DNAffinity: A Machine-Learning Approach to Predict DNA Binding Affinities of Transcription Factors

Barissi, Sandro; Sala, Alba; Wieczór, Miłosz; Battistini, Federica; Orozco, Modesto

doi:10.1101/2022.07.26.501522

Cited by 1 publication

(1 citation statement)

References 58 publications

(110 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Not only the dynamics of the protein but also the flexibility of the conformation of the DNA plays a relevant role in the identification of the binding site. Molecular dynamics of such complexes have recently been used to predict the binding affinity of TFs and to predict its corresponding PWMs 67 . We used homology modelling in the same line in our approach.…”

Section: Discussionmentioning

confidence: 99%

Structure-based learning to model complex protein-DNA interactions and transcription-factor co-operativity incis-regulatory elements

Fornés

Meseguer

Aguirre‐Plans

et al. 2022

Preprint

View full text Add to dashboard Cite

Transcription factor (TF) binding is a key component of genomic regulation. There are numerous high-throughput experimental methods to characterize TF-DNA binding specificities. Their application, however, is both laborious and expensive, which makes profiling all TFs challenging. For instance, the binding preferences of ~25% human TFs remain unknown; they neither have been determined experimentally nor inferred computationally. Here, we introduce ModCRE, a web server implementing a structure homology-modelling approach to predict TF motifs and automatically model higher-order TF regulatory complexes. Starting from a TF sequence or structure, ModCRE predicts a set of motifs for that TF. The predicted motifs are then used to scan the DNA for occurrences of each of them, and the best matches are either profiled with a binding score or collected for their subsequent modeling into a higher-order regulatory complex with DNA, as well as other TFs and co-factors. Moreover, we demonstrate that incorporating high-throughput TF binding data, such as from protein binding microarrays, addresses the protein-DNA structure scarcity problem for deriving statistical potentials. In turn, these statistical potentials are proven to be capable predictors of TF motifs. We also show the conditional advantage of using ModCRE over a nearest-neighbor approach for predicting TF binding sites as well as an improvement in prediction accuracy when using a rank-enrichment selection system. Finally, as case examples, we apply ModCRE to model the interferon beta enhanceosome and the complex of SOX2 and 11 with a nucleosome.

show abstract

Section: Discussionmentioning

confidence: 99%