A generalizable framework to comprehensively predict epigenome, chromatin organization, and transcriptome

Machine and deep learning methods for predicting 3D genome organization

Three-Dimensional (3D) chromatin interactions, such as enhancer-promoter interactions (EPIs), loops, Topologically Associating Domains (TADs), and A/B compartments play critical roles in a wide range of cellular processes by regulating gene expression. Recent development of chromatin conformation capture technologies has enabled genome-wide profiling of various 3D structures, even with single cells. However, current catalogs of 3D structures remain incomplete and unreliable due to differences in technology, tools, and low data resolution. Machine learning methods have emerged as an alternative to obtain missing 3D interactions and/or improve resolution. Such methods frequently use genome annotation data (ChIP-seq, DNAse-seq, etc.), DNA sequencing information (k-mers, Transcription Factor Binding Site (TFBS) motifs), and other genomic properties to learn the associations between genomic features and chromatin interactions. In this review, we discuss computational tools for predicting three types of 3D interactions (EPIs, chromatin interactions, TAD boundaries) and analyze their pros and cons. We also point out obstacles of computational prediction of 3D interactions and suggest future research directions.

Population-scale skeletal muscle single-nucleus multi-omic profiling reveals extensive context specific genetic regulation

Skeletal muscle, the largest human organ by weight, is relevant to several polygenic metabolic traits and diseases including type 2 diabetes (T2D). Identifying genetic mechanisms underlying these traits requires pinpointing the relevant cell types, regulatory elements, target genes, and causal variants. Here, we used genetic multiplexing to generate population-scale single nucleus (sn) chromatin accessibility (snATAC-seq) and transcriptome (snRNA-seq) maps across 287 frozen human skeletal muscle biopsies representing 456,880 nuclei. We identified 13 cell types that collectively represented 983,155 ATAC summits. We integrated genetic variation to discover 6,866 expression quantitative trait loci (eQTL) and 100,928 chromatin accessibility QTL (caQTL) (5% FDR) across the five most abundant cell types, cataloging caQTL peaks that atlas-level snATAC maps often miss. We identified 1,973 eGenes colocalized with caQTL and used mediation analyses to construct causal directional maps for chromatin accessibility and gene expression. 3,378 genome-wide association study (GWAS) signals across 43 relevant traits colocalized with sn-e/caQTL, 52% in a cell-specific manner. 77% of GWAS signals colocalized with caQTL and not eQTL, highlighting the critical importance of population-scale chromatin profiling for GWAS functional studies. GWAS-caQTL colocalization showed distinct cell-specific regulatory paradigms. For example, a C2CD4A/B T2D GWAS signal colocalized with caQTL in muscle fibers and multiple chromatin loop models nominated VPS13C, a glucose uptake gene. Sequence of the caQTL peak overlapping caSNP rs7163757 showed allelic regulatory activity differences in a human myocyte cell line massively parallel reporter assay. These results illuminate the genetic regulatory architecture of human skeletal muscle at high-resolution epigenomic, transcriptomic, and cell state scales and serve as a template for population-scale multi-omic mapping in complex tissues and traits.

ChromaFold predicts the 3D contact map from single-cell chromatin accessibility

The identification of cell-type-specific 3D chromatin interactions between regulatory elements can help to decipher gene regulation and to interpret the function of disease-associated non-coding variants. However, current chromosome conformation capture (3C) technologies are unable to resolve interactions at this resolution when only small numbers of cells are available as input. We therefore present ChromaFold, a deep learning model that predicts 3D contact maps and regulatory interactions from single-cell ATAC sequencing (scATAC-seq) data alone. ChromaFold uses pseudobulk chromatin accessibility, co-accessibility profiles across metacells, and predicted CTCF motif tracks as input features and employs a lightweight architecture to enable training on standard GPUs. Once trained on paired scATAC-seq and Hi-C data in human cell lines and tissues, ChromaFold can accurately predict both the 3D contact map and peak-level interactions across diverse human and mouse test cell types. In benchmarking against a recent deep learning method that uses bulk ATAC-seq, DNA sequence, and CTCF ChIP-seq to make cell-type-specific predictions, ChromaFold yields superior prediction performance when including CTCF ChIP-seq data as an input and comparable performance without. Finally, fine-tuning ChromaFold on paired scATAC-seq and Hi-C in a complex tissue enables deconvolution of chromatin interactions across cell subpopulations. ChromaFold thus achieves state-of-the-art prediction of 3D contact maps and regulatory interactions using scATAC-seq alone as input data, enabling accurate inference of cell-type-specific interactions in settings where 3C-based assays are infeasible.