DeepFusion-CDR Dataset：A multimodal benchmark dataset for anticancer drug response prediction

It integrates drug–target interaction information, drug molecular structural features, and multi-omics profiles of cancer cell lines. The data were collected and processed from public resources including GDSC2, CCLE/DepMap, CTRPv2, TTD, ChEMBL, and BindingDB.

The dataset contains drug response records measured by IC50, standard SMILES representations, molecular fingerprints, molecular graph structures, drug target activity profiles, and multi-omics features of cancer cell lines, including gene expression, genomic mutation, and DNA methylation data.