medRxivJClub
Deriving LD-adjusted GWAS summary statistics through linkage disequilibrium deconvolution
medRxiv · · Preprint · Open access
Nouira, Favre Moiron + 2 more
Abstract ↗AI summary
The abstract is read at the publisher; the summary is JClub's.
LDeconv uses truncated SVD to deconvolve LD and adjust GWAS summary statistics, reducing false discoveries while preserving true associations in UK Biobank analyses.
- Why it matters: LD can make non-causal variants look associated, especially for highly polygenic traits where causal signals saturate the genome, inflating false positives.
- What they did: The method deconvolves LD from summary statistics without individual-level genotypes, and its performance was evaluated via simulations across multiple LD scenarios plus sensitivity analyses.
- The result: In simulations and on UK Biobank GWAS data, LD-adjusted estimates from LDeconv reduce false discoveries while keeping true associations, enabling more reliable post-GWAS analyses such as Mendelian Randomization.
The findingWhy it mattersWhat they didThe result
AI summary of the abstract · may contain errors · not medical advice · read the paper.
- Open access