Joint sequence & chromatin neural networks characterize the differential abilities of Forkhead transcription factors to engage inaccessible chromatin.

The DNA-binding activities of transcription factors (TFs) are influenced by both intrinsic sequence preferences and extrinsic interactions with cell-specific chromatin landscapes and other regulatory proteins. Disentangling the roles of these determinants in TF-DNA binding remains challenging. For instance, the FoxA subfamily of Forkhead domain (Fox) TFs are known pioneer factors, yet their binding varies across cell types, pointing to a combination of intrinsic and extrinsic forces guiding their binding. How such sequence and chromatin influences vary across related Fox TFs remains mostly uncharacterized. Here, we present a principled approach to compare the relative contributions of intrinsic DNA sequence preference and cell-specific chromatin environments to a TF's DNA-binding activities. We over-express a selection of Fox TFs in mouse embryonic stem (mES) cells, which offer a platform to contrast each TF's binding activity within the same preexisting chromatin background. By applying a convolutional neural network that jointly models sequence and chromatin data, we evaluate how sequence and preexisting chromatin features contribute to induced TF binding, both at individual sites and genome-wide. We demonstrate that Fox TFs (FoxA1, FoxC1, FoxG1, FoxL2, and FoxP3) bind different DNA targets, and drive differential gene expression patterns, even when induced in identical chromatin settings. Differential Fox binding activities can be attributed to distinct DNA-binding preferences coupled with differential abilities to engage relatively inaccessible chromatin. We propose that a combination of divergent sequence preferences and varying preferences for preexisting chromatin states enables the functional diversification of paralogous TFs.