An AI framework can identify targets for CAR-T cell therapy in weeks, streamlining a process that often takes months to years, according to a study published June 25 in Cell.
Here are three notes from the study:
- Researchers at Philadelphia-based Penn Medicine Abramson Cancer Center built the framework using single-cell RNA sequencing datasets and validated it in melanoma, leukemia, lung and colorectal cancers.
- The framework uses large language models to screen and rank thousands of potential CAR-T targets, then repeats simulations 1,000 times to reduce the risk of AI errors.
- Researchers said the framework can be applied to any cancer type or other conditions, using different datasets or future LLMs, according to a June 25 news release from Penn Medicine.
Read the full study here.
