Ligand binding is one of the primary means of inducing conformational changes in protein. In general, proteins are dynamic entities that adapt their structure to the presence of a given ligand. An accurate prediction of a ligand binding pose should take this adaptation into account, both in the conformation of the binding site residues of the protein and their surrounding water molecules. These structural changes are complex, involving the coupled motion of many degrees of freedom, and are difficult to incorporate into standard docking algorithms. As there are many atomic positions, side chain orientations, and solvent degrees of freedom to consider, enumeration of all of the possible binding site conformations becomes impossible due to a combinatorial explosion. Molecular dynamics is useful for exploring binding site conformations, but predictions of globally stable poses can be challenging due to large free energy barriers between poses. The absolute binding free energy of a particular pose can be computed by alchemical techniques, where ligand-protein interactions are gradually turned off. These techniques are challenging to implement and are highly computationally expensive, as each of the ligand binding modes must be considered separately. We have recently developed a new approach for pose prediction called “Flexible Topology”, which allows the chemical attributes of atoms (<i>e.g.</i>, partial charges, atom size) to change over time, in response to forces from their environment. We can predict poses by running “assembly” simulations in the protein binding pocket, in which particles are initialized with random positions and chemical properties and are guided to assemble into a target molecule as the simulation progresses. By running an ensemble of assembly simulations we can gain information about the relative probabilities of different binding poses. Here, we predict ligand bound poses for a set of bromodomain-ligand systems, focusing on the first bromodomain of BRD2 human protein, which is part of the “bromodomains and extra-terminal motif” (BET) family and is involved in cancer pathogenesis. We have conducted a set of 24 flexible topology assembly simulations each for four ligands in the BRD2-BD1 binding pocket. To validate these poses, we will measure performance by comparing to an ensemble of conformations from molecular dynamics simulations. The performance of flexible topology will also be compared to poses from docking methods.

Flexible topology simulations to predict poses of bromodomain binding ligands

Where thousands of chemistry professionals meet to share ideas and advance scientific and technical knowledge. Spring 2023 abstracts, posters, and presentations will launch at the end of March.

ACS Spring 2023

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.