Also known as: Proceedings of the National Academy of Sciences
MIT chemical engineers used a large language model to optimize protein production in industrial yeast. The AI model improved yields for five of six tested proteins, including a cancer monoclonal antibody, potentially reducing drug development costs.
Studies in muscle-invasive bladder cancer found ctDNA predicted metastatic risk after bladder-sparing treatment, while urine tumor DNA was more sensitive for residual bladder disease. Data from RETAIN-2 and a PNAS study support molecular testing to help select patients for bladder preservation.
Scientists at three institutions have developed distinct nanoparticle-based drug delivery systems designed to target cancer tumors more precisely while reducing side effects, with approaches ranging from DNA nanomachines to transformable nanofibers.
CAR-T cell therapy engineering advances at Tandem 2026 focus on reducing toxicity and improving access, while McGill researchers develop targeted nanoparticle immunotherapy for lymph node metastases with fewer side effects.
