Abstract:Accurate prediction of cancer drug response (CDR) is a longstanding challenge in modern oncology that underpins personalized treatment. Current computational methods implement CDR prediction by modeling responses between entire drugs and cell lines, without the consideration that response outcomes may primarily attribute to a few finer-level 'subcomponents', such as privileged substructures of the drug or gene signatures of the cancer cell, thus producing predictions that are hard to explain. Herein, we present SubCDR, a subcomponent-guided deep learning method for interpretable CDR prediction, to recognize the most relevant subcomponents driving response outcomes. Technically, SubCDR is built upon a line of deep neural networks that enables a set of functional subcomponents to be extracted from each drug and cell line profile, and breaks the CDR prediction down to identifying pairwise interactions between subcomponents. Such a subcomponent interaction form can offer a traceable path to explicitly indicate which subcomponents contribute more to the response outcome. We verify the superiority of SubCDR over state-of-the-art CDR prediction methods through extensive computational experiments on the GDSC dataset. Crucially, we found many predicted cases that demonstrate the strength of SubCDR in finding the key subcomponents driving responses and exploiting these subcomponents to discover new therapeutic drugs. These results suggest that SubCDR will be highly useful for biomedical researchers, particularly in anti-cancer drug design. The ability to predict the response of a cancer patient to a therapeutic drug is a major goal in modern oncology. Computational methods have led to significant progress in predicting cancer drug response (CDR), but they ignore that response outcomes may be highly associated with specific 'subcomponents', such as privileged substructures of drugs or gene signatures of cancer cells. To understand the effect of subcomponents in CDR prediction, we present a subcomponents-guided deep learning method, named SubCDR. SubCDR is based on multiple deep neural networks capable of extracting functional subcomponents from the drug SMILES and cell line transcriptome, and decomposing the response prediction between the drug and cell line into identifying pairwise interactions between their respective subcomponents. With this interaction form, we can obtain a traceable path to explicitly identify subcomponents contributing significantly to the response outcome. We validate our method on the GDSC dataset and demonstrate improvements over the state-of-the-art CDR prediction methods. Furthermore, we show the strength of SubCDR in interpreting the CDRs and discovering new drugs with sensitive responses from the view of subcomponent interaction.

GraphCDR: a graph neural network method with contrastive learning for cancer drug response prediction

DeepCDR: a Hybrid Graph Convolutional Network for Predicting Cancer Drug Response

Improving anti-cancer drug response prediction using multi-task learning on graph convolutional networks

XGraphCDS: An explainable deep learning model for predicting drug sensitivity from gene pathways and chemical structures

GADRP: graph convolutional networks and autoencoders for cancer drug response prediction

MMCL-CDR: enhancing cancer drug response prediction with multi-omics and morphology images contrastive representation learning

TransCDR: a deep learning model for enhancing the generalizability of cancer drug response prediction through transfer learning and multimodal data fusion for drug representation

DualGCN: a dual graph convolutional network model to predict cancer drug response

DRN-CDR: A cancer drug response prediction model using multi-omics and drug features

Predicting cancer drug response using parallel heterogeneous graph convolutional networks with neighborhood interactions

Improving drug response prediction based on two-space graph convolution

A Novel Approach For Drug Response Prediction In Cancer Cell Lines Via Network Representation Learning

A subcomponent-guided deep learning method for interpretable cancer drug response prediction

TCR: A Transformer Based Deep Network for Predicting Cancer Drugs Response

RedCDR: Dual Relation Distillation for Cancer Drug Response Prediction

Enhancing drug and cell line representations via contrastive learning for improved anti-cancer drug prioritization

A novel heterogeneous network-based method for drug response prediction in cancer cell lines

Multi-Network Graph Contrastive Learning for cancer driver gene identification

Predicting Drug Response Based on Multi-Omics Fusion and Graph Convolution

An integrated network representation of multiple cancer-specific data for graph-based machine learning

TransCDR: a deep learning model for enhancing the generalizability of drug activity prediction through transfer learning and multimodal data fusion