Comprehensive Human Cell-Type Methylation Atlas Allows Identification of Origins of Circulating Cell-Free DNA in Health and Disease

Fragments of cell-free DNA (cfDNA) circulating in the blood are a rich source of information about cell death events taking place inside the human body, with important implications for diagnostic medicine. Tissue-specific DNA methylation markers can identify cfDNA originating from specific tissues, allowing for accurate detection of cell death from specific cell types.
Here, we present an extensive reference atlas of methylomes from 34 human tissues and cell types, and develop computational deconvolution algorithms for an unbiased and sensitive analysis of plasma cfDNA methylation patterns.
We demonstrate the applicability of our approach by extensive simulations and by quantifying the cellular contributors to cfDNA of multiple donors. As we show, cfDNA of healthy donors (young/old, female/male) typically originate from blood cells (89%), vascular endothelial cells (10%) and hepatocytes (1%). Deconvolution of cfDNA from patients reveals tissue contributions that agree with clinical findings in sepsis, pancreatic islet transplantation, colorectal cancer and patients with cancer of unknown primary source (CUP).
Our results suggest that analysis of methylation patterns, from small blood volumes and possibly over a small subset of genomic regions, allows for specific and sensitive detection of cfDNA from various tissues, can be used for early detection of cancer and other diseases, and opens a broad window into healthy and pathologic human tissue dynamics.