TABLE 1.

Four major genomic feature groups characterizing the SNVs in gold standard datasets.

Feature group	Description	Example
Structural and genomic context features	Characterizing sequence attributes of the mutation location. These features estimate the disruption in the mutations surrounding sequence both in coding and non-coding regions	Percentage of GC in a ±75 bp window
Epigenetic features	Describing epigenetic changes such as histone modifications and methylation alterations	Maximum H3K4 methylation level from Encode
Genomic distance features	Measuring the distance between a given SNV and critical functional and structural genomic elements such as transcription start and end sites	Minimum distance to Transcribed Sequence Start (TSS)
Genomic conservation features	Measuring the evolutionary conservation at the mutation alignment sites in an effort to help the training models learn the relationships between the measurements and pathogenicity of the SNVs	Scores from PhastCons and Phylop