Scientific Advances: CRISPR screens empower machine learning to predict cancer patient response to precision medicine

In an important advancement for precision oncology, researchers at the Center for Cancer Research in the US have shown how their tumor analysis pipeline, called SELECT, was 80% successful in predicting cancer patient responses to targeted therapy in over 30 clinical trials.

SELECT stands for SynthEtic LEthality and rescue-mediated precision onCology via the Transcriptome. Using machine learning, this tool analyzes both DNA and RNA from tumors to identify synthetic lethal interactions The inclusion of RNA analysis can identify vulnerabilities not readily evident by tumor profiling using standard DNA panels. The authors started by looking at a pools of synthetic lethal drug targets used for precision medicine based on published genome wide RNA interference and CRISPR screens in cancer cell lines or identified through the application of small molecule inhibitors. The authors plan to include DNA methylation information in the tool and further improve predictions in the next few years.

For more information, see: Lee, J.S., et al. (2021) Synthetic lethality-mediated precision oncology via the tumor transcriptome. Cell

Keywords: CRISPR screen, synthetic lethality, machine learning

Questions? Email: