Short talk: Compartmap: Direct inference of higher-order chromatin structure in individual cells from single-cell RNA-seq & single-cell ATAC-seq

Compartmap: Direct inference of higher-order chromatin structure in individual cells from single-cell RNA-seq & single-cell ATAC-seq

Benjamin K Johnson,Jean-Philippe Fortin,Kasper Daniel Hansen,Hui Shen,Tim Triche Van Andel Institute

Abstract

Single-cell profiling of higher-order chromatin structure remains a challenge due to cost, throughput, and resolution. We introduce compartmap as a method to reconstruct higher-order chromatin domains in individual cells, inferred from single-cell transcriptomic and epigenomic assays. In multiple cell lines and primary human samples, compartmap infers higher-order chromatin structure at least as well as chromatin capture or proximity ligation based methods, while distinguishing clinically relevant structural alterations in single cells. Using multi-omic profiling approaches such as CITE-seq, we can assess the contributions of cell-to-cell interactions in shaping both normal and dysfunctional cellular epigenomes, as well as the downstream consequences of clonal structural variants. Moreover, the relative “surprise” value for a given assay (compared to predictions based solely on higher-order chromatin structure) provides a rough metric for the contextual value of any given assay. Compartmap thus provides a lens to integrate transcriptional and epigenomic results, by linking higher-order chromatin architecture to gene regulation and phenotype across single-cell assays.

Keywords: chromatin,4dn,HiC,structural variants,epigenetics,single cell,rnaseq,atacseq,epigenomics