Researchers published a biologically informed deep‑learning framework that improves prediction of drug response in breast cancer by integrating molecular features with drug representations. The model was presented as a tool to prioritize therapeutic combinations and interpret compound activity across tumor subtypes. Separately, Exact Sciences reported early results showing its Oncodetect MRD test identified triple‑negative breast cancer (TNBC) patients at higher risk of distant recurrence after surgery; postoperative ctDNA positivity correlated with shorter distant recurrence‑free intervals. Together, the studies underscore growing momentum for AI‑driven pharmacology and circulating tumor DNA assays to guide personalized oncology strategies.