Fig. 1From: RegSNPs-intron: a computational framework for predicting pathogenic impact of intronic single nucleotide variantsWorkflow. a Data sources for model building. Pathogenic and neutral iSNVs were collected from HGMD and 1000 Genomes Project, respectively. b Three categories of features, namely RNA splicing, protein structure, and evolutionary conservation, were considered for the iSNVs. Data sources and the numbers of these features are listed on the side. For additional details for the data sources, refer to the “Methods” section. c Logic flow for prediction model development, as well as model evaluation and result validationBack to article page