Characterising a large bovine lactating mammary ATAC-seq dataset

<title>Abstract</title> <p> Background Many studies of genome-wide transcriptional regulation, or expression quantitative trait loci (eQTL), have been performed in multiple species to help improve our understanding of gene expression and regulation. However, many regulatory elements in the DNA sequence that control gene expression remain poorly known in agricultural species and humans alike. Here, we present a large-scale study using the Assay for Transposase Accessible Chromatin using Sequencing (ATAC-seq) method to detect chromatin accessibility QTLs (caQTL) in the lactating bovine mammary gland, and examine their correlations with gene expression to identify sequence motifs associated with higher or lower expression. Results Sequencing ATAC-seq libraries made from lactating mammary biopsies of 199 cows, we discovered 119,318 chromatin accessibility peaks, with 114,910 located on the autosomes. Across all samples, the average fraction of reads in peaks (FRiP) score was 0.493, above the ENCODE recommended threshold. Using high-depth mammary RNA-seq data representing duplicate biopsies of the same cows, significant ( <italic>p &lt;</italic>  1 × 10 ^− 6 ) positive correlations with peak height were seen with the expression of 1805 genes, and negative correlations with 793 genes. Significant <italic>cis</italic> -caQTL were observed for 19,339 peaks, and <italic>cis</italic> eQTL for 5003 genes. Positive correlations between eQTL and caQTL marker effects were observed for 6379 peaks, alongside 5356 for negative correlations, indicating the involvement of chromatin accessibility in the regulation of gene expression. Prediction of transcription factor binding sites in sequences under chromatin accessibility peaks showed differences in transcription factor frequencies between peaks with positive or negative effects on expression, with more STAT-domain containing factors in positive peaks, and more CTCF in negative peaks and insulator elements that decrease expression by blocking nearby enhancer elements. Conclusion These findings highlight mechanistically diverse modulation of gene expression in the lactating mammary gland and demonstrate the benefit of combining multiple omics datasets to elucidate the genetic control of gene expression in dairy cattle. </p>