Genetic Epidemiology

Background: Current Lynch syndrome (LS) prediction models do not account for detailed family structure and tumour mismatch repair (MMR) status. We developed LynchRisk, a Bayesian model that estimates individualized LS probability from pedigree, cancer history and tumour MMR test results. Methods: LynchRisk combines Mendelian inheritance, age- and sex-specific cancer hazards, affected and unaffected relatives, and detailed MMR status. Parameters were derived from the literature. Results: In two simulated cases of probands with LS-associated cancers, family structure and MMR status markedly changed LS probability estimates, with implications for germline testing decisions. LynchRisk also identified situations in which a negative germline result might be considered sufficient to rule out LS or, conversely, prompt further investigations such as somatic MMR or long-read sequencing. Discussion: LynchRisk gives individualized LS probability by combining clinical, family, and tumour data at different stages of the diagnostic process. Further work will focus on recalibrating cancer hazard functions, validating model performance through simulations, and developing an R package for use by interested research groups.

Motivation: Horizontal pleiotropy remains a major challenge for Mendelian randomization (MR), particularly as increasingly large genome-wide association studies identify many genetic instruments for complex traits. Existing pleiotropy-robust methods rely on assumptions regarding the proportion or distribution of valid instruments that may be violated when pleiotropy is widespread. Results: We developed CaPri, a Bayesian MR framework that jointly estimates causal and pleiotropic effects. CaPri allows a potentially arbitrary proportion of instruments to contribute to a shared directional pleiotropic component and estimates the posterior probability and magnitude of the causal effect, instrument-specific posterior inclusion probabilities for pleiotropy, and the shared pleiotropic effect. Across simulations, CaPri maintained appropriate false positive control and accurate causal effect estimation regardless of the proportion of pleiotropic instruments, while accurately identifying pleiotropic instruments and estimating their shared effect. Performance remained robust across varying causal and pleiotropic effect sizes and exposure-outcome confounding, although sufficiently large outcome GWAS sample sizes were important for stable inference and limited prior dependence. In real data analyses, CaPri reported low posterior probabilities of causality across negative controls and strong support for established positive control relationships. Application to alcohol consumption and coronary artery disease further demonstrated its ability to identify potentially meaningful pleiotropic instruments. CaPri provides a complementary framework for strengthening causal inference and triangulation across MR methods with different identifying assumptions. Availability and implementation: CaPri is available at https://github.com/OuRen0104/CaPri.

Polygenic scores (PGSs) have progressed toward clinical utility in recent years, demonstrating odds ratios that are comparable with both lifestyle-related and monogenic risk factors for several complex diseases. However, PGS implementation relies on developing analytically and clinically valid assays. Methods for developing polygenic models (PGMs) that incorporate tens of thousands to millions of variants require either genome-wide chip genotyping combined with imputation or whole genome sequencing, limiting their portability. Additionally, variants may be in regions where the local genomic sequence makes it difficult to assay, and multi-allelic or strand ambiguous variants can make bioinformatics more complicated. Because of these challenges, attempts to develop a reliable assay for published PGMs can produce multiple variations from the same underlying model, hindering routine clinical deployment and incorporation into multi-factorial risk prediction models. In this analysis, we present an adjustment to the previously published select and shrink with summary statistics (S4) PGM method designed to optimize performance for realistic clinical use by restricting to a pre-specified number of available SNPs that are well-captured by genotyping arrays. We compare the performance of this oligogenic S4 method (S4-Oligo) to currently published PGSs as well as the full S4 model for prediction of breast, colorectal, ovarian, and prostate cancers in the All of Us Research Program, Biobank Japan, and UK Biobank. We find that S4-Oligo matches or exceeds the accuracy of published PGMs at comparable model sizes while retaining between 88% and 99% of the performance of the original S4 model when limited to 1,000 variants.

As the use of population descriptors such as race, ethnicity, and ancestry have become increasingly common in modern genetics research, there have been growing calls to critically examine their use. Most notably, in 2023, the National Academies of Science, Engineering, and Medicine (NASEM) published a report titled Using Population Descriptors in Genetics and Genomics Research: A New Framework for an Evolving Field, which included eight specific and actionable recommendations for researchers to implement the ethical and accurate use of population descriptors in genetic research. Here, we use the 2023 NASEM report as a benchmark to analyze the use of population descriptors in genome-wide association studies (GWAS). We develop a general toolkit for large language model-based bibliometrics, operationalize the report's recommendations into an evaluation framework, and apply this framework to evaluate all 4,007 papers from the GWAS Catalog published between 2007 and 2025 with full text available on PubMedCentral. We find significant improvements in adherence to NASEM report recommendations over time. However, most improvements predate the publication of the NASEM report itself, suggesting the report functioned primarily as a synthesis of existing best practices rather than a catalyst for change. We conclude by highlighting opportunities for growth in the field of human genetics.

Background: Millions of textile and agricultural workers inhale organic dust containing endotoxin, an exposure associated with impaired respiratory function. However, the underlying biological pathways remain unclear. We conducted a two-cohort epigenome-wide association study (EWAS) to identify airborne endotoxin-associated DNA methylation (DNAm) alterations and pathways. Methods: In the discovery phase, we compared blood DNAm in 281 endotoxin-exposed cotton workers and 253 unexposed silk-workers from the Shanghai Textile Worker Study (STWS) measured using Illumina MethylationEPICv2 arrays. STWS endotoxin assessment combined detailed work histories and onsite measurements. For targeted parallel validation, we tested STWS signals in 686 incident lung cancer cases and 683 matched controls from the Shanghai Women's Health Study (SWHS), using occupation-derived low, medium, or high exposure-intensity ranks based on published evidence. To gain biological insight, we analyzed pathway enrichment, immune-cell composition, DNAm inflammation risk scores (IRSs), and respiratory phenotype associations in STWS. Results: In STWS, we identified 36 epigenome-wide significant DNAm sites (P<5e-8), of which, 35 were hypomethylated. Among 30 sites testable in SWHS, 18 had concordant effect directions, and four were nominally significant: cg11869499 (POLG; P=0.041), cg16290931 (TSKS; P=0.044), cg05019530 (EPIC1; P=0.042), and cg26033526 (PSMB9/TAP1; P=0.013). Enrichment implicated immune and signal-transduction pathways, including IgSF cell-adhesion-molecule signaling (FDR=8.2e-6;). Neutrophil were higher in cotton workers than in silk workers (63.5% vs 58.9%, p=7.52e-10). Cumulative endotoxin was associated with higher CRP-based IRSs, which showed the most consistent associations with lower lung function and airflow obstruction. Conclusions: Our EWAS findings suggest that occupational endotoxin may operate partly through immune-cell composition and inflammatory pathways.

Polygenic scores (PGS) predict an individual's genetic value for a complex trait of interest by aggregating effect size estimates from genome-wide association studies (GWAS). While powerful, this approach assumes that genetic variants affect the phenotype in an additive way, which might be problematic when dominance is part of the genetic architecture. To address this limitation, we developed PGSa,d, a simple polygenic scoring framework, rooted in quantitative genetics theory, that incorporates both additive (a) and dominance effects (d) directly. We evaluated PGSa,d against the standard GWAS-based PGS (PGS{beta}) in simulations. The results showed that PGSa,d consistently achieved lower root mean squared error (RMSE) and progressively higher R2 as dominance contributed more to the genetic architecture. We then applied the method to 16 blood cell traits, height, and body mass index (BMI) from UK Biobank. PGSa,d showed comparable R2 to PGS{beta} for all traits (including height and BMI whose architecture is predominantly additive) and significantly lower RMSE for four traits after strict Bonferroni correction. Across traits, RMSE improvement from PGSa,d was positively associated with the magnitude of the estimated dominance effects, whereby traits with larger dominance effects displayed lower RMSE from PGSa,d. As a byproduct of our analysis, we also identified 141 variants in 39 genomic regions with significant dominance effects after Bonferroni correction across traits, recapitulating known associations and identifying novel ones. Together, these results show that explicitly modeling additive and dominance effects is a robust strategy that can improve individual-level predictions, especially when dominance contributes to complex trait architecture.

AlphaGenome Atlas, released in September 2026, provides functional predictions for nearly all possible human single-nucleotide variants using the AlphaGenome Variant Impact (AVI) score. However, empirical thresholds relating AVI quantile rank (AVI_QUANTILE) to clinical variant classifications have not been established. We derived AVI_QUANTILE thresholds distinguishing ClinVar pathogenic, likely pathogenic, uncertain, conflicting, likely benign, and benign SNVs using receiver operating characteristic analyses. Pathogenic/likely pathogenic variants were strongly enriched at the upper end of the AVI_QUANTILE distribution, yielding a Youden-optimal ClinVar-derived cutoff of 0.9938. Optimal cutoffs were generally consistent across several major molecular consequence classes, although lower thresholds were observed for selected noncoding categories. Similarly, most well-powered gene-disease groups showed cutoffs close to the global threshold, with modest gene-specific variation. Among 24,989 variants of uncertain significance, 11,357 (45.45%) exceeded the global cutoff, identifying a substantial subset with AlphaGenome-predicted functional effects comparable to those observed among pathogenic/likely pathogenic variants and potentially warranting further evaluation. In a steroid-dependent asthma fine-mapping analysis, AVI_QUANTILE showed little correspondence with SuSiE posterior inclusion probabilities, indicating that AlphaGenome functional predictions capture information distinct from statistical fine-mapping.

Local genetic correlation is estimated as a ratio of a local genetic covariance to the geometric mean of two local heritabilities. LAVA, the standard implementation, returns that ratio only where both traits clear a univariate heritability test at P < 0.05, a rule its documentation presents as a stabiliser. We show that it is a selection event acting on the denominator, and characterise what it does to the estimates that survive. Under the noncentral Wishart model LAVA's own interval routine assumes, the gate-conditional mean of the estimator is affine in the true correlation, with slope and intercept available in closed form from quantities every run already reports, that is, the components retained, the two univariate P-values, and the sample-overlap correlation. Attenuation is therefore a property of the analysis, not of the method. Across six psychiatric trait pairs the measured slope runs from 0.579 to 0.923 and the closed form predicts each to within 0.007 with nothing fitted to the panel, while within a single pair it runs from 0.371 to 0.971, so no single factor can correct it. Overlap and the filter jointly manufacture correlation where the truth is zero, most where power is lowest. We invert the closed form into a gate-aware correction, which reduces mean absolute bias from 0.247 to 0.026 over 48 simulation cells, and pair it with a diagnostic that decides whether the correction is identified. Three pairs fail it, all three sharing one trait whose local heritabilities are indistinguishable from a complete local null.

Genetic association analyses of exome sequencing data pose distinct methodological challenges in small cohorts, which are common in rare disease research. Data filtering prior to association testing requires particular care, especially when analyzing datasets generated under different study designs. Most of current genetic association tools need to be integrated into complex data processing pipelines, whose implementation requires technical expertise and is prone to error. To alleviate mentioned issues we built a comprehensive, high performance Nextflow workflow for genetic association testing, which includes all commonly performed data filtering and preparation steps, controlled by a rich set of modifiable parameters. We compared our software against existing tools incorporating quality control and aggregated association testing steps and found our solution to offer the most complete suite of data preparation, association testing and reporting modules, making small-cohort association analyses accessible to teams with limited technical expertise.

Shotgun sequencing of DNA mixtures has the potential to become an important tool in forensic genetics, particularly for degraded samples or complex mixtures. However, the analysis of shotgun-sequenced mixtures is challenged by several confounding factors, including sequencing errors, population structure, the potential presence of relatives, linkage, and linkage disequilibrium. Additionally, the statistical methods currently available for such data are limited. Here, we develop likelihood ratios that summarize the evidence for the presence of a suspects DNA in a mixture, assuming genotypes of the suspect and the victim are available, while accounting for DNA from unknown contributors modeled either as general population contamination or as a single contaminating individual. We show that these likelihood ratios are not robust to model assumptions and can falsely support the presence of the suspects DNA when the true contributor is even a distant relative. We therefore propose a combined evidence score, {Lambda}: the likelihood ratio minimized over contamination models, reported only when the evidence favors the suspect over a hypothesized sibling by at least a factor of C (default C = 10) under at least one contamination model, and otherwise set to 1 (log {Lambda} = 0). In simulations based on 1000 Genomes data, {Lambda} retains nearly all power to detect true contributors at mixture proportions as low as 5% while eliminating the false positives produced by model mis-specification, population structure, and by untested relatives. The methods are implemented in the open-source program DNAMIXTURE, providing the forensic and population genetics community a resource for robustly identifying small individual contributions to a DNA mixture.

Human genetics needs benchmarks for evaluating variant scores against observed phenotypic consequences. We therefore present UKBBGym, which evaluates coding and non-coding scores against plasma protein abundance and quantitative traits observed in carriers of ultra-rare variants in the UK Biobank. This design avoids ascertainment biases of clinical labels, potential physiological mismatch of experimental assays, and confounding pertaining to common-variant associations. UKBBGym recapitulates relative performance on pathogenicity prediction while enabling comparison across variant classes and molecular mechanisms. This shows that scores designed for coding and splicing variants correlate more strongly with protein abundance and quantitative traits than scores modelling transcriptional regulation. Moreover, experimental assays do not consistently outperform computational scores for predicting missense effects. Comparing captured to detectable variance indicates scope for improvement, particularly for indels and loss-of-function variants. Finally, for coding variants, UKBBGym can be reproduced from public summary statistics, providing an accessible population-phenotype benchmark complementary to clinical labels and experimental assays.

Understanding the causal pathways linking the gut microbiome to downstream biomarkers and clinical outcomes is central to microbiome research. When the microbiome acts as the exposure, the high dimensionality and compositional structure of the data introduce unique methodological challenges for mediation analysis. To address these challenges, we propose a latent variable mediation framework that captures variation in microbial composition through a microbial balance, defined as the log-ratio between two unknown subsets of taxa. This balance serves as a latent scalar exposure and simplifies the estimation of the overall indirect effect at the community level, while simultaneously identifying specific taxa that contribute to the overall direct and indirect effects. We describe the model's estimation and inference, and illustrate the method using data from the Multiethnic Cohort--Adiposity Phenotype Study, where we examine the mediation pathway from the gut microbiome through lipopolysaccharide-binding protein, a marker of metabolic endotoxemia, to percent liver fat.

Background Staying physically active is essential for healthy ageing. Exposure to high levels of air pollutants such as particulate matter (PM) is associated with less active, more sedentary lifestyles in older adults. Whether results extend to typical fluctuations in daily PM exposure is unclear. We examined associations of higher-than-usual personal exposure to PM (PM2.5 and PM10) with next-day movement behaviour in older adults, assessed using concurrently worn personal air quality monitors and accelerometers. Methods The present study used 304 days of data from 67 adults aged 50-85 years living in the UK. For five days, participants wore accelerometers 24-hours per day and personal air quality monitors during waking hours. Average daily time spent in movement behaviours (moderate-to-vigorous physical activity [MVPA], light physical activity [LPA], sedentary behaviour [SB]) and step count was extracted from accelerometers, and average daily exposure to PM2.5 from personal air quality monitors. PM10 was examined in secondary exploratory analyses. We used linear mixed models to examine within-person associations of deviation from usual PM exposure with next-day movement outcomes, with estimates presented for 'usual' (individual mean) and 'higher-than-usual' (+1 interquartile range, equivalent to 2.6 g/m3 beyond individual mean PM2.5 exposure and 2.8 g/m3 beyond individual mean PM10 exposure) exposure. Analyses were repeated using area-based PM estimates. Results Compared with usual exposure, higher-than-usual PM2.5 was associated with 8 minutes less LPA (95% CI=-14 to -2) and 483 fewer steps (-855 to -82) on the following day. Exposure to higher than usual PM10 was associated with 8 minutes less LPA (-13 to -3), 8 minutes more SB (0 to 15), and 412 fewer steps (-739 to -85) on the following day. There was no association with MVPA, and no associations when area-level PM estimates were used as opposed to estimates from personal air quality monitors. Conclusions Typical daily variability in PM exposure in a setting with relatively low overall PM levels was associated with modest next-day differences in movement behaviour in older adults. Higher-than-usual PM exposure was associated with less next-day activity and more sedentary time. Personal monitoring may be important for capturing short-term exposure-behaviour associations. Trial registration N/A

Selection bias is a common concern in causal inference. Two mechanisms can give rise to selection bias: collider bias or stratifying on an effect modifier of the exposure-outcome causal effect. Research into the impact of selection bias on instrumental variable (IV) analyses has focused primarily on collider bias. Here, we show that IV estimates can be biased if selection is an effect modifier (or caused by an effect modifier) of the exposure-outcome effect. In a simulation study, we investigate the size and direction of the bias when estimating effects on the mean difference or odds ratio scale. Bias can also occur due to heterogeneity in the instrument-exposure relationship, although its magnitude is typically smaller. Finally, we discuss methods to deal with this bias in practice, including inverse probability weighting (IPW).

To evaluate the clinical implementation of long-read genome sequencing (lrGS) and establish the population reference resources needed to support its use in diagnostics, we integrate two complementary initiatives: Genome of Sweden (GoS), Sweden's contribution to Genome of Europe (GoE), and the Genomic Medicine Sweden - Rare Diseases (GMS-RD) long-read sequencing pilot. The initial GoS release of 576 individuals provides population-scale maps of genetic and epigenetic variation. Among 485 GMS-RD probands, 161 (33.2%) receive a molecular diagnosis, with lrGS providing added clinical value in 40 (24.8%). As a first-line test, lrGS could have avoided at least 42 follow-up analyses and an additional testing cycle in 7.2% of probands, reducing costs and the burden on patients. By integrating population and clinical long-read genomics, Sweden establishes a national ecosystem for genomic research and clinical diagnostics. Our findings support the clinical implementation of lrGS and offer a scalable model for national collaboration.

Background: Although sleep disorders are generally associated with cognitive decline, definitions of sleep conditions and cognitive function vary greatly across studies. Therefore, we evaluated associations between consensus definitions of sleep and cognitive outcomes across multiple studies. Methods: Consensus sleep measures were ascertained at baseline visits in three cohort studies (HCHS/SOL, MESA and ARIC, N=11,563): short and long sleep ([&le;]6hours and [&ge;]9hours), insomnia (Women's Health Initiative Insomnia Rating Scale [&ge;]10 or [&ge;]6 for 3 questions), daytime sleepiness (Epworth Sleepiness [&ge;]11) and obstructive sleep apnea (OSA, Apnea-Hypopnea Index [&ge;]15). Cognitive measures were ascertained at baseline and follow-up visits in each study and used to calculate harmonized outcomes: global cognitive function (GCF) at baseline, change in GCF between baseline and follow-up, poor GCF at baseline ([&ge;]1SD below the median), and incident poor GCF at follow-up. Study-specific analyses were adjusted for age, sex, center, education, race/ethnicity, APOE4 and language of cognitive assessments as needed. Results were meta-analyzed across studies. Results: Long sleep was consistently associated with lower baseline GCF ({beta}=-0.19, SE=0.03, p=3.78x10-14, FDR p=7.56x10-13, p-heterogeneity=0.78) and increased odds of poor baseline GCF (OR=1.77, 95% CI=1.49-2.09, p=2.96x10-11, FDR p=2.96x10-10, p-heterogeneity=0.22) across cohorts. Insomnia was also suggestively associated with lower baseline GCF ({beta}=-0.051, SE=0.02, p=0.01, FDR p=0.07, p-heterogeneity=0.005) and increased odds of poor baseline GCF (OR=1.22, 95% CI=1.06-1.41, p=6.76x10-3, FDR p=0.05, p-heterogeneity=0.03). Sleep measures were not associated with change in cognitive function. Discussion: Associations of long sleep and insomnia with harmonized cross-sectional GCF across three large, multiethnic cohorts suggest that cognitive function may be associated with distinct sleep phenotypes.

Amongst measures of linkage disequilibrium, the correlation measure r2, due originally to Sewall Wright, has several advantages. Although basically a haploid statistic, it can readily be estimated using covariance and correlation from diploid data. Its population expectation can also be found from probability methods. We show that the probability of linked identity by descent, LIBD, directly estimates the expectation for r2 for Wright-Fisher models. The simplest such model for two loci is the haploid model, where each haplotype is chosen independently from haplotypes of the previous generation. This gives an equilibrium between recombination (recombination frequency c) and drift (population size N) where the expected value of r2 is approximately 1/(1+4Nc). For diploid models, the expectation is similar for low values of c, but contains a term in c2 and rises to double the haploid value for the limiting value of c = 0.5. An increase in r2 is also expected for the model of monogamous mating, as opposed to random mating for each offspring. The expectation for r2 is unaffected by the use of diploid rather than haploid data, an exception being the case of monogamous mating. Computer simulation shows close agreement of r2 with probability expectations, with minor discrepancies observed for unlinked loci. Using the probabilistic definition, r2 can be intuitively understood as a parameter shaping the effective sample size in genome-wide association studies.

Background: The incidence of early-onset colorectal cancer (EOCRC), defined as CRC diagnosed age <50, is rising in the United States and is the leading cause of cancer death in among those with early-onset cancer. Epigenetic exposure signatures based on methylation risk scores (MRS) offer a novel approach to identify long-term exposures that may contribute to EOCRC risk. We sought to replicate findings from a recent study in a racially and geographically diverse CRC cohort. Methods: We computed weighted MRS across lifestyle, air pollution, and pesticide exposome traits in 146 participants from the Disparities and Cancer Epidemiology (DANCE) cohort, a diverse population-based CRC study, applying the Maas et al. (2026) framework to tumor tissue. Associations between MRS and age at CRC onset (EOCRC, n=29; intermediate-onset (IOCRC) 50-64 years, n=69; late-onset, [&ge;]65 years, n=48) were evaluated using linear and polytomous regression, with adjustment and stratification for race, geography, Area Deprivation Index (ADI), and Multi-Environmental Exposure Index (MEEI). Spearman correlations were calculated between MRSes and observed exposures. Results: EOCRC cases were predominately Black (65.5% vs. 31.3% among LOCRC) and resided in higher-deprivation neighborhoods. We found higher odds of EOCRC associated with epigenetic exposure signatures for 2,4-D herbicide, heptachlor, coarse particulate matter (PM), Mediterranean Diet Score, and smoking, and borderline associations for fine PM and nitrogen dioxide. IOCRC was associated with signatures for 2,4-D, lindane, and obesity. Heterogeneity was observed by race and area-level factors, though associations were largely consistent across strata. ADI and MEEI were correlated with multiple pesticide and pollutant epigenetic signatures. Conclusions: In a diverse US cohort, epigenetic exposure signatures distinguish CRC cohorts, particularly EOCRC from later-onset disease, with associations consistent with but not identical to previous research. EOCRC in DANCE was characterized by higher epigenetic exposure signatures for air pollution, smoking, and distinct pesticide exposure profiles. IOCRC had overlapping characteristics with early- and late-onset and were the only group with a significant obesity signature. These findings support the hypothesis that exposome-driven epigenetic changes contribute to EOCRC risk and highlight the importance of studying diverse populations to characterize the full spectrum of environmental risk factors.

Purpose A significant proportion of cases with rare inherited retinal disease (IRD) remain genetically unresolved following short-read whole genome sequencing (WGS). This study aimed to increase the diagnostic yield of two cohorts in a research setting. One cohort was previously screened by short-read WGS (n=120 cases), while the second cohort (n=28 cases) was not included in any previous short-read WGS studies. In contrast to the larger cohort, these cases only underwent either exome sequencing (ES) or single-molecule molecular inversion probe (smMIPs) sequencing as pre-screening in earlier studies. For all cases, we performed a stepwise, case by case reanalysis in which short-read WGS was generated (n=28 cases) and (re)analyzed short-read WGS data using an optimized approach. For a subset of cases (n=20) that remained unresolved, this was followed by long read WGS. Together, these steps aimed to improve the diagnostic yield in these cohorts. Methods Short-read WGS data of 148 IRD-cases were (re)analyzed using updated allele frequency databases, new variant caller and variant predictor tools, as well as updated and extended gene-panels to identify causal single nucleotide variants (SNVs) and structural variants (SVs). For 20 genetically unresolved cases with sufficient high-quality DNA available, long-read WGS was performed. Results The (re)analysis of short-read WGS data in the 120 previously WGS-screened unresolved cases resulted in a diagnostic yield of 15% (18/120). For the 28 individuals without prior WGS, performing WGS analysis within this study provided an additional yield of 32% (9/28). Furthermore, long-read WGS contributed two genetic diagnoses among the 20 long-read WGS-sequenced probands. Amongst other reasons, new genetic diagnoses were attributed to pathogenic variants in genes newly associated with IRDs and in genes that were not included in the gene panels applied in previous WGS studies. The Mobile Element Locator Tool facilitated the identification of two pathogenic Alu insertions in short-read WGS. Long-read WGS identified two additional pathogenic variants. Conclusion These results support periodic reanalysis of existing short-read WGS data as a key component of IRD genetic diagnostics, with selective long-read WGS representing a valuable additional approach for unresolved cases. Together, these complementary strategies offer a practical framework for improving molecular diagnosis and narrowing the remaining diagnostic gap in IRDs.

The rapid growth of technologies such as wearable devices and high-throughput sequencing has generated unprecedented volumes of longitudinal data, often with multiple correlated features. This trend highlights the need for robust clustering methods tailored to multi-feature longitudinal datasets. Although previous reviews have benchmarked clustering methods in this context, most evaluations relied heavily on simulated and overly simplified datasets, which substantially limit their interpretability and applicability to real-world data. Missing data is also ubiquitous in longitudinal data, yet understudied in the context of clustering. To address these gaps, we systematically evaluate clustering methods across three distinct types of real-world longitudinal datasets: (1) a high-dimensional, low-timepoint metagenomic study, (2) a low-dimensional, high-timepoint multi-site repeated-measure study, and (3) a high-dimensional, high-timepoint repeated-measure study. We further investigate how missing data and different imputation strategies influence clustering performance. In addition, we introduce a novel resampling framework based on the empirical cumulative distribution function (eCDF) and a copula. This framework enables us to simulate multi-feature longitudinal data that preserves the correlation structure of real datasets, while also allowing for controlled variations such as sample size. Our simulation framework naturally accommodates missing data, thereby providing a flexible tool for benchmarking and comparing clustering methods in more realistic longitudinal settings.