European Commission logo
italiano italiano
CORDIS - Risultati della ricerca dell’UE

Dynamical Redesign of Biomolecular Networks

Periodic Reporting for period 3 - BioNet (Dynamical Redesign of Biomolecular Networks)

Periodo di rendicontazione: 2020-07-01 al 2021-12-31

This project includes three main objectives. First and foremost, to go beyond the state-of-the-art, we develop novel enhanced sampling algorithms that enable more efficient biomolecular simulations. Biological processes take place on a much longer timescale than current molecular simulations can be afforded. Therefore, to be able to study these key biomolecular processes, enhanced sampling algorithms are necessary. Here, our objective is to use modern statistical mechanics tools to address this problem and enable our multiscale simulations to reach the required timescales. In addition, novel methods developed will be beneficial to the molecular modelling community.

Secondly, we aim to understand how mechanistic changes affect catalytic reactions. Here we focus on proton transfer coupled phosphate transfer reactions as our main objective. However, to understand these complex processes, we aim to create model systems that can be studied without the need to include the biomolecular catalytic machinery. Here we use host-guest systems, such as cucurbituril-based (CB) macrocycles as molecular hosts, with catalyzed reactions including Diels-Alder reactions and the electrocatalytic reduction of CO2.

Our main goal is to understand the catalytic reaction mechanisms in key phosphate catalytic enzymes. To this end, we currently investigate several examples of phosphate catalytic enzymes to identify important structural and mechanistic aspects shared by these systems. We aim to use these underlying principles to design catalytic systems aiding biomedical applications.
Method development:
We developed methods aiding the understanding underlying the kinetic network for conformational changes in biomolecular systems. We worked in the framework of stochastic simulations that can be discretized and clustered to obtain the kinetically most representative reduced network of lower dimensionality. These types of methods help to capture long timescale behavior by building a tractable network which is computationally more efficient to study. We derived novel relationships between mean first passage times, kinetic rates, and variational properties of dynamical coarse graining approaches.
We also developed novel statistical methods to speed up Monte Carlo simulations using a framework based on irreversible samplers.
We published several joint experimental-computational studies addressing host-guest ligand complexation and catalysis. Here, one of the examples includes CO2 reduction on gold surface by electrocatalysis, aided via host-guest complexation with CBs. We also analyzed key factors in the rate enhancement for Diels-Alder reactions catalyzed by CB complexation. Here, our computational tools using multiscale quantum classical methods were able to reproduce the catalytic effects accurately for various substrates.
Biological applications:
We have demonstrated that accurate quantum-classical multiscale simulations can reveal the mechanistic details of several phosphate catalytic reactions of biomedical interest. In collaboration with the Cherepanov group at the Crick Institute, we demonstrated that computational methods can help distinguish first and second-generation viral drugs in wild type and drug resistant mutant HIV-1 integrase.
Our group made significant progress, developing novel methods and theoretical relationships. Our joint results have already proved useful for various experimental collaborations. We now aim to apply these in the design for phosphate catalytic reactions together with experimental collaborations to achieve our main goals until the end of this project.
First and second generation drugs bound in the HIV integrase active site