Sensitive detection of copy number alterations in low-pass liquid biopsy sequencing data.

Liquid biopsies, coupled with analysis of copy number alterations (CNAs), have emerged as a promising tool for non-invasive monitoring of cancer progression and tumor composition. However, methods utilizing CNA data from liquid biopsies are limited by the low signal in the samples, caused by a low percentage of cancer DNA in the blood, and inherent noise introduced in the sequencing. To address this challenge, we developed BayesCNA, a method designed to improve signal extraction from low-pass liquid biopsy sequencing data, by utilizing a Bayesian changepoint detection algorithm. We use information of the posterior changepoint probabilities to identify likely changepoints, where a changepoint indicates a shift in the copy number state. The signal is then reconstructed using the identified partition. We show the effectiveness of the method on synthetically generated datasets and compare the method with state-of-the-art bioinformatics tools under noisy conditions. Our results show that this novel approach increases sensitivity in detecting CNAs, particularly in low-quality cases.