CoRE-ATAC: A Deep Learning model for the functional Classification of Regulatory Elements from single cell and bulk ATAC-seq data

Abstract

Cis-Regulatory elements (cis-REs) include promoters, enhancers, and insulators that regulate gene expression programs via binding of transcription factors. ATAC-seq technology effectively identifies active cis-REs in a given cell type (including from single cells) by mapping accessible chromatin at base-pair resolution. However, these maps are not immediately useful for inferring specific functions of cis-REs. For this purpose, we developed a deep learning framework (CoRE-ATAC) with novel data encoders that integrate DNA sequence (reference or personal genotypes) with ATAC-seq cut sites and read pileups. CoRE-ATAC was trained on 4 cell types (n=6 samples/replicates) and accurately predicted known cis-RE functions from 7 cell types (n=40 samples) that were not used in model training (mean average precision=0.80). CoRE-ATAC enhancer predictions from 19 human islet samples coincided with genetically modulated gain/loss of enhancer activity, which was confirmed by massively parallel reporter assays (MPRAs). Finally, CoRE-ATAC effectively inferred cis-RE function from aggregate single nucleus ATAC-seq (snATAC) data from human blood-derived immune cells that overlapped with known functional annotations in sorted immune cells, which established the efficacy of these models to study cis-RE functions of rare cells without the need for cell sorting. ATAC-seq maps from primary human cells reveal individual- and cell-specific variation in cis-RE activity. CoRE-ATAC increases the functional resolution of these maps, a critical step for studying regulatory disruptions behind diseases.Author Summary Non-coding DNA sequences serve different functional roles to regulate gene expression. For these sequences to be active, they must be accessible for proteins and other factors to bind in order to carry out a specific regulatory function. Even so, mutations within these sequences or other regulatory events may modulate their activity or regulatory function. It is therefore critical that we identify these non-coding sequences and their specific regulatory function to fully understand how specific genes are regulated. Current sequencing technologies allow us to identify accessible sequences via chromatin accessibility maps from low cell numbers, enabling the study of clinical samples. However, determining the functional role associated with these sequences remains a challenge. Towards this goal, we harnessed the power of deep learning to unravel the intricacies of chromatin accessibility maps to infer their associated gene regulatory functions. We demonstrate that our method, CoRE-ATAC, can infer regulatory functions in diverse cell types, captures activity differences modulated by genetic mutations, and can be applied to accessibility maps of single cell clusters to infer regulatory functions of rare cell populations. These inferences will further our understanding of how genes are regulated and enable the study of these mechanisms as they relate to disease.Competing Interest StatementThe authors have declared no competing interest.View Full Text