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, mean F1 score = 0.70). 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 cellswithout 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.

Inferring the selective history of CNVs using a maximum likelihood model

“…For this purpose, we utilized ATAC-seq data to identify the open and closed chromatin segments (Daugherty et al, 2017; Jänes et al, 2018; Thibodeau et al, 2021). The ATAC-seq peak data from C. elegans (Jänes et al, 2018), includes 42,245 peaks with an average width of 152 bp.…”

maxATAC: Genome-scale transcription-factor binding prediction from ATAC-seq with deep neural networks

“…maxATAC is a unique resource, representing the largest collection of deep CNN TFBS prediction models for chromatin accessibility and a first-time benchmark of trans-cell type prediction on scATAC-seq. There are numerous deep CNN methods for ATAC-seq data [21,[66][67][68][69][70], but these methods have different modeling objectives, including: prediction of ATAC-seq PLOS COMPUTATIONAL BIOLOGY signal from DNA sequence (e.g., AI-TAC [68], chromBPNet [66], deepMEL [67]), denoising of ATAC-seq and scATAC-seq signal (AtacWorks [70]), classification of enhancers, promoters and insulators based on ATAC-derived inputs (CoRE-ATAC [69]) and clustering of scATACseq data (scFAN [21]).…”

Chromatin Structure and Dynamics: Focus on Neuronal Differentiation and Pathological Implication

“…The coupling of different techniques allows the visualization of different activating or repressive histone marks correlated with the chromatin opening state [97,98].…”