Researchers built a publicly available tumor organoid biobank covering 256 clinically annotated models across colorectal, oesophageal, ovarian, pancreatic, and gastric cancers. The dataset links whole-genome and transcriptome profiles to gene-dependency maps derived from genome-wide CRISPR–Cas9 screens. The resource aims to overcome limitations of traditional cell lines by preserving patient context and enabling paired comparisons across tumor evolution, including pre- and post-treatment samples. In colorectal cancer, functional tests of the EGFR–RAS–MAPK axis showed differential effects by KRAS variant allele. By combining multi-omics with dependency vulnerabilities and making the resource open, the study strengthens the translational bridge between genomics and precision oncology targeting—particularly for rare subtypes where cell-line representation is often weak.