cfDECOR: a novel approach for estimating cell type contributions to cfDNA based on chromatin accessibility patterns

Cell-free DNA (cfDNA) sequencing has emerged as a promising approach in liquid biopsy, offering minimally invasive methods for diagnosing and monitoring various diseases. Determining cfDNA tissue of origin can provide a clinically relevant signal, as the share of diseased tissue in the plasma increases under many conditions. Fragmentation-based biomarkers, as one class of epigenetic signals in cfDNA, have advantages, including the large number of tissue-specific regions across the genome and the lack of requirements for high-depth sequencing. This study introduces a novel deconvolution model, cfDECOR, for predicting the contribution of different cell types in cfDNA on the basis of chromatin accessibility profiles obtained from tissue/cell type-specific ATAC-seq (assay for transposase-accessible chromatin using sequencing) data. This study provides proof-of-concept evidence that deconvolution of cfDNA based on tissue/cell type-specific chromatin accessibility patterns can improve the accuracy and sensitivity of disease detection.