Comprehensive, high-resolution binding energy landscapes reveal context dependencies of transcription factor binding

Abstract

SignificanceTranscription factors (TFs) are key proteins that bind DNA targets to coordinate gene expression in cells. Understanding how TFs recognize their DNA targets is essential for predicting how variations in regulatory sequence disrupt transcription to cause disease. Here, we develop a high-throughput assay and analysis pipeline capable of measuring binding energies for over one million sequences with high resolution and apply it toward understanding how nucleotides flanking DNA targets affect binding energies for two model yeast TFs. Through systematic comparisons between models trained on these data, we establish that considering dinucleotide (DN) interactions is sufficient to accurately predict binding and further show that sites used by TFs in vivo are both energetically and mutationally distant from the highest affinity sequence.

Deep neural networks identify context-specific determinants of transcription factor binding affinity

“…Furthermore, these features and their importance for binding may vary greatly across cell types, loci, or development timepoints 1 . Studies have shown that many of these features are related to sequence context in the immediate vicinity of the TF motif itself [14][15][16] , implying that TF binding may be predicted directly from sequence information.…”

Mapping DNA sequence to transcription factor binding energy in vivo

“…Massively parallel reporter assays (MPRAs) have made it possible to read out transcription factor binding position and occupancy in vivo with base-pair resolution, and provide a means for analyzing additional features such as "insulator" sequences [6][7][8]. In vitro methods based on protein-binding microarrays [9], SELEX [10][11][12], MITOMI [13][14][15], and binding assays performed in high-throughput sequencing flow cells [16,17] have made it possible to measure transcription factor affinity to a broad array of possible binding sites and can also account for features such as flanking sequences [15,18,19]. However, in vitro methods cannot fully account for the in vivo consequences of binding site context and interactions with other proteins.…”

Paralogs in the PKA regulon traveled different evolutionary routes to divergent expression in budding yeast

“…This would be strengthened by a tighter link between pathway mutants and transcription factor binding via traditional assays such as ChIP-seq and complementary methods such as the calling card method (Mayhew and Mitra, 2016). It would also be strengthened by incorporating the fuller understanding of the links between transcription factors and the DNA sequences they bind, both in vitro and in vivo (Le et al, 2018;Samee et al, 2019;Zhu et al, 2018). A combination of all such methods will allow us to more fully understand the role that changes in gene regulation play in evolution.…”