Structural predictions of protein–DNA binding: MELD-DNA

Reza Esmaeeli¹, Antonio Bauzá², Alberto Perez¹

1. Department of Chemistry, Quantum theory project, University of Florida, Gainesville, FL 32611, USA
2. Department of Chemistry, Universitat de les Illes Balears, Palma de Mallorca (Baleares), 07122, Spain

Proteins play a crucial role in maintaining a healthy and functional genome by interacting with DNA. However, our understanding of the structural basis of these interactions is limited. In this study, we propose a new computational method called MELD-DNA that uses a combination of molecular dynamics simulations and general knowledge or experimental information to predict the structures of protein-DNA complexes. Our approach is designed to take into account sequence-dependent properties and conformational changes required for binding, and it can a) sample multiple binding modes, b) identify the preferred binding mode, and c) provide qualitative binding preferences between DNA sequences. We evaluate the performance of MELD-DNA on a dataset of 15 protein-DNA complexes and compare it to other state-of-the-art methods. Additionally, we demonstrate the sequence dependence effects of binding in MELD predictions for three selected complexes. MELD-DNA can provide insights into relative binding affinity through competitive binding as well. We believe that the freely available software and the results presented in this study will have a significant impact on the field of structural biology by complementing DNA structural databases and on the study of molecular recognition by providing new insights into the mechanisms underlying protein-DNA interactions.

This work is published in Nucleic Acids Research: doi.org/10.1093/nar/gkad013
I'd also be open to presenting it in the chemical biology/biochemistry division as it has practical applications and open to potential collaborations.
In case there is no possibility of giving a talk, I'm open to a poster presentation as well.