003cli\u003eRansohoff RM, Engelhardt B. The anatomical and cellular basis of immune surveillance in the central nervous system. Nat Rev Immunol. 2012;12(9):623-635.\u003c/li\u003e\n\u003cli\u003eMucida D, Husain MM, Muroi S, et al. Transcriptional reprogramming of mature CD4+ helper T cells generates distinct MHC class II-restricted cytotoxic T lymphocytes. Nat Immunol. 2013;14(3):281-289.\u003c/li\u003e\n\u003cli\u003eDendrou CA, Fugger L, Friese MA. Immunopathology of multiple sclerosis. Nat Rev Immunol. 2015;15(9):545-558.\u003c/li\u003e\n\u003cli\u003eOkada Y, Wu D, Trynka G, et al. Genetics of rheumatoid arthritis contributes to biology and drug discovery. Nature. 2014;506(7488):376-381.\u003c/li\u003e\n\u003cli\u003eEllinghaus D, Jostins L, Spain SL, et al. Analysis of five chronic inflammatory diseases identifies 27 new associations and highlights disease-specific patterns. Nat Genet. 2016;48(5):510-518.\u003c/li\u003e\n\u003cli\u003eCross-Disorder Group of the Psychiatric Genomics Consortium. Identification of risk loci with shared effects on five major psychiatric disorders. Lancet. 2013;381(9875):1371-1379.\u003c/li\u003e\n\u003cli\u003eCotsapas C, Voight BF, Rossin E, et al. Pervasive sharing of genetic effects in autoimmune disease. PLoS Genet. 2011;7(8):e1002254.\u003c/li\u003e\n\u003cli\u003eJavierre BM, Burren OS, Wilder SP, et al. Lineage-specific genome organization links enhancers and non-coding disease variants to target gene promoters. Cell. 2016;167(5):1369-1384.\u003c/li\u003e\n\u003cli\u003eFarh KK, Marson A, Zhu J, et al. Genetic and epigenetic fine mapping of causal autoimmune disease variants. Nature. 2015;518(7539):337-343.\u003c/li\u003e\n\u003cli\u003eRoadmap Epigenomics Consortium. Integrative analysis of 111 reference human epigenomes. Nature. 2015;518(7539):317-330.\u003c/li\u003e\n\u003cli\u003eErnst J, Kellis M. ChromHMM: automating chromatin-state discovery and characterization. Nat Methods. 2012;9(3):215-216.\u003c/li\u003e\n\u003cli\u003eBoyle EA, Li YI, Pritchard JK. An expanded view of complex traits: from polygenic to omnigenic. Cell. 2017;169(7):1177-1186.\u003c/li\u003e\n\u003cli\u003eENCODE Project Consortium. An integrated encyclopedia of DNA elements in the human genome. Nature. 2012;489(7414):57-74.\u003c/li\u003e\n\u003cli\u003eSoskic B, Cano-Gamez E, Smyth DJ, et al. Chromatin activity at GWAS loci identifies T cell states driving complex immune diseases. Nat Genet. 2019;51(10):1486-1493.\u003c/li\u003e\n\u003cli\u003eLyons JJ, Yu X, Hughes JD, et al. Elevated basal serum tryptase identifies a multisystem disorder associated with increased TPSAB1 copy number. Nat Genet. 2016;48(12):1564-1569.\u003c/li\u003e\n\u003cli\u003eHormozdiari F, van de Bunt M, Segr\u0026egrave; AV, et al. Colocalization of GWAS and eQTL signals detects target genes. Am J Hum Genet. 2016;99(6):1245-1260.\u003c/li\u003e\n\u003cli\u003eWallace C. Eliciting priors and relaxing the single causal variant assumption in colocalisation analyses. PLoS Genet. 2020;16(4):e1008720.\u003c/li\u003e\n\u003cli\u003eGiambartolomei C, Vukcevic D, Schadt EE, et al. Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. 2014;10(5):e1004383.\u003c/li\u003e\n\u003cli\u003eChoi SW, Mak TS, O\u0026apos;Reilly PF. Tutorial: a guide to performing polygenic risk score analyses. Nat Protoc. 2020;15(9):2759-2772.\u003c/li\u003e\n\u003cli\u003eMars N, Kerminen S, Feng YCA, et al. Polygenic and clinical risk scores and their impact on age at onset and prediction of cardiometabolic diseases and common cancers. Nat Med. 2020;26(4):549-557.\u003c/li\u003e\n\u003cli\u003eVilhj\u0026aacute;lmsson BJ, Yang J, Finucane HK, et al. Modeling linkage disequilibrium increases accuracy of polygenic risk scores. Am J Hum Genet. 2015;97(4):576-592.\u003c/li\u003e\n\u003cli\u003eSchmidt AF, Finan C, Gordillo-Mara\u0026ntilde;\u0026oacute;n M, et al. Genetic drug target validation using Mendelian randomisation. Nat Commun. 2020;11(1):3255.\u003c/li\u003e\n\u003cli\u003eBurgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol. 2013;37(7):658-665.\u003c/li\u003e\n\u003cli\u003eBowden J, Davey Smith G, Haycock PC, Burgess S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet Epidemiol. 2016;40(4):304-314.\u003c/li\u003e\n\u003cli\u003eVerbanck M, Chen CY, Neale B, Do R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat Genet. 2018;50(5):693-698.\u003c/li\u003e\n\u003cli\u003eNelson MR, Tipney H, Painter JL, et al. The support of human genetic evidence for approved drug indications. Nat Genet. 2015;47(8):856-860.\u003c/li\u003e\n\u003cli\u003eRicklin D, Hajishengallis G, Yang K, Lambris JD. Complement: a key system for immune surveillance and homeostasis. Nat Immunol. 2010;11(9):785-797.\u003c/li\u003e\n\u003cli\u003eKalliolias GD, Ivashkiv LB. TNF biology, pathogenic mechanisms and emerging therapeutic strategies. Nat Rev Rheumatol. 2016;12(1):49-62.\u003c/li\u003e\n\u003cli\u003eAbdelhak A, Huss A, Kassubek J, et al. Serum GFAP as a biomarker for disease severity in multiple sclerosis. Sci Rep. 2018;8(1):14798.\u003c/li\u003e\n\u003cli\u003eKerschensteiner M, Gallmeier E, Behrens L, et al. Activated human T cells, B cells, and monocytes produce brain-derived neurotrophic factor. Ann Neurol. 1999;46(1):90-100.\u003c/li\u003e\n\u003cli\u003ePendergrass SA, Browning BL, Dudek SM, et al. Phenome-wide association study (PheWAS) for detection of pleiotropy within the Population Architecture using Genomics and Epidemiology (PAGE) Network. PLoS Genet. 2013;9(1):e1003087.\u003c/li\u003e\n\u003cli\u003eDenny JC, Ritchie MD, Basford MA, et al. PheWAS: demonstrating the feasibility of a phenome-wide scan to discover gene-disease associations. Bioinformatics. 2010;26(9):1205-1210.\u003c/li\u003e\n\u003cli\u003eMedici M, Visser WE, Visser TJ, Peeters RP. Genetic determination of the hypothalamic-pituitary-thyroid axis: where do we stand? Endocr Rev. 2015;36(2):214-244.\u003c/li\u003e\n\u003cli\u003eGusev A, Mancuso N, Won H, et al. Transcriptome-wide association study of schizophrenia and chromatin activity yields mechanistic disease insights. Nat Genet. 2018;50(4):538-548.\u003c/li\u003e\n\u003cli\u003eZhu Z, Zhang F, Hu H, et al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nat Genet. 2016;48(5):481-487.\u003c/li\u003e\n\u003cli\u003eHauberg ME, Zhang W, Giambartolomei C, et al. Large-scale identification of common trait and disease variants affecting gene expression. Am J Hum Genet. 2017;101(1):157-173.\u003c/li\u003e\n\u003cli\u003eWu Y, Zeng J, Zhang F, et al. Integrative analysis of omics summary data reveals putative mechanisms underlying complex traits. Nat Commun. 2018;9(1):918.\u003c/li\u003e\n\u003cli\u003ePopejoy AB, Fullerton SM. Genomics is failing on diversity. Nature. 2016;538(7624):161-164.\u003c/li\u003e\n\u003cli\u003eMartin AR, Kanai M, Kamatani Y, et al. Clinical use of current polygenic risk scores may exacerbate health disparities. Nat Genet. 2019;51(4):584-591.\u003c/li\u003e\n\u003cli\u003ePeterson RE, Kuchenbaecker K, Walters RK, et al. Genome-wide association studies in ancestrally diverse populations: opportunities, methods, pitfalls, and recommendations. Cell. 2019;179(3):589-603.\u003c/li\u003e\n\u003cli\u003eDuncan L, Shen H, Gelaye B, et al. Analysis of polygenic risk score usage and performance in diverse human populations. Nat Commun. 2019;10(1):3328.\u003c/li\u003e\n\u003cli\u003eHunter DJ. Gene-environment interactions in human diseases. Nat Rev Genet. 2005;6(4):287-298.\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003e\u003cstrong\u003eTable 1. Parameters and Quality Metrics for Genomic Structural Equation Modeling.\u003c/strong\u003e\u003c/p\u003e\n \u003ctable border=\"1\" cellspacing=\"0\" cellpadding=\"0\" width=\"99%\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003ePhenotype\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 10px;\"\u003e\n \u003cp\u003eNSNPs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003eh2_se\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e\u0026lambda;GC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 18px;\"\u003e\n \u003cp\u003eMean_ChiSquare\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003eIntercept_se\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003eRatio_se\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003eGBS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 10px;\"\u003e\n \u003cp\u003e1159723\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.0013 (0.0013)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e1.0132\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 18px;\"\u003e\n \u003cp\u003e1.0055\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.9927 (0.0067)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e1.3374 (1.2246)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003eMG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 10px;\"\u003e\n \u003cp\u003e1159735\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.0019 (0.001)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e1.0216\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 18px;\"\u003e\n \u003cp\u003e1.0157\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.9975 (0.0068)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.1612 (0.435)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003eMS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 10px;\"\u003e\n \u003cp\u003e1159738\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.0084 (0.0013)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e1.089\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 18px;\"\u003e\n \u003cp\u003e1.113\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e1.0287 (0.0084)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.254 (0.0745)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003eSLE\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 10px;\"\u003e\n \u003cp\u003e1159730\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.0011 (0.001)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e1.0072\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 18px;\"\u003e\n \u003cp\u003e1.0036\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.9927 (0.0077)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e2.0275 (2.1232)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 12px;\"\u003e\n \u003cp\u003eSCTD\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 10px;\"\u003e\n \u003cp\u003e1159740\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.0085 (0.0017)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 8px;\"\u003e\n \u003cp\u003e1.1112\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 18px;\"\u003e\n \u003cp\u003e1.1285\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e1.0385 (0.0095)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 16px;\"\u003e\n \u003cp\u003e0.2994 (0.0737)\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u003cstrong\u003eNotes:\u003c/strong\u003e h2_se represents SNP-based heritability estimates with standard errors in parentheses. \u0026lambda;GC indicates genomic control lambda for population stratification assessment. LDSC intercept values assess confounding due to population stratification and cryptic relatedness.\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":"genomic structural equation modeling, neuroimmune syndromes, genome-wide association study, transcriptome-wide association study","lastPublishedDoi":"10.21203/rs.3.rs-7328766/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7328766/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The genetic architecture underlying neuroimmune syndrome-related traits remains poorly characterized. We employed genomic structural equation modeling (genomic SEM) alongside comprehensive post-GWAS analytical frameworks to identify causal single nucleotide polymorphisms (SNPs) associated with phenotypic variance independent of measured traits, revealing 33 genome-wide significant loci. Multi-tissue transcriptome-wide association analyses were conducted across tissue, cellular, and genomic regulatory elements to characterize susceptibility gene signals and regulatory components with high relevance to neuroimmune syndrome GWAS. We subsequently leveraged extensive human disease datasets to determine neuroimmune syndrome-associated risk factors and explored potential therapeutic targets through plasma proteomics and drug target prioritization analyses. Additionally, summary statistics-based polygenic scoring assessed chromosomal contributions to neuroimmune syndrome risk. Our investigation represents the first comprehensive genetic architecture mapping of neuroimmune syndromes through GWAS of a latent, unmeasured phenotype, providing unprecedented insights into shared genetic mechanisms underlying these complex disorders.","manuscriptTitle":"Cross-disorder genomic structural equation modeling reveals common genetic basis of neuroimmune diseases","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-22 11:32:22","doi":"10.21203/rs.3.rs-7328766/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":"5e3f202e-41ba-4d04-9786-3107ea62b24c","owner":[],"postedDate":"August 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":53270094,"name":"Biological sciences/Computational biology and bioinformatics"},{"id":53270095,"name":"Health sciences/Diseases"},{"id":53270096,"name":"Biological sciences/Genetics"},{"id":53270097,"name":"Health sciences/Neurology"},{"id":53270098,"name":"Biological sciences/Neuroscience"}],"tags":[],"updatedAt":"2025-09-07T09:38:28+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-22 11:32:22","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7328766","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7328766","identity":"rs-7328766","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}