Segmenting Proteins into Tripeptides to Enhance Conformational Sampling with Monte Carlo Methods

Laurent Denarie (1) , Ibrahim Al Bluwi (1) , Marc Vaisset (2) , Thierry Simeon (1) , Juan Cortés (1)

1 LAAS-RIS - Équipe Robotique et InteractionS
2 LAAS-IDEA - Service Informatique : Développement, Exploitation et Assistance

This paper presents an approach to enhance conformational sampling of proteins employing stochastic algorithms such as Monte Carlo (MC) methods. The approach is based on a mechanistic representation of proteins and on the application of methods originating from robotics. We outline the general ideas of our approach and detail how it can be applied to construct several MC move classes, all operating on a shared representation of the molecule and using a single mathematical solver. We showcase these sampling techniques on several types of proteins. Results show that combining several move classes, which can be easily implemented thanks to the proposed approach, significantly improves sampling efficiency.

Mots clés: proteins; Monte Carlo; conformational sampling; robotics-inspired approach

Domaines: Robotique [cs.RO]; Bio-informatique [q-bio.QM]; Biochimie [q-bio.BM]; Chimie théorique et/ou physique

Laurent Denarie, Ibrahim Al Bluwi, Marc Vaisset, Thierry Simeon, Juan Cortés. Segmenting Proteins into Tripeptides to Enhance Conformational Sampling with Monte Carlo Methods. Molecules, 2018, 23 (2), pp.373. ⟨10.3390/molecules23020373⟩. ⟨hal-01708710⟩