A set cover algorithm identifies minimal circulating tumour DNA sequencing targets for colorectal cancer detection

Abstract

Advances in high-throughput sequencing and decreasing costs have made cell-free DNA sequencing a promising approach for cancer detection. Sequencing assays require high read depth to detect low-frequency somatic mutations, so cell-free DNA panels must support deep sequencing while still assaying broadly enough to detect as many malignancies as possible. We developed OPTIC (Oncogene Panel Tester for Identifying Cancers), a pipeline employing a set cover algorithm, to identify the minimal set of genomic targets capturing the maximal proportion of tumours. Using three cohorts totalling 2,940 colorectal cancer samples, OPTIC was utilized to design a targeted sequencing panel spanning just 10,975 bases across APC, TP53, KRAS, BRAF, NRAS, PIK3CA, CTNNB1, RNF43, and ACVR2A. Collectively, these loci contain pathogenic mutations in 96.3% of cases. Our pipeline enables compact panel design without compromising sample coverage. This enables higher throughput, greater sequencing depth, and lower costs per-sample in early colorectal cancer detection from cell-free DNA.