Controlled Generation of Cancer Drug-Resistance Data through Barcoded Variant Profiling

Predictive models of cancer evolution require systematic experimental data linking defined genetic variants to drug-response trajectories. However, longitudinal datasets generated under controlled genetic backgrounds that link defined driver variants, drug exposures, population trajectories, and acquired genomic alterations remain limited. Here, we present a barcoded, doxycycline (Dox)-inducible expression platform in human induced pluripotent stem cells (iPSCs) for generating longitudinal variant-drug response data. As a proof of concept, we constructed 20 cancer-associated TP53 variant lines, together with wild-type (WT) and mock controls, each carrying a variant-specific genomic barcode at a defined genomic locus. Pooled selection with the topoisomerase I inhibitor topotecan revealed time-dependent enrichment of three TP53 variants, and individual colony-formation assays independently validated variant-dependent differences in drug resistance. The platform is designed to scale across genetic variants, drug conditions, and cellular contexts and can be coupled with whole-genome sequencing of resistant clones to characterize secondary genomic alterations. This controlled experimental framework provides a foundation for generating training and benchmark datasets for predictive models of cancer evolution and drug resistance.