Each variant is scored to quantify the deleteriousness

The hypothesis whether GRIPT test is the deleterious mutation loads of a disease-causing gene is significantly higher in the case cohort than in the control cohort. To quantify the deleteriousness of variants, in this study, we applied Combined Annotation Dependent Depletion (CADD v1.3) score to each variant of each gene in every individual [19]. CADD score is an integrative score derived from the integration of diverse annotations and is highly predictive of molecular functionality and pathogenicity [19]. Higher CADD score indicates more deleteriousness of the mutation. In addition, CADD not only provides integrative prediction scores for SNVs but also for INDELs which are missing for most other variant effect prediction tools. We further normalized the variant score on a scale of 0 to 1 as s = 1 − 10^−C/10. C is the PHRED-like scaled C-score as described in CADD. Moreover, CADD score can be easily replaced by any other score that users provide in order to better predict the variant’s deleteriousness. To test the reliability and robustness of the statistic test framework of GRIPT, the ranked REVEL [30] and DANN [29] scores were also applied as the variant scores respectively. The CADD score (v1.3) was downloaded from https://cadd.gs.washington.edu/download. The ranked DANN score was extracted from dbNSFP3.4a downloaded from https://sites.google.com/site/jpopgen/dbNSFP. The ranked REVEL score was downloaded from https://sites.google.com/site/revelgenomics/downloads [38, 39].