Design and Statistical Analysis of Pooled Next Generation Sequencing for Rare Variants
Figure 2
Empirical power at a level of 5% as a function of the difference of the allele frequency between cases and controls for the proposed PB test based on various estimates of the covariance matrix for testing multiple rare variants under the unbalanced design. The allele frequency of controls was 0.01; the sample size was set at 500 cases and 500 controls; the error rates (e) were set at 0, 0.005, and 0.01; the depths of coverage were set at 10× and 20×, and the numbers of pools (r) were set at 5 and 10. Figure (a) shows the power of the PB test based on an identity covariance matrix (I), the PB test based on the shrinkage estimate of the covariance matrix (S) and the single-variant test with Bonferroni correction the number of variants for independent variants (min). Figure (b) shows the power of the PB test based on the empirical estimate of the covariance matrix (E), the PB test based on the shrinkage estimate of the covariance matrix (S) and the single-variant test with Bonferroni correction (min) the number of variants for independent variants.
