A team led by Prof.Andrew TESCHENDORFF from the Shanghai Institute of Nutrition and Health (SINH), Chinese Academy of Sciences, has now cracked this problem in  a recent work published in Nature Methods.

The researchers elegantly demonstrate how it is possible to generate representative DNA methylation profiles for all main cell-types in a tissue without doing any experiments, by predicting these DNA methylation values from single-cell RNA-Seq data using a validated statistical prediction model. They use this validated strategy to build the first DNA methylation atlas for 13 tissue-types and 40 cell-types. This resource is freely available to the epigenomics community.

According to Prof. TESCHENDORFF this study is important because “This DNA methylation atlas meets an urgent need for a general pan-tissue wide resource that enables research groups all over the world to better interpret epigenome-wide data generated in complex tissues, such as colon, lung, brain or skin”. With this DNA methylation atlas it will be possible to better understand which cell-types change and how they are altered in association with disease risk factors such as age or obesity.

Ideally one would measure DNA methylation at single-cell resolution in hundreds if not thousands of individuals to accurately capture the heterogeneity at the cellular and human population levels, but such studies are at present too costly. Instead, this resource could be used to infer DNA methylation changes at cell-type resolution from 100s to 1000s of individuals without the need to do single-cell sequencing in any of them.

In their study, the researchers demonstrate how the DNA methylation atlas can be used to identify different cells of origin of specific cancer-types, or to identify neuron-specific alterations in schizophrenia. “The appealing aspect of our resource is its general applicability to most tissue-types”, says the lead author.

This work was published online under the title “A pan-tissue DNA methylation atlas enables in silico decomposition of human tissue methylomes at cell-type resolution” on 11^th March 2022. It was funded by the Chinese Academy of Sciences and National Natural Science Foundation of China. The datasets analyzed in the manuscript are from open databases.

The image depicts a globe with continents representing organs/tissues, and with DNA methylation "lollipop" marks depicted on double-stranded DNA, itself depicted as a key. Cell-types within each organ are shown. The image intends to convey the generation of a DNA m-atlas for different tissues/organs. (Image provided by Prof.Andrew TESCHENDORFF’s group)