AI Meets Oncology: Tailoring Cancer Treatments with MMRDetect

Researchers from the University of Cambridge have developed a new algorithm called MMRDetect to identify tumors sensitive to specific immunotherapies. This tool uses data from the 100,000 Genomes Project, analyzing thousands of NHS cancer patient samples to detect "mismatch repair deficiencies" in tumors. The study identified nine critical DNA repair genes that protect the genome from damage caused by oxygen, water, and cell division errors. Using CRISPR-Cas9, the team knocked out these genes in healthy stem cells to observe mutational signatures, which serve as biomarkers for repair pathway defects. These signatures can help determine which treatments are most effective for individual patients. The MMRDetect algorithm was trained on whole genome sequencing data to identify tumors with mismatch repair deficiency, making them sensitive to checkpoint inhibitors. This breakthrough is a result of collaboration with the 100,000 Genomes Project, demonstrating its value in precision medicine. The algorithm could transform cancer treatment by personalizing therapies based on genetic weaknesses in tumors. It is planned to be rolled out across all cancers identified by Genomics England. The research highlights the potential of genomic data to improve patient outcomes by tailoring treatments to specific genetic vulnerabilities.