Exome-wide rare-variant associations vary across disease trajectories

Genetic studies of chronic disease typically collapse longitudinal clinical histories into binary case-control outcomes, obscuring disease timing and differences in genetic association patterns across disease transitions. Time-to-event analysis can preserve this information, but rare-variant analysis of post-diagnosis progression is challenged by sparse carrier-events and extensive censoring. We developed GATE-GENE, a scalable gene-based framework that maintains calibration under these conditions. Across 300 onset phenotypes in UK Biobank (UKB), GATE-GENE identified 107 exome-wide significant germline associations involving 89 genes and 67 diseases. Across 95 multi-state trajectories comprising baseline (T0), an upstream disease state (T1) and a downstream disease state (T2), disease onset (T0→T1) and subsequent progression (T1→T2) showed limited detectable sharing of rare-variant association signals. Progression showed greater concordance of gene-level association signals with downstream phenotypes (T0→T2) when the analyses shared T2 events occurring after T1. Removing overlapping downstream events reduced the median correlation between gene-level association signals from 0.063 to 0.001, after accounting for background correlation. Null simulations reproduced the relationship between event overlap and concordance without genetic effects, whereas additional simulations showed that near-zero concordance could persist despite extensive underlying causal-gene sharing. This attenuation was reproduced across three ancestry groups in All of Us. Cross-biobank meta-analysis identified 12 additional exome-wide significant progression associations beyond three UKB discoveries, although clinical sensitivity analyses suggested that the strongest PKD1 signals largely reflected underlying cystic kidney disease. Together, these findings establish GATE-GENE for investigating rare-variant associations with disease timing and progression, while demonstrating how overlapping clinical events and limited carrier-event information complicate inference about genetic sharing across disease transitions.

Authors

Institutions

Publication Details

Journal
medRxiv
Published
2026-11-16
DOI
https://doi.org/10.64898/2026.11.14.26368035
Primary Topic
Genetic Associations and Epidemiology
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Exome-wide rare-variant associations vary across disease trajectories

Satoshi Koyama, Benjamin M. Neale, Ida L. Surakka, Connor Dowd et al.
medRxiv
Genetic Associations and Epidemiology
preprint

Exome-wide rare-variant associations vary across disease trajectories

Satoshi Koyama, Benjamin M. Neale, Ida L. Surakka, Connor Dowd, Ying Wang, Wei Zhou, Wenhan Lu, Pradeep Nataranjan, Mark Daly, Tony Chen, Cindy Wen, Wenjian Bi, Weijun Luo, Linfeng Hu, Alicia Martin
preprint en

Abstract

Genetic studies of chronic disease typically collapse longitudinal clinical histories into binary case-control outcomes, obscuring disease timing and differences in genetic association patterns across disease transitions. Time-to-event analysis can preserve this information, but rare-variant analysis of post-diagnosis progression is challenged by sparse carrier-events and extensive censoring. We developed GATE-GENE, a scalable gene-based framework that maintains calibration under these conditions. Across 300 onset phenotypes in UK Biobank (UKB), GATE-GENE identified 107 exome-wide significant germline associations involving 89 genes and 67 diseases. Across 95 multi-state trajectories comprising baseline (T0), an upstream disease state (T1) and a downstream disease state (T2), disease onset (T0→T1) and subsequent progression (T1→T2) showed limited detectable sharing of rare-variant association signals. Progression showed greater concordance of gene-level association signals with downstream phenotypes (T0→T2) when the analyses shared T2 events occurring after T1. Removing overlapping downstream events reduced the median correlation between gene-level association signals from 0.063 to 0.001, after accounting for background correlation. Null simulations reproduced the relationship between event overlap and concordance without genetic effects, whereas additional simulations showed that near-zero concordance could persist despite extensive underlying causal-gene sharing. This attenuation was reproduced across three ancestry groups in All of Us. Cross-biobank meta-analysis identified 12 additional exome-wide significant progression associations beyond three UKB discoveries, although clinical sensitivity analyses suggested that the strongest PKD1 signals largely reflected underlying cystic kidney disease. Together, these findings establish GATE-GENE for investigating rare-variant associations with disease timing and progression, while demonstrating how overlapping clinical events and limited carrier-event information complicate inference about genetic sharing across disease transitions.

medRxiv
Broad Institute (US), Georgia Institute of Technology (US), Amgen (United States) (US), Harvard University (US), University of Michigan (US), Institute of Computing Technology (CN), Michigan Medicine (US)
Genetic Associations and Epidemiology
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon