Using interpretable deep learning to model cancer dependencies

Motivation: Cancer dependencies provide potential drug targets. Unfortunately, dependencies differ among cancers and even individuals. To this end, visible neural networks (VNNs) are promising due to robust performance and the interpretability required for the biomedical field. Results: We design Biological visible neural network (BioVNN) using pathway knowledge to predict cancer dependencies. Despite having fewer parameters, BioVNN marginally outperforms traditional neural networks (NNs) and converges faster. BioVNN also outperforms an NN based on randomized pathways. More importantly, dependency predictions can be explained by correlating with the neuron output states of relevant pathways, which suggest dependency mechanisms. In feature importance analysis, BioVNN recapitulates known reaction partners and proposes new ones. Such robust and interpretable VNNs may facilitate the understanding of cancer dependency and the development of targeted therapies.

Automated summary

BioVNN, utilizing pathway knowledge, outperforms traditional neural networks in predicting cancer dependencies with faster convergence, and offers explainable dependency predictions by correlating with neuron output states.