Population-Specific Structural Variant Landscape in a Puerto Rican Rare Disease Cohort

Population-Specific Structural Variant Landscape in a Puerto Rican Rare Disease Cohort | 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 Research Article Population-Specific Structural Variant Landscape in a Puerto Rican Rare Disease Cohort Cas LeMaster, Emily Farrow, Rebecca McLennan, Elinette Albino, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8544782/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 4 You are reading this latest preprint version Abstract The integration of long-read PacBio High-Fidelity (HiFi) sequencing with the complete Telomere-to-Telomere CHM13 (T2T-CHM13) reference genome has enabled thorough characterization of structural variants (SVs) in previously inaccessible genomic regions, yet Puerto Rican and broader admixed populations remain critically underrepresented in these advances. We performed HiFi whole genome sequencing on 90 samples across 30 parent-proband trios in the Genomic Answers for Kids (GA4K) program (15 European, 15 Puerto Rican) aligned to T2T-CHM13, identifying 1,729,471 deletions, 18,805 duplications, 1,203,260 insertions, and 2,872 inversions with stringent filtering. Puerto Rican individuals carried significantly more SVs, with enrichment in centromeric/pericentromeric and telomeric regions. SV genotypes provided strong ancestry discrimination (72.3% total variance by MDS vs 8.6% for SNVs), and ancestry-associated SVs were predominantly Puerto Rican for deletions and duplications. Functionally, Puerto Rican-enriched SVs intersected constrained and dosage-sensitive genes, including recurrent UTR and coding events with plausible regulatory or dosage effects. Together, these findings demonstrate that structural variants exhibit significant population-specific distributions and underscore the importance of combining complete reference genomes with long-read sequencing for ancestry-considerate interpretation. structural variants Puerto Rican genomics T2T HiFi long-read health disparities ancestry-specific Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 BACKGROUND The integration of long-read PacBio High-Fidelity (HiFi) sequencing with the complete Telomere-to-Telomere CHM13 (T2T-CHM13) reference genome has transformed structural variant (SV) detection by resolving previously inaccessible genomic regions. T2T-CHM13 adds nearly 200 million base pairs while resolving segmental duplications, satellites, and other repeats that confound short reads, universally improving variant calling across diverse populations 1 – 3 . This complete reference enables a 21% increase in inversion detection sensitivity and reduces false positives in medically relevant genes by up to 12-fold compared to GRCh38 4,5 . Clinical applications demonstrate significantly enhanced SV detection in complex genomic regions, directly improving diagnostic outcomes 6 . To date, genomic studies have largely focused on European-centric populations while other genetic backgrounds remain underrepresented, particularly those examining SVs 7 – 9 . Recent population analyses identified Puerto Ricans among the most underrepresented founder populations in genomic databases, creating significant disparities for variant interpretation 9 , 10 . This underrepresentation manifests clinically, as Caribbean Hispanic populations, including Puerto Ricans, show distinct pharmacogenomic profiles that differ substantially from European ancestry references, highlighting population-specific genetic burdens 11 , 12 . Thus, the exclusion of these populations from large-scale genomic initiatives perpetuates health disparities and limits precision medicine interventions in these communities 13 . The clinical implications of T2T‑CHM13–enabled SV discovery extend beyond a single diease area and are especially relevant for ancestry-informed interpretation. Pericentromeres, centromeres, telomeres, and segmental duplications harbor ancestry-specific repeats and mobile elements, accumulating SVs that vary in location and type across populations–regions now resolvable with a complete reference and accurate long reads 2 , 3 , 14 . Here, we used HiFi long-read sequencing aligned to the T2T reference to quantify ancestry-specific differences in SV location and type and characterized their potential for disease impacts in both Puerto Rican and European parent-proband trios. We show that Puerto Rican-specific SVs are enriched for intersecting high impact gene regions, including constrained and dosage-sensitive genes, and that complex clinical case interpretations—such as compound UTR disruptions, overlapping CDS/UTR intervals, and multi-locus combinations—require SV analysis that traditional approaches cannot reliably resolve. METHODS Study Cohort The study cohort includes 30 affected probands from 30 unrelated families (n = 90). Participants from Puerto Rico were consented and enrolled at the University of Puerto Rico Medical Science Campus (IRB Study #2305114422). Control subjects were enrolled in the Genomics Answers for Kids (GA4K) program. Eligibility criteria included a suspected genetic diagnosis based on clinical presentation and/or existing molecular or cytogenic findings. Providers introduced the study and asked if families were interested in participation. The study complies with all relevant ethical regulations as approved by both the University of Puerto Rico School of Medicine and Children's Mercy Institutional Review Board (IRB) (Study #11120514). Informed written consent was obtained from all participants prior to enrollment. Participants did not receive financial compensation for participation. Sequencing and Alignment PacBio HiFi raw sequencing reads were aligned to the CHM13v2.0 reference genome using pbmm2 (PacBio minimap2) with the CCS preset. Multiple BAM files per sample were merged using samtools merge when necessary. Alignment quality was assessed using samtools flagstat for mapping statistics and samtools depth for coverage analysis. BAM files were indexed using both samtools index and pbindex for downstream compatibility. Structural Variant Discovery and Genotyping with Local Haplotype Modeling Structural variants (SVs) were identified using Sawfish (v0.12.9), which performs local haplotype modeling for accurate breakpoint resolution and genotyping in complex regions. SV discovery was performed per sample using Sawfish discover, utilizing all genomic regions (coverage regex ".") to identify clusters of SV alignment signatures. Individual sample genotyping was performed using the default Sawfish joint-call option, which models and genotypes all SV breakpoints using local haplotypes. Trio joint-calling was implemented using Sawfish's multi-sample joint-call functionality to improve variant calling accuracy through family-based genotyping and Mendelian consistency. Discovery results from all three trio members (proband, mother, and father) were jointly analyzed using sawfish joint-call. This approach consolidates duplicate SV haplotypes across samples and performs genotyping in the context of overlapping haplotype pools. Trio VCFs were merged using samtools, and SVs were filtered for quality > 30 (Phred-scaled, 99.9% confidence), depth > 10 (to ensure adequate coverage for confident genotyping), and minimum length > 50bp (standard SV size threshold). Additionally, we required SVs to pass Sawfish's internal FILTER field to exclude low-quality calls. Single Nucleotide Variant Calling Small variants and single nucleotide variants (SNVs) were called using Clair3 with the HiFi Revio model, optimized for PacBio HiFi sequencing data. Clair3 analysis was performed using the "hifi" preset to maximize calling accuracy for HiFi reads. SNVs were filtered for QUAL > 20 and GQ > 20 to ensure high-confidence calls. Ancestry PCAs Ancestry analysis was performed using both SNVs and SVs with parallel analytical frameworks to enable comparison. The analysis included 93 individuals from 31 families across three populations: European (EUR, n = 45), Puerto Rican (PUR, n = 45), and African (AFR, n = 3). Each family consisted of a trio (proband, mother, and father). For SNV analysis, we used Somalier to identify 2,515 ancestry-informative SNPs and calculate pairwise relatedness between samples using Somalier's internal algorithms. The resulting relatedness values were converted to distance matrices (1 - relatedness) and classical multidimensional scaling (MDS) was applied using R's cmdscale function for dimensionality reduction. For SV analysis, we extracted 4,971 high-quality SVs (allele frequency 0.01–0.99, size ≥ 50bp, missingness ≤ 10% per variant) from joint-called VCF files with SV type-specific length thresholds. Pairwise relatedness was calculated from genotype correlations and normalized to range 0–1. Classical MDS was implemented in Python with double centering of squared distance matrices followed by eigen decomposition to extract coordinates. Genes, Regions, Constraint Mapping, and Dosage Sensitivity Variants were annotated with intersecting genes via bedtools intersect and systematically collapsed on unique variant IDs. Gene annotations were obtained from the Telomere-to-Telomere Consortium GitHub under T2T-CHM13v2.0 (T2T-CHM13 + Y) via the HPRC EBI GENCODEv38 r2 gff3 file 15 . Regional classes and constraint metrics were assigned to SVs using the following hierarchy: (1) for SVs intersecting multiple genes, we used the mean constraint score among all intersected genes; (2) for SVs intersecting multiple genomic regions of the same gene, we prioritized CDS > UTR > intronic overlaps; (3) categorical haploinsufficiency data was obtained from the Clinical Genome Resource (ClinGen). Constraint scores were provided by gnomADv4 resources using the pLOF z-score for canonical genes. Chromosomal regions were defined using the T2T-CHM13v2.0 censat file available via the UCSC genome browser. Regions p, q, and centromeric were defined via bed coordinates, telomeric regions were defined via 1Mb sections from the end of chromosome arms (with consideration for acrocentric chromosomes), and pericentromeric regions were 5Mb sections from the centromere end coordinate in respective directions to the arms they associated with. For gnomAD comparison, we downloaded the v4.1 SV call set and harmonized coordinates from GRCh38 to CHM13v2.0 using a chain-based liftover, retained unique SVs, and excluded centromeric/satellite/rDNA intervals on CHM13. When matching between the cohort and gnomAD, we used ≥ 50% reciprocal overlapping for DEL/DUP/INV and \(\:\pm\:\) 100 bp start tolerance for INS. Statistics Count comparisons across ancestries were tested for significance using Mann-Whitney U Tests. Unique structural variant comparisons across ancestries were made using Fisher’s Exact Test. RESULTS Structural Variant Discovery and Characterization in European and Puerto Rican Trios To investigate structural variation across the genome from different ancestry groups, we focused on 90 samples across 30 parent-child trios with 15 of European (EUR) and 15 of Puerto Rican (PUR) ancestry. Using HiFi long-read whole genome sequencing aligned to the T2T-CHM13v2.0 reference genome 2 , we achieved high coverage and depth across autosomes (mean coverage = 98.75%, mean depth = 31x) and sex chromosomes (mean coverage = 95.6%, mean depth = 22.9x; Supplemental Table 1 ; Supplemental Fig. 1A—B ). This approach systematic downstream variant detection, enabling the identification of 1,729,471 deletions (DELs), 18,805 duplications (DUPs), 1,203,260 insertions (INSs), and 2,872 inversions (INVs) across the cohort, with an average of 32,827 SVs per individual (Fig. 1 A; Supplemental Table 1 ). These counts align well with previous long-read sequencing studies 16 – 18 , with one trio (EUR4) showing comparatively lower counts while maintaining high coverage (95%) and depth (23.2x). Size distribution analysis revealed that DELs and INSs, while constituting the majority of variants, typically occupied smaller genomic regions with the highest density under 500 bp and notable peaks at 300 bp and 6 kb corresponding to Alu element and LINE-1 retrotransposon activities, respectively (Fig. 1 B) 24 – 26 . DUPs and INVs showed larger size distributions extending beyond 1 Mb, consistent with their complex formation mechanisms. Ancestry-Specific Distribution of Genomic Variation Reveals Enhanced Discriminatory Power of Structural Variants To understand population differences in SV counts and build on our comprehensive SV catalog, we investigated ancestry-specific patterns of genomic variation. Investigating these patterns across the genome, we constructed circos plots displaying mean variant densities for probands in each ancestry group (EUR n = 15, PUR n = 15, Fig. 2 A-B), visualizing chromosomal regions enriched for both SVs and single nucleotide variants (SNVs). While SNVs were not our primary focus, recent studies have demonstrated that different chromosomal regions exhibit varying susceptibility to SNV and SV accumulation due to repetitive sequences, recombination hotspots, and structural rearrangement (Fig. 2 A-B) 19 . This revealed large peaks around telomeric and centromeric regions for both variant classes that often paralleled lower coverage. To quantify ancestry-specific enrichment, we calculated log2 ratios (log2r) of variant counts between populations, highlighting regions with the largest relative differences (> 0.3 log2r) for each ancestry cohort (Fig. 2 A-B; Supplemental Table 1 ). We observed that the pericentromeric (pc) region chr20pc_q (log2r = 0.51; p = 0.0004), telomeric (t) region chr1t_p (log2r = 0.39; p = 0.0007), and the centromere (c) of chr7c (log2r = 0.34, p = 0.02) showed significantly higher structural variation in PUR individuals, particularly for DELs and INSs. In contrast, no regions were more associated with EUR ancestry ( Supplemental Table 1; Supplemental Fig. 2 ). These findings are consistent with recent work demonstrating that centromeric regions can harbor non-pathogenic population-specific satellite DNA variants and higher-order repeats that correlate with ancestry 20 , 21 . Intriguingly, SNV patterns showed inverse relationships in some regions: EUR-enriched SV regions at chr12c and chr4c harbored more SNVs in PUR individuals (log2r = 0.25 & 0.19, respectively), while PUR-enriched SV regions like chr1t_p showed higher SNV counts in EUR individuals (log2r = 0.15). Additional PUR-enriched SNV regions included chr21c centromere (log2r = 0.42) and telomeric regions chr18t_p, chr8t_p, and chr10t_p (log2r = 0.33, 0.32, & 0.32, respectively). Notably, all EUR regions with > 0.3 log2r SNVs were telomeric (chr19t_p, chr16t_p, chr12t_q; log2r = 0.55, 0.44, & 0.37, respectively), suggesting potential ancestry-specific telomeric patterns for SNVs ( Supplemental Table 1; Supplemental Figs. 3 and 4 ). Genome-wide variant quantification revealed higher overall variant counts in PUR individuals for both SNVs (PUR median = 891,820, EUR median = 830,008; p = 0.005) and SVs (PUR median = 32,923, EUR median = 31,379; p 2%) compared to SNVs suggests that structural variation may be particularly informative for ancestry inference. To directly test this hypothesis, we performed parallel dimensionality reduction analyses comparing the ancestry discrimination power of SNVs versus SVs using conceptually similar multidimensional scaling (MDS) approaches. We included an additional African (AFR) ancestry trio to enhance population stratification. For SV analysis, we extracted 4,971 high-quality structural variants (allele frequency 0.01–0.99, size ≥ 50bp, missingness ≤ 10%) ( Supplemental Table 1 ). We observed that SV-based analysis led to substantial discriminatory power, with the first two principal coordinates explaining 72.3% of total variance (PC1: 38.0%, PC2: 34.3%, n = 4,971) compared to 8.6% for SNV-based analysis (PC1: 5.0%, PC2: 3.6%, n = 2,515). While both approaches successfully separated the three ancestry groups (EUR, PUR, AFR), the SV-based method produced higher-resolution population clusters with greater separation (Fig. 2 D-E; Supplemental Table 1 ). Ancestry-Specific Structural Variant Type Distributions Reveal Predominant Puerto Rican Associations Having established the ancestry SV count differences, we next investigated the distribution of unique SVs by the significance of their association and their distribution by type and chromosome (Fig. 3 A-D). For DELs, we identified 275,063 variants across the cohort, of which 10,937 (4% of total DELs) showed significant ancestry associations (unique SVs possessing significantly higher carrier frequency in one population; Fisher’s Exact), with 56% linked to PUR and 44% to EUR ancestry ( Supplemental Table 1; Fig. 3 A). These ancestry-associated DELs were predominantly located on chromosomes 2, 9, and 22, consistent with our chromosomal distribution analysis ( Supplemental Fig. 2 ). Among the top 50 most significant ancestry-associated DELs, only one was found exclusively in EUR individuals, while 18 were PUR-specific, 6 of which were intersecting genes ( Supplemental Table 1; Fig. 3 A). Duplications, while less frequent overall (4,040 total), showed similar ancestry association rates (4% of total DUPs) but with an even stronger PUR bias (69% of ancestry-significant associations) ( Supplemental Table 1; Fig. 3 B). Chromosomes 17 and 20 harbored the majority of unique DUPs, with both showing predominant PUR associations. Among the top 50 ancestry-significant DUPs, 11 were EUR-specific while 22 were PUR-specific, and only one ancestry-shared DUP overlapped with a gene ( Supplemental Table 1; Fig. 3 B). Insertions also exhibited distinct patterns, with a lower proportion showing significant ancestry associations (3%), but nearly all chromosomes had PUR-enriched insertions except for chrY, which had the lowest overall count ( Supplemental Table 1; Fig. 3 C). Among the top 50 ancestry-significant INSs, 15 were PUR-specific while 35 were shared between populations. Notably, INSs showed the highest rate of gene intersection among all SV types: 35,262 of 122,751 unique INSs (29%) overlapped genes, compared to 16% for DELs, 12% for DUPs, and 22% for INVs ( Supplemental Table 1 ). This high INS rate could reflect the activity of mobile elements, particularly Alu and LINE-1 elements, which preferentially insert into open chromatin regions associated with active genes 22 . The 6,234 unique genes affected by INSs exceeded all other SV types (DEL: 5,792, DUP: 417, INV: 49). Lastly, inversions, while least frequent, followed similar patterns with a majority PUR-specific (Fig. 3 D). Overall, this suggests that structural variants show distinct population-specific patterns, with Puerto Rican ancestry exhibiting higher rates across all SV types. To validate our findings in a larger population context, we harmonized gnomAD v4 SVs to CHM13v2.0 and compared them with our cohort, acknowledging methodological differences that limit direct comparisons 23 . GnomAD employs short-read sequencing aligned to the incomplete GRCh38 reference, uses broad ancestry definitions (Admixed/AMR vs. our specific Puerto Rican samples), and utilizes GATK-SV calling, which is largely haplotype-agnostic by default; in contrast, our study uses HiFi long reads, T2T alignment and Sawfish, leveraging local haplotype modeling and trio-aware joint calling. Despite these limitations, we anticipated that general patterns should parallel our findings. We identified 847,913 unique SVs in gnomAD, with 125,699 showing significant ancestry associations (Admixed American/AMR: 101,721; non-Finnish European/NFE: 23,978; Supplemental Table 1 ). Consistent with our results, DELs were the most frequent SV type (72,130 total), with most associating with AMR ancestry (60,801). DUPs, INSs, and INVs all occurred at ~ 3.0-4.0x higher rates in AMR compared to NFE genomes (DUP: 3.35x; INS: 3.03x; INV: 4.03x), supporting our findings of increased SV counts in admixed populations. When matching variants between datasets, we identified 41,568 overlapping variants (~ 10.3% of our cohort) – comprising 26,019 DELs, 495 DUPs, 15,026 INSs, and 28 INVs ( Supplemental Fig. 5 ). Overall, across matched variants more were found between AMR and PUR than NFE and EUR, and there was a greater frequency SVs in AMR than NFE, paralleling our findings. Structural Variants intersecting Coding Regions Reveal Ancestry-Specific Patterns of Potential Gene Disruption With the observation that PUR individuals consistently exhibit higher variant counts than EUR individuals, we next assessed the differential potential impact of these variants on genes. Using GENCODE v38 r2 annotations from the Human Pangenome Reference Consortium (HPRC), we systematically evaluated SV overlaps with coding sequence (CDS), 5′UTR, 3’UTR, intronic, and intergenic regions (Fig. 4 ) 15 . Because the functional consequences of SVs depend strongly on their genomic context, we ranked intersecting regions by potential impact, prioritizing CDS overlaps as most disruptive to gene function and following descending impact with 5’UTR, 3’UTR, intronic, and intergenic, respectively. Across SV types, unique INSs had a higher occurrence in intronic and 3’ UTRs—consistent with retrotransposons or mobile element activity, while DELs disproportionately affected coding, 5’UTR, and intergenic regions. Focusing on the highest impact, DELs in coding regions represented the strongest potential for disruption (n = 388), followed by INSs (n = 148), DUPs (n = 48), and INVs (n = 2). This illuminates a potentially higher impact DEL burden in the Puerto Rican population. Gene Constraint and Dosage Sensitivity Reveal Functionally Intolerant SVs in Puerto Rican Probands To assess the functional severity of SV overlaps across coding and regulatory regions, we integrated gene-level constraint metrics. We evaluated functional ancestry-exclusive SV intersections (CDS, 5′UTR, 3′UTR) using gene-level loss-of-function (LoF) constraint z-scores from gnomAD v4 (Fig. 5 A, Supplemental Fig. 6 ). Constraint density peaked between LoF z = 1 & 2, with less tolerance across PUR-exclusive SVs (PUR mean z = 1.69, n = 384; EUR mean z = 1.55, n = 256). Positive mean z-scores for LoF constraint were seen across regional intersects for both populations, with PUR higher in both CDS and 5’ UTR regions and EUR higher in 3’ UTR (CDS mean PUR = 1.58, EUR = 1.27; 5′UTR mean PUR = 1.99, EUR = 1.76; 3′UTR mean PUR = 1.72, EUR = 1.93; Supplemental Fig. 6A-C ). Ancestry-agnostic SV density peaks between LoF z = 4 & 6 reflected enrichment at EXOC3 (18%), driven by complex DEL–INS events and clusters at INF2 , BZW1 , FLI1 , and XPR1 , none with established pathogenic relevance ( Supplemental Figure D ). High-constraint loci (z > 6) were predominately PUR-associated variants (60%). Gene-specific enrichments included DOCK8 (10/14 PUR; various intervals), ACAN (15/25 PUR; various intervals), KDM4B (14/23 PUR; chr19:5,130,725-5,130,875), KSR2 (6/8 PUR; chr12:117,452,891 − 117,452,947), and TTBK2 (4/4 EUR; chr15:40,608,293 − 40,608,450). Given that these high-constraint variants were uniformly heterozygous, we next evaluated their dosage sensitivity using ClinGen haploinsufficiency scores. Among our cohort genes with annotated dosage data, 24 were benign, 4 likely benign, 3 uncertain, 16 pathogenic, and 66 autosomal recessive (AR) linked (Fig. 5 B), corresponding to 30, 5, 3, 19, and 116 unique SVs, respectively (Fig. 5 C). As an impact threshold, we filtered SVs for haploinsufficiency or constraint > 1, consistent with our constraint density peaks, and prioritized PUR probands with SVs overlapping functional gene regions. One SNV-negative proband (PUR3P) presenting with autism, hypermobility, cataracts, anemia, and autoimmunity had a maternally inherited 3.8 kb heterozygous HBA2 (LoFz=-0.6; CG = AR) coding deletion (chr16:166,741–170,548) only observed in this proband, exclusive to the anemia trait, and consistent with enrichment in Puerto Rican populations 24 , 25 ( Supplemental Fig. 7 ). More complex phenotypes (> 10 HPO terms) with diagnostic SNV findings suggested multi-locus effects. PUR2P had compound heterozygous YLPM1 SNVs and also carried compound ANKRD11 (LoFz = 10.8; CG = Pathogenic) 3’UTR deletions (61 bp; chr16:95,379,667–95,379,728 and 63 bp; chr16:95,379,980–95,380,043), each DEL observed with co-occurrence in 4 probands (PUR = 2, EUR = 2). An 86 bp EHMT1 (LoFz = 9.1; CG = Pathogenic) 3’UTR deletion (chr9:150,131,808–150,131,894) was present in 7 probands (PUR = 5, EUR = 2), and a paternal 779 bp GABRB3 (LoFz = 4.3; CG = Unknown) 5′UTR deletion (chr15:24,457,716–24,458,495) in 11 probands (PUR = 6, EUR = 5). While individual events recur in the cohort, no other proband carried the full combination observed here, implicating combined disruption of transcriptional regulation ( YLPM1 , ANKRD11 , EHMT1 ) and GABAergic signaling ( GABRB3 ), and suggesting a potential polygenic contribution to the proband’s neurodevelopmental features ( Supplemental Fig. 8A-C ). PUR6P had pathogenic NF1 and TMEM147 SNVs and exhibited a 1.89 kb CHMP1A (LoFz = 1.3; CG = AR) deletion (chr16:95,731,319–95,733,209) spanning CDS and UTR, and a 135 bp 3’UTR deletion (chr16:95,733,580–95,733,715) observed only in this proband and 8 other probands (PUR = 5, EUR = 4), respectively; co-occurrence of these two CHMP1A variants was seen only in this proband—consistent with biallelic CHMP1A disruption and the proband’s microcephaly, hypotonia, and abnormal movements 26 ( Supplemental Fig. 9 ). Furthermore, two additional SNV-negative probands presented plausible regulatory hits. PUR5P, with ataxia, dystonia, and dysphagia, carried a 2.15 kb FGF12 (LoFz = 1.6; CG = Unknown) 5′UTR deletion (chr3:195,368,031–195,370,181) observed in 2 probands (PUR = 1, EUR = 1), at a locus that modulates NaV channel activity and plausibly contributes to early-onset cerebellar ataxia, movement, and intellectual disabilities – also shared by the EUR proband with the variant ( Supplemental Fig. 10 ) 27 . PUR15P, with ventricular septal defect and idiopathic epilepsy, was negative but had two compound heterozygous SZT2 VUS potentially driving mTOR-pathway dysregulation, alongside a 59 bp KCNQ2 (LoFz = 6.1; CG = Pathogenic) 3′UTR deletion (chr20:65,209,664–65,209,723) observed in 4 probands (PUR = 2, EUR = 2) and a 1.1 kb GTF2I (LoFz = 1.7; CG = Unknown) CDS deletion (chr7:75,944,926–75,946,021) observed in only this proband, with co-occurrence of these two loci also only in this proband ( Supplemental Fig. 11 ). The 3’ UTR deletion in KCNQ2 plausibly reduces channel gene expression and lowers seizure threshold, while the coding deletion in the transcription factor GTF2I may perturb developmental cardiac pathways. Collectively, these results demonstrate that integrating gene-level constraint and dosage sensitivity helps reveal functionally intolerant SVs, with PUR-enriched variants disproportionately affecting dosage-sensitive loci across different biological pathways and phenotypes. DISCUSSION T2T-CHM13 alignment paired with HiFi long-read sequencing reveals ancestry-specific SV landscapes that are largely inaccessible to short read analyses 2 , 3 , 15 . Between Puerto Rican and European trios, we observe a consistently higher SV count in Puerto Rican individuals (median = 32,923 vs 31,379; p < 0.0001), enrichment at pericentromeric/centromeric and telomeric regions, and enhanced ancestry discrimination power from SVs in relation to SNVs (72.3% vs 8.6% variance). Notably, our per-genome SV count (∼32.8k SVs) is consistent with prior long-read SV catalogs reporting on the order of ~ 27.6k SVs per genome, while finding additional signals in complex regions enabled by the T2T reference 1 , 28 . Compared with prior work, our study sits at the intersection of three advances that have often been addressed separately: improved SV calling with long read sequencing, improved reference completeness with T2T alignment, and improved representation of admixed populations. The shift to a complete reference is especially beneficial for SV analyses because T2T-CHM13 resolves large tracts of repetitive sequence and segmental duplications that were previously incomplete or misrepresented, opening these regions to variation discovery and interpretation 2 , 29 . In parallel, long-read population studies demonstrate better SV discovery and genotyping, supporting the use of long-read SV callsets as a foundation for variant interpretation 30 . Finally, because admixed American cohorts are frequently aggregated despite having meaningful genetic substructure, ancestry-matched catalogs are necessary to avoid diluting subgroup-specific variation relevant to clinical interpretation 31 . Consistent with broader significance, our results also suggest that the additional SVs called in Puerto Rican genomes are not merely quantitative but enriched in a genetic context that can affect clinical interpretation. We show that Puerto Ricans have SVs that intersect constrained and dosage-sensitive genes in higher frequency compared to Europeans, and several case interpretations required integration of SVs in review to resolve complex genetic architectures that SNV-centric analyses could not. This included overlapping CDS/UTR intervals with plausible dosage effects, and multi-locus combinations uncovered by long-read genotypes (e.g., HBA2 coding deletion, biallelic CHMP1A , and combined regulatory/coding hits such as ANKRD11 / EHMT1 / GABRB3 , SZT2 with KCNQ2 and GTF2I ). These examples illustrate how ancestry-specific patterns may translate into clinically relevant findings when integrating an ancestry-matched SV reference 32 – 34 . Together, our results suggest that an inclusive cohort design and repeat-resolving technologies could help deliver better precision genomics: ancestry-specific SV catalogs, standardized SV interpretations, and harmonized cross-platform benchmarks to improve diagnostic yield and reduce misinterpretation in underrepresented populations, including Puerto Ricans 15 , 32 , 35 . Limitations involve residual uncertainties at the most repetitive loci and differences in SV definitions across short-read resources. Continued integration of complete references and long-read data should further resolve cross-ancestry comparisons and case interpretations 2 , 15 . CONCLUSION A complete reference and long-read sequencing reveal ancestry-informative and clinically relevant structural variants in a rare disease Puerto Rican cohort. By integrating constraint, dosage, and functional gene region annotations, we can identify high impact variants with clinically relevant interpretation. Inclusion of underrepresented populations with a higher frequency of structural variation is essential to realize equitable precision medicine. Declarations ADDITIONAL INFORMATION Further information and requests for resources should be directed to and will be fulfilled by the corresponding authors, Cas LeMaster ( [email protected] ) and Craig Smail ( [email protected] ). DATA AVAILABILITY GA4K study data can be found at the ANVIL host at https://anvilproject.org/data/studies/phs002206/workspaces . Code used can be found in a git-hub repository at https://github.com/smail-lab-cmh/ . COMPETING INTERESTS The authors declare no competing interests. Author Contribution Conceptualization: C.L., E.F., E.A., T.P., C.S.; data analysis: C.L.; writing: C.L.; editing: all authors; funding: C.S. Acknowledgement C.S. is supported by NIH grant R35GM146966. The work of GA4K was made possible in part by generous gifts to Children’s Mercy Research Institute at Children’s Mercy Kansas City. This work was funded through internal institutional funds from Children’s Mercy Research Institute and Children’s Mercy Kansas City. References Aganezov S, et al. A complete reference genome improves analysis of human genetic variation. Science. 2022;376:eabl3533. 10.1126/science.abl3533 . Nurk S, et al. The complete sequence of a human genome. Science. 2022;376:44–53. 10.1126/science.abj6987 . Ebert P, et al. Haplotype-resolved diverse human genomes and integrated analysis of structural variation. Science. 2021;372. 10.1126/science.abf7117 . Porubsky D, et al. Gaps and complex structurally variant loci in phased genome assemblies. Genome Res. 2023;33:496–510. 10.1101/gr.277334.122 . Chen S, et al. A genomic mutational constraint map using variation in 76,156 human genomes. Nature. 2024;625:92–100. 10.1038/s41586-023-06045-0 . Banu S, et al. Enhanced resolution of optical genome mapping utilizing telomere-to-telomere reference in genetic disorders. Eur J Hum Genet. 2025;33:956–9. 10.1038/s41431-024-01763-z . Fatumo S, et al. A roadmap to increase diversity in genomic studies. Nat Med. 2022;28:243–50. 10.1038/s41591-021-01672-4 . Chair SY, et al. Structural Variations Identified in Patients with Autism Spectrum Disorder (ASD) in the Chinese Population: A Systematic Review of Case-Control Studies. Genes (Basel). 2024;15. 10.3390/genes15081082 . Isshiki M, et al. Genetic disease risks of under-represented founder populations in New York City. PLoS Genet. 2025;21:e1011755. 10.1371/journal.pgen.1011755 . Zimin AV, et al. A reference-quality, fully annotated genome from a Puerto Rican individual. Genetics. 2022;220. 10.1093/genetics/iyab227 . Duconge J, et al. Pharmacogenomic polygenic risk score for clopidogrel responsiveness among Caribbean Hispanics: A candidate gene approach. Clin Transl Sci. 2021;14:2254–66. 10.1111/cts.13124 . Zeiger AM, et al. Identification of CFTR variants in Latino patients with cystic fibrosis from the Dominican Republic and Puerto Rico. Pediatr Pulmonol. 2020;55:533–40. 10.1002/ppul.24549 . Srinivasan T, et al. Integrating Genomic Screening into Primary Care: Provider Experiences Caring for Latino Patients at a Community-Based Health Center. J Prim Care Community Health. 2021;12:21501327211000242. 10.1177/21501327211000242 . Altemose N, et al. Complete genomic and epigenetic maps of human centromeres. Science. 2022;376:eabl4178. 10.1126/science.abl4178 . Liao WW, et al. A draft human pangenome reference. Nature. 2023;617:312–24. 10.1038/s41586-023-05896-x . Saunders CT, et al. Sawfish: improving long-read structural variant discovery and genotyping with local haplotype modeling. Bioinformatics. 2025;41. 10.1093/bioinformatics/btaf136 . Joe S, et al. Comparison of structural variant callers for massive whole-genome sequence data. BMC Genomics. 2024;25:318. 10.1186/s12864-024-10239-9 . Paulin LF, et al. Closing the gaps, and improving somatic structural variant analysis and benchmarking using CHM13-T2T. Genome Res. 2025;35:621–31. 10.1101/gr.279352.124 . Logsdon GA, et al. The variation and evolution of complete human centromeres. bioRxiv. 2023. 10.1101/2023.05.30.542849 . Said I, Barbash DA, Clark AG. The Structure of Simple Satellite Variation in the Human Genome and Its Correlation With Centromere Ancestry. Genome Biol Evol. 2024;16. 10.1093/gbe/evae153 . Suzuki Y, Morishita S. The time is ripe to investigate human centromeres by long-read sequencingdagger. DNA Res. 2021;28. 10.1093/dnares/dsab021 . Kojima S, et al. Mobile element variation contributes to population-specific genome diversification, gene regulation and disease risk. Nat Genet. 2023;55:939–51. 10.1038/s41588-023-01390-2 . Koenig Z, et al. A harmonized public resource of deeply sequenced diverse human genomes. Genome Res. 2024;34:796–809. 10.1101/gr.278378.123 . Soler AM, et al. Alpha thalassemia and alpha-MRE haplotypes in Uruguayan patients with microcytosis and hypochromia without anemia. Genet Mol Biol. 2021;44:e20200399. 10.1590/1678-4685-GMB-2020-0399 . Bryce RM, et al. The prevalence, correlates and impact of anaemia among older people in Cuba, Dominican Republic, Mexico, Puerto Rico and Venezuela. Br J Haematol. 2013;160:387–98. 10.1111/bjh.12153 . Sakamoto M, et al. A novel homozygous CHMP1A variant arising from segmental uniparental disomy causes pontocerebellar hypoplasia type 8. J Hum Genet. 2023;68:247–53. 10.1038/s10038-022-01098-x . Willemsen MH, et al. Epilepsy phenotype in individuals with chromosomal duplication encompassing FGF12. Epilepsia Open. 2020;5:301–6. 10.1002/epi4.12396 . Chaisson MJP, et al. Multi-platform discovery of haplotype-resolved structural variation in human genomes. Nat Commun. 2019;10:1784. 10.1038/s41467-018-08148-z . Vollger MR, et al. Segmental duplications and their variation in a complete human genome. Science. 2022;376:eabj6965. 10.1126/science.abj6965 . Beyter D, et al. Long-read sequencing of 3,622 Icelanders provides insight into the role of structural variants in human diseases and other traits. Nat Genet. 2021;53:779–86. 10.1038/s41588-021-00865-4 . Sharma J, et al. Genetic ancestry influences gene-environment interactions with sociocultural factors: Results from the Hispanic Community Health Study/Study of Latinos. HGG Adv. 2025;6:100451. 10.1016/j.xhgg.2025.100451 . Collins RL, et al. A structural variation reference for medical and population genetics. Nature. 2020;581:444–51. 10.1038/s41586-020-2287-8 . Karczewski KJ, et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020;581:434–43. 10.1038/s41586-020-2308-7 . Del Gobbo GF, et al. Long-read genome sequencing reveals a novel intronic retroelement insertion in NR5A1 associated with 46,XY differences of sexual development. Am J Med Genet A. 2024;194:e63522. 10.1002/ajmg.a.63522 . Zook JM, et al. A robust benchmark for detection of germline large deletions and insertions. Nat Biotechnol. 2020;38:1347–55. 10.1038/s41587-020-0538-8 . Additional Declarations No competing interests reported. Supplementary Files SupplementalTable1.xlsx SupplementalFigureLegends.docx Cite Share Download PDF Status: Under Review Version 1 posted Reviewers invited by journal 19 Feb, 2026 Editor assigned by journal 21 Jan, 2026 Submission checks completed at journal 08 Jan, 2026 First submitted to journal 07 Jan, 2026 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-8544782","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":595804567,"identity":"01366620-9af2-4d85-8ff5-13c7aa024efb","order_by":0,"name":"Cas LeMaster","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAwklEQVRIiWNgGAWjYDACdjijgYEZSDE2ENTCDCYNGBh4DpCsRSKBSC38zcxHN/z480fOfObjp5sLGGxkNxwgoEXiMFvazd42A2OZ22lmt2cwpBkT1GLAzGN2g7fBIHGGdA7bbR6Gw4lEaOH/dvPPH4P6GZJnQFr+E6OFB6iSzSBBQgLEYDhAWAvQL2a3ZduMDWfwAP3CY5BsPJOQFv725mc33/yRk5dgP/zsNk+FnWwfIS3o7iRN+SgYBaNgFIwCHAAAa+g+u7K3/WkAAAAASUVORK5CYII=","orcid":"","institution":"Children's Mercy Hospital","correspondingAuthor":true,"prefix":"","firstName":"Cas","middleName":"","lastName":"LeMaster","suffix":""},{"id":595804568,"identity":"81d5714c-4276-48f3-88c4-1daef67efe34","order_by":1,"name":"Emily Farrow","email":"","orcid":"","institution":"Children's Mercy Hospital","correspondingAuthor":false,"prefix":"","firstName":"Emily","middleName":"","lastName":"Farrow","suffix":""},{"id":595804569,"identity":"8ae65757-084d-44e7-b729-b9128e93b216","order_by":2,"name":"Rebecca McLennan","email":"","orcid":"","institution":"Children's Mercy Hospital","correspondingAuthor":false,"prefix":"","firstName":"Rebecca","middleName":"","lastName":"McLennan","suffix":""},{"id":595804570,"identity":"b25804ba-12be-4e2a-b274-2fede90bc805","order_by":3,"name":"Elinette Albino","email":"","orcid":"","institution":"University of Puerto Rico, Medical Sciences Campus","correspondingAuthor":false,"prefix":"","firstName":"Elinette","middleName":"","lastName":"Albino","suffix":""},{"id":595804571,"identity":"b4ada802-8573-4388-bf1b-616cdaef1635","order_by":4,"name":"Tomi Pastinen","email":"","orcid":"","institution":"Children's Mercy Hospital","correspondingAuthor":false,"prefix":"","firstName":"Tomi","middleName":"","lastName":"Pastinen","suffix":""},{"id":595804572,"identity":"f3c09052-f9b4-4763-8078-09b4a472d70b","order_by":5,"name":"Craig Smail","email":"","orcid":"","institution":"Children's Mercy Hospital","correspondingAuthor":false,"prefix":"","firstName":"Craig","middleName":"","lastName":"Smail","suffix":""}],"badges":[],"createdAt":"2026-01-07 19:38:12","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8544782/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8544782/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":103301729,"identity":"c41356a4-d6a7-4c00-8fd5-ed92bbef078e","added_by":"auto","created_at":"2026-02-24 08:17:53","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":728873,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eLong read sequencing of a Puerto Rican and European trio cohort.\u003c/strong\u003e(\u003cstrong\u003eA\u003c/strong\u003e) Individual bars illustrate SV counts across the cohort for deletions (blue), duplications (orange), insertions (green), and inversions (red). (\u003cstrong\u003eB\u003c/strong\u003e) An overlapping density of SVs by length and type (deletions (blue), duplications (orange), insertions (green), and inversions (red)).\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/f7f62fafe9eeb927eec0b2a5.png"},{"id":103301726,"identity":"013a7b7b-eb1b-40cb-b70a-eeb154e339f5","added_by":"auto","created_at":"2026-02-24 08:17:53","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":115425,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eDistinct patterns of ancestry differences revealed by SNVs and SVs.\u003c/strong\u003e (\u003cstrong\u003eA-B\u003c/strong\u003e) Circos plots showing mean densities of proband sequencing depths (light blue), SNVs (green), SVs (gray), de novo deletions (red), de novo insertions (blue), and de novo duplications (light orange). The left circos displays Puerto Rican ancestry probands and the right European ancestry probands. (\u003cstrong\u003eC\u003c/strong\u003e) Violin plots of individual SNVs and SV type counts, separated by ancestry (EUR=blue, PUR=red). Mann-Whitney P-values are displayed between plots. (\u003cstrong\u003eD-E\u003c/strong\u003e) PCA plots displaying Puerto Rican trios (red, n=15), European trios (blue, n=15), and a single trio of African ancestry (light green). Each trio member is distinguished by shape (proband=square, mother=triangle, father=circle). (D) represents SNV-based clustering (n=2,515) while (E) is SV-based (n=4,971).\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/833a488c999c3e5c62892d85.png"},{"id":103301728,"identity":"5d111df2-6ceb-46a8-be34-a8b49453ce85","added_by":"auto","created_at":"2026-02-24 08:17:53","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":894547,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSV counts in Puerto Ricans and European cohorts.\u003c/strong\u003e (\u003cstrong\u003eLeft\u003c/strong\u003e) Dot plot showing carrier allele frequencies between ancestries for unique SVs, with PUR frequency increasing from left to right along the x-axis and EUR frequency increasing from bottom to top along the y-axis. Variants significantly associated with a particular ancestry are colored accordingly. Non-significant variants are colored gray nearer the center of the correlation line. (\u003cstrong\u003eMiddle\u003c/strong\u003e) Stacked bar chart showing ancestry-significant SV counts between ancestries across each chromosome. A triangle at the top of the bar indicates which ancestry has a higher count. (\u003cstrong\u003eRight\u003c/strong\u003e) Stacked bar chart of the top 50 ancestry-significant SVs and a count of the number of carriers of the associated ancestral group. A triangle at the top of the bar indicates which ancestry has more carriers. If the variant overlaps a gene body, the gene name is displayed atop the triangle (PR=red, EU=blue). SV types are separated by deletions (\u003cstrong\u003eA\u003c/strong\u003e), duplications (\u003cstrong\u003eB\u003c/strong\u003e), insertions (\u003cstrong\u003eC\u003c/strong\u003e), and inversions (\u003cstrong\u003eD\u003c/strong\u003e).\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/5e1dabdfc96ab555bfda5992.png"},{"id":103301727,"identity":"ebb534cd-d801-464e-bf9a-2af397da7ecb","added_by":"auto","created_at":"2026-02-24 08:17:53","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":31672,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eUnique SV count in across gene regions is higher in Puerto Ricans for all SV types.\u003c/strong\u003e Bar plots show the unique SV count differences between ancestries, x-axis divided by SV type (PUR = red, EUR=blue). Plots are separated by gene regions where SVs are found.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/2d4d34bf2ff79c50dd66cb60.png"},{"id":103301730,"identity":"11578a29-53f7-4152-92fe-f8ef91713b52","added_by":"auto","created_at":"2026-02-24 08:17:53","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":107046,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePuerto Rican SVs show greater propensity for pathogenicity.\u003c/strong\u003e(\u003cstrong\u003eA\u003c/strong\u003e) A density plot shows the constraint differences across ancestries (PUR=red, EUR=blue) of genes intersected by ancestry-exclusive SVs at functional regions (CDS, 5’ UTR, and 3’ UTR). (\u003cstrong\u003eB\u003c/strong\u003e) A categorical bar plot of the count of unique genes with ClinGen haploinsuffiency (HI) intersected by unique SVs in the cohort (separated by ancestry; PUR=red, EUR=blue). (\u003cstrong\u003eC\u003c/strong\u003e) A categorical bar plot of the unique SV count at ClinGen genes (regions CDS, 5’ UTR, 3’ UTR) with HI (separated by ancestry; PUR=red, EUR=blue).\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/a59d9db66690f80da7b156d2.png"},{"id":103506514,"identity":"d9c40240-dc62-4916-8863-c006d6c4081d","added_by":"auto","created_at":"2026-02-26 13:37:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3015339,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/f4d6774c-e795-4bb0-92e3-a715ffd4b209.pdf"},{"id":103301733,"identity":"3dd0b02a-3e68-495d-84d2-2865b3566b1c","added_by":"auto","created_at":"2026-02-24 08:17:55","extension":"xlsx","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":71446161,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementalTable1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/09598f57292f5e68044f914a.xlsx"},{"id":103301732,"identity":"d6137465-31e7-4a09-a135-6c2bdbb0c921","added_by":"auto","created_at":"2026-02-24 08:17:54","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":31490673,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementalFigureLegends.docx","url":"https://assets-eu.researchsquare.com/files/rs-8544782/v1/1f53fd21d5c241b3a255444d.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Population-Specific Structural Variant Landscape in a Puerto Rican Rare Disease Cohort","fulltext":[{"header":"BACKGROUND","content":"\u003cp\u003eThe integration of long-read PacBio High-Fidelity (HiFi) sequencing with the complete Telomere-to-Telomere CHM13 (T2T-CHM13) reference genome has transformed structural variant (SV) detection by resolving previously inaccessible genomic regions. T2T-CHM13 adds nearly 200\u0026nbsp;million base pairs while resolving segmental duplications, satellites, and other repeats that confound short reads, universally improving variant calling across diverse populations\u003csup\u003e\u003cspan additionalcitationids=\"CR2\" citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e\u003c/sup\u003e. This complete reference enables a 21% increase in inversion detection sensitivity and reduces false positives in medically relevant genes by up to 12-fold compared to GRCh38\u003csup\u003e4,5\u003c/sup\u003e. Clinical applications demonstrate significantly enhanced SV detection in complex genomic regions, directly improving diagnostic outcomes\u003csup\u003e\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTo date, genomic studies have largely focused on European-centric populations while other genetic backgrounds remain underrepresented, particularly those examining SVs\u003csup\u003e\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e\u003c/sup\u003e. Recent population analyses identified Puerto Ricans among the most underrepresented founder populations in genomic databases, creating significant disparities for variant interpretation\u003csup\u003e\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e,\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e\u003c/sup\u003e. This underrepresentation manifests clinically, as Caribbean Hispanic populations, including Puerto Ricans, show distinct pharmacogenomic profiles that differ substantially from European ancestry references, highlighting population-specific genetic burdens\u003csup\u003e\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e,\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e\u003c/sup\u003e. Thus, the exclusion of these populations from large-scale genomic initiatives perpetuates health disparities and limits precision medicine interventions in these communities\u003csup\u003e\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eThe clinical implications of T2T‑CHM13\u0026ndash;enabled SV discovery extend beyond a single diease area and are especially relevant for ancestry-informed interpretation. Pericentromeres, centromeres, telomeres, and segmental duplications harbor ancestry-specific repeats and mobile elements, accumulating SVs that vary in location and type across populations\u0026ndash;regions now resolvable with a complete reference and accurate long reads\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eHere, we used HiFi long-read sequencing aligned to the T2T reference to quantify ancestry-specific differences in SV location and type and characterized their potential for disease impacts in both Puerto Rican and European parent-proband trios. We show that Puerto Rican-specific SVs are enriched for intersecting high impact gene regions, including constrained and dosage-sensitive genes, and that complex clinical case interpretations\u0026mdash;such as compound UTR disruptions, overlapping CDS/UTR intervals, and multi-locus combinations\u0026mdash;require SV analysis that traditional approaches cannot reliably resolve.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eStudy Cohort\u003c/h2\u003e \u003cp\u003eThe study cohort includes 30 affected probands from 30 unrelated families (n\u0026thinsp;=\u0026thinsp;90). Participants from Puerto Rico were consented and enrolled at the University of Puerto Rico Medical Science Campus (IRB Study #2305114422). Control subjects were enrolled in the Genomics Answers for Kids (GA4K) program. Eligibility criteria included a suspected genetic diagnosis based on clinical presentation and/or existing molecular or cytogenic findings. Providers introduced the study and asked if families were interested in participation. The study complies with all relevant ethical regulations as approved by both the University of Puerto Rico School of Medicine and Children's Mercy Institutional Review Board (IRB) (Study #11120514). Informed written consent was obtained from all participants prior to enrollment. Participants did not receive financial compensation for participation.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eSequencing and Alignment\u003c/h3\u003e\n\u003cp\u003ePacBio HiFi raw sequencing reads were aligned to the CHM13v2.0 reference genome using pbmm2 (PacBio minimap2) with the CCS preset. Multiple BAM files per sample were merged using samtools merge when necessary. Alignment quality was assessed using samtools flagstat for mapping statistics and samtools depth for coverage analysis. BAM files were indexed using both samtools index and pbindex for downstream compatibility.\u003c/p\u003e\n\u003ch3\u003eStructural Variant Discovery and Genotyping with Local Haplotype Modeling\u003c/h3\u003e\n\u003cp\u003eStructural variants (SVs) were identified using Sawfish (v0.12.9), which performs local haplotype modeling for accurate breakpoint resolution and genotyping in complex regions. SV discovery was performed per sample using Sawfish discover, utilizing all genomic regions (coverage regex \".\") to identify clusters of SV alignment signatures. Individual sample genotyping was performed using the default Sawfish joint-call option, which models and genotypes all SV breakpoints using local haplotypes.\u003c/p\u003e \u003cp\u003eTrio joint-calling was implemented using Sawfish's multi-sample joint-call functionality to improve variant calling accuracy through family-based genotyping and Mendelian consistency. Discovery results from all three trio members (proband, mother, and father) were jointly analyzed using sawfish joint-call. This approach consolidates duplicate SV haplotypes across samples and performs genotyping in the context of overlapping haplotype pools.\u003c/p\u003e \u003cp\u003eTrio VCFs were merged using samtools, and SVs were filtered for quality\u0026thinsp;\u0026gt;\u0026thinsp;30 (Phred-scaled, 99.9% confidence), depth\u0026thinsp;\u0026gt;\u0026thinsp;10 (to ensure adequate coverage for confident genotyping), and minimum length\u0026thinsp;\u0026gt;\u0026thinsp;50bp (standard SV size threshold). Additionally, we required SVs to pass Sawfish's internal FILTER field to exclude low-quality calls.\u003c/p\u003e\n\u003ch3\u003eSingle Nucleotide Variant Calling\u003c/h3\u003e\n\u003cp\u003eSmall variants and single nucleotide variants (SNVs) were called using Clair3 with the HiFi Revio model, optimized for PacBio HiFi sequencing data. Clair3 analysis was performed using the \"hifi\" preset to maximize calling accuracy for HiFi reads. SNVs were filtered for QUAL\u0026thinsp;\u0026gt;\u0026thinsp;20 and GQ\u0026thinsp;\u0026gt;\u0026thinsp;20 to ensure high-confidence calls.\u003c/p\u003e\n\u003ch3\u003eAncestry PCAs\u003c/h3\u003e\n\u003cp\u003eAncestry analysis was performed using both SNVs and SVs with parallel analytical frameworks to enable comparison. The analysis included 93 individuals from 31 families across three populations: European (EUR, n\u0026thinsp;=\u0026thinsp;45), Puerto Rican (PUR, n\u0026thinsp;=\u0026thinsp;45), and African (AFR, n\u0026thinsp;=\u0026thinsp;3). Each family consisted of a trio (proband, mother, and father).\u003c/p\u003e \u003cp\u003eFor SNV analysis, we used Somalier to identify 2,515 ancestry-informative SNPs and calculate pairwise relatedness between samples using Somalier's internal algorithms. The resulting relatedness values were converted to distance matrices (1 - relatedness) and classical multidimensional scaling (MDS) was applied using R's cmdscale function for dimensionality reduction.\u003c/p\u003e \u003cp\u003eFor SV analysis, we extracted 4,971 high-quality SVs (allele frequency 0.01\u0026ndash;0.99, size\u0026thinsp;\u0026ge;\u0026thinsp;50bp, missingness\u0026thinsp;\u0026le;\u0026thinsp;10% per variant) from joint-called VCF files with SV type-specific length thresholds. Pairwise relatedness was calculated from genotype correlations and normalized to range 0\u0026ndash;1. Classical MDS was implemented in Python with double centering of squared distance matrices followed by eigen decomposition to extract coordinates.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eGenes, Regions, Constraint Mapping, and Dosage Sensitivity\u003c/h2\u003e \u003cp\u003eVariants were annotated with intersecting genes via bedtools intersect and systematically collapsed on unique variant IDs. Gene annotations were obtained from the Telomere-to-Telomere Consortium GitHub under T2T-CHM13v2.0 (T2T-CHM13\u0026thinsp;+\u0026thinsp;Y) via the HPRC EBI GENCODEv38 r2 gff3 file\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eRegional classes and constraint metrics were assigned to SVs using the following hierarchy: (1) for SVs intersecting multiple genes, we used the mean constraint score among all intersected genes; (2) for SVs intersecting multiple genomic regions of the same gene, we prioritized CDS\u0026thinsp;\u0026gt;\u0026thinsp;UTR\u0026thinsp;\u0026gt;\u0026thinsp;intronic overlaps; (3) categorical haploinsufficiency data was obtained from the Clinical Genome Resource (ClinGen). Constraint scores were provided by gnomADv4 resources using the pLOF z-score for canonical genes.\u003c/p\u003e \u003cp\u003eChromosomal regions were defined using the T2T-CHM13v2.0 censat file available via the UCSC genome browser. Regions p, q, and centromeric were defined via bed coordinates, telomeric regions were defined via 1Mb sections from the end of chromosome arms (with consideration for acrocentric chromosomes), and pericentromeric regions were 5Mb sections from the centromere end coordinate in respective directions to the arms they associated with.\u003c/p\u003e \u003cp\u003eFor gnomAD comparison, we downloaded the v4.1 SV call set and harmonized coordinates from GRCh38 to CHM13v2.0 using a chain-based liftover, retained unique SVs, and excluded centromeric/satellite/rDNA intervals on CHM13. When matching between the cohort and gnomAD, we used\u0026thinsp;\u0026ge;\u0026thinsp;50% reciprocal overlapping for DEL/DUP/INV and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\pm\\:\\)\u003c/span\u003e\u003c/span\u003e100 bp start tolerance for INS.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eStatistics\u003c/h3\u003e\n\u003cp\u003eCount comparisons across ancestries were tested for significance using Mann-Whitney U Tests. Unique structural variant comparisons across ancestries were made using Fisher\u0026rsquo;s Exact Test.\u003c/p\u003e"},{"header":"RESULTS","content":"\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eStructural Variant Discovery and Characterization in European and Puerto Rican Trios\u003c/h2\u003e \u003cp\u003eTo investigate structural variation across the genome from different ancestry groups, we focused on 90 samples across 30 parent-child trios with 15 of European (EUR) and 15 of Puerto Rican (PUR) ancestry. Using HiFi long-read whole genome sequencing aligned to the T2T-CHM13v2.0 reference genome \u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e\u003c/sup\u003e, we achieved high coverage and depth across autosomes (mean coverage\u0026thinsp;=\u0026thinsp;98.75%, mean depth\u0026thinsp;=\u0026thinsp;31x) and sex chromosomes (mean coverage\u0026thinsp;=\u0026thinsp;95.6%, mean depth\u0026thinsp;=\u0026thinsp;22.9x; \u003cb\u003eSupplemental Table\u0026nbsp;1\u003c/b\u003e; \u003cb\u003eSupplemental Fig.\u0026nbsp;1A\u0026mdash;B\u003c/b\u003e). This approach systematic downstream variant detection, enabling the identification of 1,729,471 deletions (DELs), 18,805 duplications (DUPs), 1,203,260 insertions (INSs), and 2,872 inversions (INVs) across the cohort, with an average of 32,827 SVs per individual (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA; \u003cb\u003eSupplemental Table\u0026nbsp;1\u003c/b\u003e). These counts align well with previous long-read sequencing studies\u003csup\u003e\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e\u003c/sup\u003e, with one trio (EUR4) showing comparatively lower counts while maintaining high coverage (95%) and depth (23.2x). Size distribution analysis revealed that DELs and INSs, while constituting the majority of variants, typically occupied smaller genomic regions with the highest density under 500 bp and notable peaks at 300 bp and 6 kb corresponding to Alu element and LINE-1 retrotransposon activities, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) \u003csup\u003e\u003cspan additionalcitationids=\"CR25\" citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e. DUPs and INVs showed larger size distributions extending beyond 1 Mb, consistent with their complex formation mechanisms.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eAncestry-Specific Distribution of Genomic Variation Reveals Enhanced Discriminatory Power of Structural Variants\u003c/h2\u003e \u003cp\u003eTo understand population differences in SV counts and build on our comprehensive SV catalog, we investigated ancestry-specific patterns of genomic variation. Investigating these patterns across the genome, we constructed circos plots displaying mean variant densities for probands in each ancestry group (EUR n\u0026thinsp;=\u0026thinsp;15, PUR n\u0026thinsp;=\u0026thinsp;15, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA-B), visualizing chromosomal regions enriched for both SVs and single nucleotide variants (SNVs). While SNVs were not our primary focus, recent studies have demonstrated that different chromosomal regions exhibit varying susceptibility to SNV and SV accumulation due to repetitive sequences, recombination hotspots, and structural rearrangement (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA-B)\u003csup\u003e19\u003c/sup\u003e. This revealed large peaks around telomeric and centromeric regions for both variant classes that often paralleled lower coverage. To quantify ancestry-specific enrichment, we calculated log2 ratios (log2r) of variant counts between populations, highlighting regions with the largest relative differences (\u0026gt;\u0026thinsp;0.3 log2r) for each ancestry cohort (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA-B; \u003cb\u003eSupplemental Table\u0026nbsp;1\u003c/b\u003e). We observed that the pericentromeric (pc) region chr20pc_q (log2r\u0026thinsp;=\u0026thinsp;0.51; p\u0026thinsp;=\u0026thinsp;0.0004), telomeric (t) region chr1t_p (log2r\u0026thinsp;=\u0026thinsp;0.39; p\u0026thinsp;=\u0026thinsp;0.0007), and the centromere (c) of chr7c (log2r\u0026thinsp;=\u0026thinsp;0.34, p\u0026thinsp;=\u0026thinsp;0.02) showed significantly higher structural variation in PUR individuals, particularly for DELs and INSs. In contrast, no regions were more associated with EUR ancestry (\u003cb\u003eSupplemental Table\u0026nbsp;1; Supplemental Fig.\u0026nbsp;2\u003c/b\u003e). These findings are consistent with recent work demonstrating that centromeric regions can harbor non-pathogenic population-specific satellite DNA variants and higher-order repeats that correlate with ancestry\u003csup\u003e\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e,\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eIntriguingly, SNV patterns showed inverse relationships in some regions: EUR-enriched SV regions at chr12c and chr4c harbored more SNVs in PUR individuals (log2r\u0026thinsp;=\u0026thinsp;0.25 \u0026amp; 0.19, respectively), while PUR-enriched SV regions like chr1t_p showed higher SNV counts in EUR individuals (log2r\u0026thinsp;=\u0026thinsp;0.15). Additional PUR-enriched SNV regions included chr21c centromere (log2r\u0026thinsp;=\u0026thinsp;0.42) and telomeric regions chr18t_p, chr8t_p, and chr10t_p (log2r\u0026thinsp;=\u0026thinsp;0.33, 0.32, \u0026amp; 0.32, respectively). Notably, all EUR regions with \u0026gt;\u0026thinsp;0.3 log2r SNVs were telomeric (chr19t_p, chr16t_p, chr12t_q; log2r\u0026thinsp;=\u0026thinsp;0.55, 0.44, \u0026amp; 0.37, respectively), suggesting potential ancestry-specific telomeric patterns for SNVs (\u003cb\u003eSupplemental Table\u0026nbsp;1; Supplemental Figs.\u0026nbsp;3 and 4\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eGenome-wide variant quantification revealed higher overall variant counts in PUR individuals for both SNVs (PUR median\u0026thinsp;=\u0026thinsp;891,820, EUR median\u0026thinsp;=\u0026thinsp;830,008; p\u0026thinsp;=\u0026thinsp;0.005) and SVs (PUR median\u0026thinsp;=\u0026thinsp;32,923, EUR median\u0026thinsp;=\u0026thinsp;31,379; p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001) (Mann-Whitney U test; Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). The magnitude of SV differences (\u0026gt;\u0026thinsp;2%) compared to SNVs suggests that structural variation may be particularly informative for ancestry inference. To directly test this hypothesis, we performed parallel dimensionality reduction analyses comparing the ancestry discrimination power of SNVs versus SVs using conceptually similar multidimensional scaling (MDS) approaches. We included an additional African (AFR) ancestry trio to enhance population stratification. For SV analysis, we extracted 4,971 high-quality structural variants (allele frequency 0.01\u0026ndash;0.99, size\u0026thinsp;\u0026ge;\u0026thinsp;50bp, missingness\u0026thinsp;\u0026le;\u0026thinsp;10%) (\u003cb\u003eSupplemental Table\u0026nbsp;1\u003c/b\u003e). We observed that SV-based analysis led to substantial discriminatory power, with the first two principal coordinates explaining 72.3% of total variance (PC1: 38.0%, PC2: 34.3%, n\u0026thinsp;=\u0026thinsp;4,971) compared to 8.6% for SNV-based analysis (PC1: 5.0%, PC2: 3.6%, n\u0026thinsp;=\u0026thinsp;2,515). While both approaches successfully separated the three ancestry groups (EUR, PUR, AFR), the SV-based method produced higher-resolution population clusters with greater separation (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD-E; \u003cb\u003eSupplemental Table\u0026nbsp;1\u003c/b\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eAncestry-Specific Structural Variant Type Distributions Reveal Predominant Puerto Rican Associations\u003c/h2\u003e \u003cp\u003eHaving established the ancestry SV count differences, we next investigated the distribution of unique SVs by the significance of their association and their distribution by type and chromosome (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-D). For DELs, we identified 275,063 variants across the cohort, of which 10,937 (4% of total DELs) showed significant ancestry associations (unique SVs possessing significantly higher carrier frequency in one population; Fisher\u0026rsquo;s Exact), with 56% linked to PUR and 44% to EUR ancestry (\u003cb\u003eSupplemental Table\u0026nbsp;1;\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). These ancestry-associated DELs were predominantly located on chromosomes 2, 9, and 22, consistent with our chromosomal distribution analysis (\u003cb\u003eSupplemental Fig.\u0026nbsp;2\u003c/b\u003e). Among the top 50 most significant ancestry-associated DELs, only one was found exclusively in EUR individuals, while 18 were PUR-specific, 6 of which were intersecting genes (\u003cb\u003eSupplemental Table\u0026nbsp;1;\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). Duplications, while less frequent overall (4,040 total), showed similar ancestry association rates (4% of total DUPs) but with an even stronger PUR bias (69% of ancestry-significant associations) (\u003cb\u003eSupplemental Table\u0026nbsp;1;\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Chromosomes 17 and 20 harbored the majority of unique DUPs, with both showing predominant PUR associations. Among the top 50 ancestry-significant DUPs, 11 were EUR-specific while 22 were PUR-specific, and only one ancestry-shared DUP overlapped with a gene (\u003cb\u003eSupplemental Table\u0026nbsp;1;\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). Insertions also exhibited distinct patterns, with a lower proportion showing significant ancestry associations (3%), but nearly all chromosomes had PUR-enriched insertions except for chrY, which had the lowest overall count (\u003cb\u003eSupplemental Table\u0026nbsp;1;\u003c/b\u003e Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). Among the top 50 ancestry-significant INSs, 15 were PUR-specific while 35 were shared between populations. Notably, INSs showed the highest rate of gene intersection among all SV types: 35,262 of 122,751 unique INSs (29%) overlapped genes, compared to 16% for DELs, 12% for DUPs, and 22% for INVs (\u003cb\u003eSupplemental Table\u0026nbsp;1\u003c/b\u003e). This high INS rate could reflect the activity of mobile elements, particularly Alu and LINE-1 elements, which preferentially insert into open chromatin regions associated with active genes\u003csup\u003e\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e\u003c/sup\u003e. The 6,234 unique genes affected by INSs exceeded all other SV types (DEL: 5,792, DUP: 417, INV: 49). Lastly, inversions, while least frequent, followed similar patterns with a majority PUR-specific (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD). Overall, this suggests that structural variants show distinct population-specific patterns, with Puerto Rican ancestry exhibiting higher rates across all SV types.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo validate our findings in a larger population context, we harmonized gnomAD v4 SVs to CHM13v2.0 and compared them with our cohort, acknowledging methodological differences that limit direct comparisons\u003csup\u003e\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u003c/sup\u003e. GnomAD employs short-read sequencing aligned to the incomplete GRCh38 reference, uses broad ancestry definitions (Admixed/AMR vs. our specific Puerto Rican samples), and utilizes GATK-SV calling, which is largely haplotype-agnostic by default; in contrast, our study uses HiFi long reads, T2T alignment and Sawfish, leveraging local haplotype modeling and trio-aware joint calling. Despite these limitations, we anticipated that general patterns should parallel our findings. We identified 847,913 unique SVs in gnomAD, with 125,699 showing significant ancestry associations (Admixed American/AMR: 101,721; non-Finnish European/NFE: 23,978; \u003cb\u003eSupplemental Table\u0026nbsp;1\u003c/b\u003e). Consistent with our results, DELs were the most frequent SV type (72,130 total), with most associating with AMR ancestry (60,801). DUPs, INSs, and INVs all occurred at ~\u0026thinsp;3.0-4.0x higher rates in AMR compared to NFE genomes (DUP: 3.35x; INS: 3.03x; INV: 4.03x), supporting our findings of increased SV counts in admixed populations. When matching variants between datasets, we identified 41,568 overlapping variants (~\u0026thinsp;10.3% of our cohort) \u0026ndash; comprising 26,019 DELs, 495 DUPs, 15,026 INSs, and 28 INVs (\u003cb\u003eSupplemental Fig.\u0026nbsp;5\u003c/b\u003e). Overall, across matched variants more were found between AMR and PUR than NFE and EUR, and there was a greater frequency SVs in AMR than NFE, paralleling our findings.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eStructural Variants intersecting Coding Regions Reveal Ancestry-Specific Patterns of Potential Gene Disruption\u003c/h2\u003e \u003cp\u003eWith the observation that PUR individuals consistently exhibit higher variant counts than EUR individuals, we next assessed the differential potential impact of these variants on genes. Using GENCODE v38 r2 annotations from the Human Pangenome Reference Consortium (HPRC), we systematically evaluated SV overlaps with coding sequence (CDS), 5\u0026prime;UTR, 3\u0026rsquo;UTR, intronic, and intergenic regions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e)\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Because the functional consequences of SVs depend strongly on their genomic context, we ranked intersecting regions by potential impact, prioritizing CDS overlaps as most disruptive to gene function and following descending impact with 5\u0026rsquo;UTR, 3\u0026rsquo;UTR, intronic, and intergenic, respectively.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAcross SV types, unique INSs had a higher occurrence in intronic and 3\u0026rsquo; UTRs\u0026mdash;consistent with retrotransposons or mobile element activity, while DELs disproportionately affected coding, 5\u0026rsquo;UTR, and intergenic regions. Focusing on the highest impact, DELs in coding regions represented the strongest potential for disruption (n\u0026thinsp;=\u0026thinsp;388), followed by INSs (n\u0026thinsp;=\u0026thinsp;148), DUPs (n\u0026thinsp;=\u0026thinsp;48), and INVs (n\u0026thinsp;=\u0026thinsp;2). This illuminates a potentially higher impact DEL burden in the Puerto Rican population.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eGene Constraint and Dosage Sensitivity Reveal Functionally Intolerant SVs in Puerto Rican Probands\u003c/h2\u003e \u003cp\u003eTo assess the functional severity of SV overlaps across coding and regulatory regions, we integrated gene-level constraint metrics. We evaluated functional ancestry-exclusive SV intersections (CDS, 5\u0026prime;UTR, 3\u0026prime;UTR) using gene-level loss-of-function (LoF) constraint z-scores from gnomAD v4 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA, \u003cb\u003eSupplemental Fig.\u0026nbsp;6\u003c/b\u003e). Constraint density peaked between LoF z\u0026thinsp;=\u0026thinsp;1 \u0026amp; 2, with less tolerance across PUR-exclusive SVs (PUR mean z\u0026thinsp;=\u0026thinsp;1.69, n\u0026thinsp;=\u0026thinsp;384; EUR mean z\u0026thinsp;=\u0026thinsp;1.55, n\u0026thinsp;=\u0026thinsp;256). Positive mean z-scores for LoF constraint were seen across regional intersects for both populations, with PUR higher in both CDS and 5\u0026rsquo; UTR regions and EUR higher in 3\u0026rsquo; UTR (CDS mean PUR\u0026thinsp;=\u0026thinsp;1.58, EUR\u0026thinsp;=\u0026thinsp;1.27; 5\u0026prime;UTR mean PUR\u0026thinsp;=\u0026thinsp;1.99, EUR\u0026thinsp;=\u0026thinsp;1.76; 3\u0026prime;UTR mean PUR\u0026thinsp;=\u0026thinsp;1.72, EUR\u0026thinsp;=\u0026thinsp;1.93; \u003cb\u003eSupplemental Fig.\u0026nbsp;6A-C\u003c/b\u003e). Ancestry-agnostic SV density peaks between LoF z\u0026thinsp;=\u0026thinsp;4 \u0026amp; 6 reflected enrichment at \u003cem\u003eEXOC3\u003c/em\u003e (18%), driven by complex DEL\u0026ndash;INS events and clusters at \u003cem\u003eINF2\u003c/em\u003e, \u003cem\u003eBZW1\u003c/em\u003e, \u003cem\u003eFLI1\u003c/em\u003e, and \u003cem\u003eXPR1\u003c/em\u003e, none with established pathogenic relevance (\u003cb\u003eSupplemental Figure D\u003c/b\u003e). High-constraint loci (z\u0026thinsp;\u0026gt;\u0026thinsp;6) were predominately PUR-associated variants (60%). Gene-specific enrichments included \u003cem\u003eDOCK8\u003c/em\u003e (10/14 PUR; various intervals), \u003cem\u003eACAN\u003c/em\u003e (15/25 PUR; various intervals), \u003cem\u003eKDM4B\u003c/em\u003e (14/23 PUR; chr19:5,130,725-5,130,875), \u003cem\u003eKSR2\u003c/em\u003e (6/8 PUR; chr12:117,452,891\u0026thinsp;\u0026minus;\u0026thinsp;117,452,947), and \u003cem\u003eTTBK2\u003c/em\u003e (4/4 EUR; chr15:40,608,293\u0026thinsp;\u0026minus;\u0026thinsp;40,608,450).\u003c/p\u003e \u003cp\u003eGiven that these high-constraint variants were uniformly heterozygous, we next evaluated their dosage sensitivity using ClinGen haploinsufficiency scores. Among our cohort genes with annotated dosage data, 24 were benign, 4 likely benign, 3 uncertain, 16 pathogenic, and 66 autosomal recessive (AR) linked (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB), corresponding to 30, 5, 3, 19, and 116 unique SVs, respectively (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eAs an impact threshold, we filtered SVs for haploinsufficiency or constraint\u0026thinsp;\u0026gt;\u0026thinsp;1, consistent with our constraint density peaks, and prioritized PUR probands with SVs overlapping functional gene regions. One SNV-negative proband (PUR3P) presenting with autism, hypermobility, cataracts, anemia, and autoimmunity had a maternally inherited 3.8 kb heterozygous \u003cem\u003eHBA2\u003c/em\u003e (LoFz=-0.6; CG\u0026thinsp;=\u0026thinsp;AR) coding deletion (chr16:166,741\u0026ndash;170,548) only observed in this proband, exclusive to the anemia trait, and consistent with enrichment in Puerto Rican populations\u003csup\u003e\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e,\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e\u003c/sup\u003e (\u003cb\u003eSupplemental Fig.\u0026nbsp;7\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eMore complex phenotypes (\u0026gt;\u0026thinsp;10 HPO terms) with diagnostic SNV findings suggested multi-locus effects. PUR2P had compound heterozygous \u003cem\u003eYLPM1\u003c/em\u003e SNVs and also carried compound \u003cem\u003eANKRD11\u003c/em\u003e (LoFz\u0026thinsp;=\u0026thinsp;10.8; CG\u0026thinsp;=\u0026thinsp;Pathogenic) 3\u0026rsquo;UTR deletions (61 bp; chr16:95,379,667\u0026ndash;95,379,728 and 63 bp; chr16:95,379,980\u0026ndash;95,380,043), each DEL observed with co-occurrence in 4 probands (PUR\u0026thinsp;=\u0026thinsp;2, EUR\u0026thinsp;=\u0026thinsp;2). An 86 bp \u003cem\u003eEHMT1\u003c/em\u003e (LoFz\u0026thinsp;=\u0026thinsp;9.1; CG\u0026thinsp;=\u0026thinsp;Pathogenic) 3\u0026rsquo;UTR deletion (chr9:150,131,808\u0026ndash;150,131,894) was present in 7 probands (PUR\u0026thinsp;=\u0026thinsp;5, EUR\u0026thinsp;=\u0026thinsp;2), and a paternal 779 bp \u003cem\u003eGABRB3\u003c/em\u003e (LoFz\u0026thinsp;=\u0026thinsp;4.3; CG\u0026thinsp;=\u0026thinsp;Unknown) 5\u0026prime;UTR deletion (chr15:24,457,716\u0026ndash;24,458,495) in 11 probands (PUR\u0026thinsp;=\u0026thinsp;6, EUR\u0026thinsp;=\u0026thinsp;5). While individual events recur in the cohort, no other proband carried the full combination observed here, implicating combined disruption of transcriptional regulation (\u003cem\u003eYLPM1\u003c/em\u003e, \u003cem\u003eANKRD11\u003c/em\u003e, \u003cem\u003eEHMT1\u003c/em\u003e) and GABAergic signaling (\u003cem\u003eGABRB3\u003c/em\u003e), and suggesting a potential polygenic contribution to the proband\u0026rsquo;s neurodevelopmental features (\u003cb\u003eSupplemental Fig.\u0026nbsp;8A-C\u003c/b\u003e). PUR6P had pathogenic \u003cem\u003eNF1\u003c/em\u003e and \u003cem\u003eTMEM147\u003c/em\u003e SNVs and exhibited a 1.89 kb \u003cem\u003eCHMP1A\u003c/em\u003e (LoFz\u0026thinsp;=\u0026thinsp;1.3; CG\u0026thinsp;=\u0026thinsp;AR) deletion (chr16:95,731,319\u0026ndash;95,733,209) spanning CDS and UTR, and a 135 bp 3\u0026rsquo;UTR deletion (chr16:95,733,580\u0026ndash;95,733,715) observed only in this proband and 8 other probands (PUR\u0026thinsp;=\u0026thinsp;5, EUR\u0026thinsp;=\u0026thinsp;4), respectively; co-occurrence of these two CHMP1A variants was seen only in this proband\u0026mdash;consistent with biallelic \u003cem\u003eCHMP1A\u003c/em\u003e disruption and the proband\u0026rsquo;s microcephaly, hypotonia, and abnormal movements\u003csup\u003e\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e\u003c/sup\u003e (\u003cb\u003eSupplemental Fig.\u0026nbsp;9\u003c/b\u003e).\u003c/p\u003e \u003cp\u003eFurthermore, two additional SNV-negative probands presented plausible regulatory hits. PUR5P, with ataxia, dystonia, and dysphagia, carried a 2.15 kb \u003cem\u003eFGF12\u003c/em\u003e (LoFz\u0026thinsp;=\u0026thinsp;1.6; CG\u0026thinsp;=\u0026thinsp;Unknown) 5\u0026prime;UTR deletion (chr3:195,368,031\u0026ndash;195,370,181) observed in 2 probands (PUR\u0026thinsp;=\u0026thinsp;1, EUR\u0026thinsp;=\u0026thinsp;1), at a locus that modulates NaV channel activity and plausibly contributes to early-onset cerebellar ataxia, movement, and intellectual disabilities \u0026ndash; also shared by the EUR proband with the variant (\u003cb\u003eSupplemental Fig.\u0026nbsp;10\u003c/b\u003e)\u003csup\u003e\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u003c/sup\u003e. PUR15P, with ventricular septal defect and idiopathic epilepsy, was negative but had two compound heterozygous \u003cem\u003eSZT2\u003c/em\u003e VUS potentially driving mTOR-pathway dysregulation, alongside a 59 bp \u003cem\u003eKCNQ2\u003c/em\u003e (LoFz\u0026thinsp;=\u0026thinsp;6.1; CG\u0026thinsp;=\u0026thinsp;Pathogenic) 3\u0026prime;UTR deletion (chr20:65,209,664\u0026ndash;65,209,723) observed in 4 probands (PUR\u0026thinsp;=\u0026thinsp;2, EUR\u0026thinsp;=\u0026thinsp;2) and a 1.1 kb GTF2I (LoFz\u0026thinsp;=\u0026thinsp;1.7; CG\u0026thinsp;=\u0026thinsp;Unknown) CDS deletion (chr7:75,944,926\u0026ndash;75,946,021) observed in only this proband, with co-occurrence of these two loci also only in this proband (\u003cb\u003eSupplemental Fig.\u0026nbsp;11\u003c/b\u003e). The 3\u0026rsquo; UTR deletion in \u003cem\u003eKCNQ2\u003c/em\u003e plausibly reduces channel gene expression and lowers seizure threshold, while the coding deletion in the transcription factor GTF2I may perturb developmental cardiac pathways. Collectively, these results demonstrate that integrating gene-level constraint and dosage sensitivity helps reveal functionally intolerant SVs, with PUR-enriched variants disproportionately affecting dosage-sensitive loci across different biological pathways and phenotypes.\u003c/p\u003e \u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eT2T-CHM13 alignment paired with HiFi long-read sequencing reveals ancestry-specific SV landscapes that are largely inaccessible to short read analyses\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e. Between Puerto Rican and European trios, we observe a consistently higher SV count in Puerto Rican individuals (median\u0026thinsp;=\u0026thinsp;32,923 vs 31,379; p\u0026thinsp;\u0026lt;\u0026thinsp;0.0001), enrichment at pericentromeric/centromeric and telomeric regions, and enhanced ancestry discrimination power from SVs in relation to SNVs (72.3% vs 8.6% variance). Notably, our per-genome SV count (\u0026sim;32.8k SVs) is consistent with prior long-read SV catalogs reporting on the order of ~\u0026thinsp;27.6k SVs per genome, while finding additional signals in complex regions enabled by the T2T reference\u003csup\u003e\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e,\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eCompared with prior work, our study sits at the intersection of three advances that have often been addressed separately: improved SV calling with long read sequencing, improved reference completeness with T2T alignment, and improved representation of admixed populations. The shift to a complete reference is especially beneficial for SV analyses because T2T-CHM13 resolves large tracts of repetitive sequence and segmental duplications that were previously incomplete or misrepresented, opening these regions to variation discovery and interpretation\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u003c/sup\u003e. In parallel, long-read population studies demonstrate better SV discovery and genotyping, supporting the use of long-read SV callsets as a foundation for variant interpretation\u003csup\u003e\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e\u003c/sup\u003e. Finally, because admixed American cohorts are frequently aggregated despite having meaningful genetic substructure, ancestry-matched catalogs are necessary to avoid diluting subgroup-specific variation relevant to clinical interpretation\u003csup\u003e\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eConsistent with broader significance, our results also suggest that the additional SVs called in Puerto Rican genomes are not merely quantitative but enriched in a genetic context that can affect clinical interpretation. We show that Puerto Ricans have SVs that intersect constrained and dosage-sensitive genes in higher frequency compared to Europeans, and several case interpretations required integration of SVs in review to resolve complex genetic architectures that SNV-centric analyses could not. This included overlapping CDS/UTR intervals with plausible dosage effects, and multi-locus combinations uncovered by long-read genotypes (e.g., HBA2 coding deletion, biallelic \u003cem\u003eCHMP1A\u003c/em\u003e, and combined regulatory/coding hits such as \u003cem\u003eANKRD11\u003c/em\u003e/\u003cem\u003eEHMT1\u003c/em\u003e/\u003cem\u003eGABRB3\u003c/em\u003e, SZT2 with \u003cem\u003eKCNQ2\u003c/em\u003e and \u003cem\u003eGTF2I\u003c/em\u003e). These examples illustrate how ancestry-specific patterns may translate into clinically relevant findings when integrating an ancestry-matched SV reference\u003csup\u003e\u003cspan additionalcitationids=\"CR33\" citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003eTogether, our results suggest that an inclusive cohort design and repeat-resolving technologies could help deliver better precision genomics: ancestry-specific SV catalogs, standardized SV interpretations, and harmonized cross-platform benchmarks to improve diagnostic yield and reduce misinterpretation in underrepresented populations, including Puerto Ricans\u003csup\u003e\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e,\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e,\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u003c/sup\u003e. Limitations involve residual uncertainties at the most repetitive loci and differences in SV definitions across short-read resources. Continued integration of complete references and long-read data should further resolve cross-ancestry comparisons and case interpretations\u003csup\u003e\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e,\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e\u003c/sup\u003e.\u003c/p\u003e"},{"header":"CONCLUSION","content":"\u003cp\u003eA complete reference and long-read sequencing reveal ancestry-informative and clinically relevant structural variants in a rare disease Puerto Rican cohort. By integrating constraint, dosage, and functional gene region annotations, we can identify high impact variants with clinically relevant interpretation. Inclusion of underrepresented populations with a higher frequency of structural variation is essential to realize equitable precision medicine.\u003c/p\u003e "},{"header":"Declarations","content":"\u003cdiv id=\"Sec18\" class=\"Section2\"\u003e \u003ch2\u003eADDITIONAL INFORMATION\u003c/h2\u003e \u003cp\u003eFurther information and requests for resources should be directed to and will be fulfilled by the corresponding authors, Cas LeMaster ([email protected]) and Craig Smail ([email protected]).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eDATA AVAILABILITY\u003c/h2\u003e \u003cp\u003eGA4K study data can be found at the ANVIL host at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://anvilproject.org/data/studies/phs002206/workspaces\u003c/span\u003e\u003cspan address=\"https://anvilproject.org/data/studies/phs002206/workspaces\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. Code used can be found in a git-hub repository at \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://github.com/smail-lab-cmh/\u003c/span\u003e\u003cspan address=\"https://github.com/smail-lab-cmh/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e \u003c/div\u003e\u003cp\u003e \u003ch2\u003eCOMPETING INTERESTS\u003c/h2\u003e \u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eConceptualization: C.L., E.F., E.A., T.P., C.S.; data analysis: C.L.; writing: C.L.; editing: all authors; funding: C.S.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003eC.S. is supported by NIH grant R35GM146966. The work of GA4K was made possible in part by generous gifts to Children\u0026rsquo;s Mercy Research Institute at Children\u0026rsquo;s Mercy Kansas City. This work was funded through internal institutional funds from Children\u0026rsquo;s Mercy Research Institute and Children\u0026rsquo;s Mercy Kansas City.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAganezov S, et al. A complete reference genome improves analysis of human genetic variation. Science. 2022;376:eabl3533. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.abl3533\u003c/span\u003e\u003cspan address=\"10.1126/science.abl3533\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNurk S, et al. The complete sequence of a human genome. Science. 2022;376:44\u0026ndash;53. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.abj6987\u003c/span\u003e\u003cspan address=\"10.1126/science.abj6987\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEbert P, et al. Haplotype-resolved diverse human genomes and integrated analysis of structural variation. Science. 2021;372. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.abf7117\u003c/span\u003e\u003cspan address=\"10.1126/science.abf7117\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePorubsky D, et al. Gaps and complex structurally variant loci in phased genome assemblies. Genome Res. 2023;33:496\u0026ndash;510. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/gr.277334.122\u003c/span\u003e\u003cspan address=\"10.1101/gr.277334.122\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen S, et al. A genomic mutational constraint map using variation in 76,156 human genomes. Nature. 2024;625:92\u0026ndash;100. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-023-06045-0\u003c/span\u003e\u003cspan address=\"10.1038/s41586-023-06045-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBanu S, et al. Enhanced resolution of optical genome mapping utilizing telomere-to-telomere reference in genetic disorders. Eur J Hum Genet. 2025;33:956\u0026ndash;9. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41431-024-01763-z\u003c/span\u003e\u003cspan address=\"10.1038/s41431-024-01763-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFatumo S, et al. A roadmap to increase diversity in genomic studies. Nat Med. 2022;28:243\u0026ndash;50. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41591-021-01672-4\u003c/span\u003e\u003cspan address=\"10.1038/s41591-021-01672-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChair SY, et al. Structural Variations Identified in Patients with Autism Spectrum Disorder (ASD) in the Chinese Population: A Systematic Review of Case-Control Studies. Genes (Basel). 2024;15. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/genes15081082\u003c/span\u003e\u003cspan address=\"10.3390/genes15081082\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eIsshiki M, et al. Genetic disease risks of under-represented founder populations in New York City. PLoS Genet. 2025;21:e1011755. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1371/journal.pgen.1011755\u003c/span\u003e\u003cspan address=\"10.1371/journal.pgen.1011755\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZimin AV, et al. A reference-quality, fully annotated genome from a Puerto Rican individual. Genetics. 2022;220. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/genetics/iyab227\u003c/span\u003e\u003cspan address=\"10.1093/genetics/iyab227\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuconge J, et al. Pharmacogenomic polygenic risk score for clopidogrel responsiveness among Caribbean Hispanics: A candidate gene approach. Clin Transl Sci. 2021;14:2254\u0026ndash;66. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/cts.13124\u003c/span\u003e\u003cspan address=\"10.1111/cts.13124\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZeiger AM, et al. Identification of CFTR variants in Latino patients with cystic fibrosis from the Dominican Republic and Puerto Rico. Pediatr Pulmonol. 2020;55:533\u0026ndash;40. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ppul.24549\u003c/span\u003e\u003cspan address=\"10.1002/ppul.24549\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSrinivasan T, et al. Integrating Genomic Screening into Primary Care: Provider Experiences Caring for Latino Patients at a Community-Based Health Center. J Prim Care Community Health. 2021;12:21501327211000242. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1177/21501327211000242\u003c/span\u003e\u003cspan address=\"10.1177/21501327211000242\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAltemose N, et al. Complete genomic and epigenetic maps of human centromeres. Science. 2022;376:eabl4178. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.abl4178\u003c/span\u003e\u003cspan address=\"10.1126/science.abl4178\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLiao WW, et al. A draft human pangenome reference. Nature. 2023;617:312\u0026ndash;24. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-023-05896-x\u003c/span\u003e\u003cspan address=\"10.1038/s41586-023-05896-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaunders CT, et al. Sawfish: improving long-read structural variant discovery and genotyping with local haplotype modeling. Bioinformatics. 2025;41. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/bioinformatics/btaf136\u003c/span\u003e\u003cspan address=\"10.1093/bioinformatics/btaf136\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJoe S, et al. Comparison of structural variant callers for massive whole-genome sequence data. BMC Genomics. 2024;25:318. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1186/s12864-024-10239-9\u003c/span\u003e\u003cspan address=\"10.1186/s12864-024-10239-9\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaulin LF, et al. Closing the gaps, and improving somatic structural variant analysis and benchmarking using CHM13-T2T. Genome Res. 2025;35:621\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/gr.279352.124\u003c/span\u003e\u003cspan address=\"10.1101/gr.279352.124\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLogsdon GA, et al. The variation and evolution of complete human centromeres. bioRxiv. 2023. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/2023.05.30.542849\u003c/span\u003e\u003cspan address=\"10.1101/2023.05.30.542849\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaid I, Barbash DA, Clark AG. The Structure of Simple Satellite Variation in the Human Genome and Its Correlation With Centromere Ancestry. Genome Biol Evol. 2024;16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/gbe/evae153\u003c/span\u003e\u003cspan address=\"10.1093/gbe/evae153\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuzuki Y, Morishita S. The time is ripe to investigate human centromeres by long-read sequencingdagger. DNA Res. 2021;28. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1093/dnares/dsab021\u003c/span\u003e\u003cspan address=\"10.1093/dnares/dsab021\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKojima S, et al. Mobile element variation contributes to population-specific genome diversification, gene regulation and disease risk. Nat Genet. 2023;55:939\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41588-023-01390-2\u003c/span\u003e\u003cspan address=\"10.1038/s41588-023-01390-2\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoenig Z, et al. A harmonized public resource of deeply sequenced diverse human genomes. Genome Res. 2024;34:796\u0026ndash;809. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1101/gr.278378.123\u003c/span\u003e\u003cspan address=\"10.1101/gr.278378.123\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSoler AM, et al. Alpha thalassemia and alpha-MRE haplotypes in Uruguayan patients with microcytosis and hypochromia without anemia. Genet Mol Biol. 2021;44:e20200399. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1590/1678-4685-GMB-2020-0399\u003c/span\u003e\u003cspan address=\"10.1590/1678-4685-GMB-2020-0399\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBryce RM, et al. The prevalence, correlates and impact of anaemia among older people in Cuba, Dominican Republic, Mexico, Puerto Rico and Venezuela. Br J Haematol. 2013;160:387\u0026ndash;98. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1111/bjh.12153\u003c/span\u003e\u003cspan address=\"10.1111/bjh.12153\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSakamoto M, et al. A novel homozygous CHMP1A variant arising from segmental uniparental disomy causes pontocerebellar hypoplasia type 8. J Hum Genet. 2023;68:247\u0026ndash;53. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s10038-022-01098-x\u003c/span\u003e\u003cspan address=\"10.1038/s10038-022-01098-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWillemsen MH, et al. Epilepsy phenotype in individuals with chromosomal duplication encompassing FGF12. Epilepsia Open. 2020;5:301\u0026ndash;6. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/epi4.12396\u003c/span\u003e\u003cspan address=\"10.1002/epi4.12396\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChaisson MJP, et al. Multi-platform discovery of haplotype-resolved structural variation in human genomes. Nat Commun. 2019;10:1784. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41467-018-08148-z\u003c/span\u003e\u003cspan address=\"10.1038/s41467-018-08148-z\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVollger MR, et al. Segmental duplications and their variation in a complete human genome. Science. 2022;376:eabj6965. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1126/science.abj6965\u003c/span\u003e\u003cspan address=\"10.1126/science.abj6965\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeyter D, et al. Long-read sequencing of 3,622 Icelanders provides insight into the role of structural variants in human diseases and other traits. Nat Genet. 2021;53:779\u0026ndash;86. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41588-021-00865-4\u003c/span\u003e\u003cspan address=\"10.1038/s41588-021-00865-4\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharma J, et al. Genetic ancestry influences gene-environment interactions with sociocultural factors: Results from the Hispanic Community Health Study/Study of Latinos. HGG Adv. 2025;6:100451. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.xhgg.2025.100451\u003c/span\u003e\u003cspan address=\"10.1016/j.xhgg.2025.100451\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCollins RL, et al. A structural variation reference for medical and population genetics. Nature. 2020;581:444\u0026ndash;51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-020-2287-8\u003c/span\u003e\u003cspan address=\"10.1038/s41586-020-2287-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKarczewski KJ, et al. The mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020;581:434\u0026ndash;43. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41586-020-2308-7\u003c/span\u003e\u003cspan address=\"10.1038/s41586-020-2308-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDel Gobbo GF, et al. Long-read genome sequencing reveals a novel intronic retroelement insertion in NR5A1 associated with 46,XY differences of sexual development. Am J Med Genet A. 2024;194:e63522. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/ajmg.a.63522\u003c/span\u003e\u003cspan address=\"10.1002/ajmg.a.63522\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZook JM, et al. A robust benchmark for detection of germline large deletions and insertions. Nat Biotechnol. 2020;38:1347\u0026ndash;55. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1038/s41587-020-0538-8\u003c/span\u003e\u003cspan address=\"10.1038/s41587-020-0538-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"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":"genome-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Genome Medicine](https://genomemedicine.biomedcentral.com/)","snPcode":"13073","submissionUrl":"https://submission.springernature.com/new-submission/13073/3","title":"Genome Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"structural variants, Puerto Rican genomics, T2T, HiFi, long-read, health disparities, ancestry-specific","lastPublishedDoi":"10.21203/rs.3.rs-8544782/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8544782/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe integration of long-read PacBio High-Fidelity (HiFi) sequencing with the complete Telomere-to-Telomere CHM13 (T2T-CHM13) reference genome has enabled thorough characterization of structural variants (SVs) in previously inaccessible genomic regions, yet Puerto Rican and broader admixed populations remain critically underrepresented in these advances. We performed HiFi whole genome sequencing on 90 samples across 30 parent-proband trios in the Genomic Answers for Kids (GA4K) program (15 European, 15 Puerto Rican) aligned to T2T-CHM13, identifying 1,729,471 deletions, 18,805 duplications, 1,203,260 insertions, and 2,872 inversions with stringent filtering. Puerto Rican individuals carried significantly more SVs, with enrichment in centromeric/pericentromeric and telomeric regions. SV genotypes provided strong ancestry discrimination (72.3% total variance by MDS vs 8.6% for SNVs), and ancestry-associated SVs were predominantly Puerto Rican for deletions and duplications. Functionally, Puerto Rican-enriched SVs intersected constrained and dosage-sensitive genes, including recurrent UTR and coding events with plausible regulatory or dosage effects. Together, these findings demonstrate that structural variants exhibit significant population-specific distributions and underscore the importance of combining complete reference genomes with long-read sequencing for ancestry-considerate interpretation.\u003c/p\u003e","manuscriptTitle":"Population-Specific Structural Variant Landscape in a Puerto Rican Rare Disease Cohort","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-24 08:17:48","doi":"10.21203/rs.3.rs-8544782/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewersInvited","content":"","date":"2026-02-20T02:11:42+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-01-21T20:51:12+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-01-08T06:20:38+00:00","index":"","fulltext":""},{"type":"submitted","content":"Genome Medicine","date":"2026-01-07T19:22:35+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"genome-medicine","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"Learn more about [Genome Medicine](https://genomemedicine.biomedcentral.com/)","snPcode":"13073","submissionUrl":"https://submission.springernature.com/new-submission/13073/3","title":"Genome Medicine","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC/SO AJ","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"a3d5ce85-a559-4bc2-a950-1faa53135968","owner":[],"postedDate":"February 24th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-02-24T08:17:48+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-24 08:17:48","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8544782","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8544782","identity":"rs-8544782","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}