Leveraging Whole-Exome Sequencing and Mutational Signatures to Detect Homologous Recombination Deficiency in Cancer

Homologous recombination is a high-fidelity DNA repair mechanism essential for maintaining genome stability. Impairment of this pathway, often due to BRCA1 or BRCA2 inactivation, leads to homologous recombination deficiency (HRD), forcing cells to rely on error-prone mechanisms for repairing DNA double-strand breaks, such as nonhomologous or microhomology-mediated end joining. HRD is a clinically important biomarker, particularly in breast and ovarian cancers, as it predicts responsiveness to platinum-based chemotherapies and PARP inhibitors. However, current tests in the clinical setting, mostly based on targeted panel sequencing, lack specificity and lead to a substantial number of false positives. In contrast, whole-genome sequencing, despite its high accuracy, remains largely confined to research because of high costs and logistical constraints. In this issue of Cancer Research, Abbasi and colleagues present HRProfiler, a machine learning–based tool that accurately detects HRD using whole-exome sequencing (WES) data, which is increasingly accessible in clinical oncology. Notably, it demonstrates improved sensitivity in the WES setting compared with existing tools, such as HRDetect and SigMA. As WES continues to gain traction, HRProfiler offers a promising step toward democratizing HRD detection and enabling more precise, genomics-guided treatment strategies. This article is part of a special series: Driving Cancer Discoveries with Computational Research, Data Science, and Machine Learning/AI. See related article by Abbasi et al., p. 2504