MultiCapsNet An interpretable deep learning classifier integrate data from multiple sources

MultiCapsNet is a deep learning model built on CapsNet and scCapsNet, which possesses the merits of both easier data integration and higher model interpretability. By performing the labeled variant call dataset analysis, MultiCapsNet shows a similar performance with the previous feed forward neural network model with much less parameters. Furthermore, MultiCapsNet provides importance scores for each data source directly without an extra importance determination step required by the previous model.