Fig. 10

The performance of GRIPT and CMC on detecting genes with modest-effect variants. The AR model was tested with 0.5%, 1%, 2%, and 3% of patients concurrently carrying the modest-effect variants (p.Asn1868Ile and p.Gly863Ala) of ABCA4, respectively. The patient cohort size was 600. The control cohort size was 5000. The performance of GRIPT and CMC are shown in red and blue, respectively. a The ranking of ABCA4 under the AR model. b The power of the two tests for ABCA4. c The number of significant autosomal candidate genes under the AR model. The rankings of ABCA4 generated by GRIPT were compared to those generated by CMC with one-tailed WRST. The methods that generated significantly worse ranking than GRIPT were marked with “*” if p value < 0.05, “**” if p value < 0.01, and “***” if p value < 0.001