Extended Data Fig. 8: Power comparison between a Cochran‐Mantel‐Haenszel (CMH) and beta-binomial approach.
From: Dietary stress remodels the genetic architecture of lifespan variation in outbred Drosophila
(A-B) The proportion of simulated true positives (that is, SNPs with a significant effect on lifespan) detected by a beta-binomial versus a CMH approach as a function of effect size. Different p-value thresholds are used to detect true positive SNPs in panels A and B as noted on the y-axis. Details on the simulations and analyses of simulated data are provided in the SI Methods (see Text S5. Comparison of two statistical approaches to test for allele frequency changes with age). Correlation between the -log10 p-value for a genetic effect on lifespan from the real data, estimated in (C) CTRL or (D) HS conditions using a beta-binomial model (x-axis) or a CMH test (y-axis). As described in the main text, the p-values for the beta-binomial were estimated while controlling for covariates of sex, sequencing batch, and cage. R values from a Pearson correlation are reported on each plot. Plots were created using the “smoothScatter” function in R with default parameters.
