Abstract
The democratization of genome sequencing for genetic diseases is leading to the identification of a large amount of variants in noncoding regions. Unless supported by a strong clinically oriented diagnostic hypothesis, these variants remain largely under-analyzed due to the limited availability of in silico prediction tools for prioritization and the lack of functional assays for validation. We discuss here the current state of whole-genome analysis using the combined annotation-dependent depletion (CADD) score, one of the most efficient genome-wide prediction and most popular prioritization tool for genetic variants. In light of the worldwide participative ClinVar database that stores the disease classification of millions of human genetic variants, we evaluated the use of genomic region-specific thresholds to guide the geneticists in prioritizing noncoding region variants using the CADD score.