Show Notes
️ Episode 154: Multiple-testing corrections in selection scans using identity-by-descent segments
In this episode of PaperCast Base by Base, we explore how Temple and Browning develop principled genome-wide significance thresholds for IBD-based scans of recent positive selection by explicitly modeling correlation along the genome.
Study Highlights:
The authors model standardized IBD-rate scan statistics as an Ornstein–Uhlenbeck process and derive both an analytical threshold and a fast simulation-based alternative that control the family-wise error rate while adapting to genetic-map spacing and autocorrelation. In extensive coalescent simulations, the approach achieves approximate FWER control and shows that Bonferroni can be overly conservative or dependent on arbitrary test spacing, whereas OU-based thresholds better reflect the effective number of tests. Power analyses indicate more than 50% power for hard sweeps with selection coefficients around or above 0.01 when the sweeping allele is at intermediate present-day frequency, and negligible power for weaker or nearly fixed sweeps. Applying the framework to TOPMed and UK Biobank cohorts, they identify a limited set of genome-wide significant loci across ancestry groups and show that some large cross-ancestry signals concentrate near structural-variant–rich regions rather than reflecting recent adaptation.
Conclusion:
OU-based multiple-testing corrections make IBD selection scans more reliable by calibrating significance to genomic correlation, improving reproducibility and reducing false positives in large biobank-scale analyses.
Reference:
Temple SD, Browning SR (2025). Multiple-testing corrections in selection scans using identity-by-descent segments. The American Journal of Human Genetics 112:1–21. https://doi.org/10.1016/j.ajhg.2025.09.004
License:
This episode is based on an open-access article published under the Creative Commons Attribution 4.0 International License (CC BY 4.0) – https://creativecommons.org/licenses/by/4.0/
Support:
If you'd like to support Base by Base, you can make a one-time or monthly donation here: https://basebybase.castos.com/
