Table 1.

Summary Results of top scoring Variant selection by VarClass per dataset.

Dataset	Diseasea	Unique genes	Risk Modelsb	Protective Models (IDI score < = −0.02)c	Pro-disease Models (IDI score > = 0.02)c	No. of top scoring VarClass Variantsc
GSE7226 Platform: GPL2005	Intellectual disability	2486	16551	82		82
GSE7226 Platform: GPL2004	Intellectual disability	2676	18992	51		51
GSE58356	Gastric cancer	9256	15272		12	10
PPMI	Parkinson’s	20264	6253		25	22

^aDisease keywords are used in stages 1 and 2 of the VarClass flow Chart (see Fig. 1) to initiate the disease profile and extract information from ClinVar.

bRisk Models are derived from stages 3–8 of the VarClass pipeline, where each model represents a different type of network used to obtain information of VUS by gene association.

^cProtective and Pro-disease models derived after filtering the results from step 7 (final outcome) of VarClass by applying IDI Thresholds and selecting a final number of Protecting and Pro-disease models and their associated variants.