Guiding the Search for Native-like Protein Conformations with an Ab-initio Tree-based Exploration

Abstract

In this paper we propose a robotics-inspired method to enhance sampling of native-like conformations when employing only aminoacid sequence information for a protein at hand. Computing such conformations, essential to associating structural and functional information with gene sequences, is challenging due to the high-dimensionality and the rugged energy surface of the protein conformational space. The contribution of this paper is a novel two-layered method to enhance the sampling of geometrically distinct low-energy conformations at a coarse-grained level of detail. The method grows a tree in conformational space reconciling two goals: (i) guiding the tree towards lower energies; and (ii) not oversampling geometrically similar conformations. Discretizations of the energy surface and a low-dimensional projection space are employed to select more often for expansion low-energy conformations in under-explored regions of the conformational space. The tree is expanded with low-energy conformations through a Metropolis Monte Carlo framework that uses a move set of physical fragment configurations. Testing on sequences of eight small-to-medium structurally diverse proteins shows that the method rapidly samples native-like conformations in a few hours on a single CPU. Analysis shows that computed conformations are good candidates for further detailed energetic refinements by larger studies in protein engineering and design.

Protein conformational search with geometric projections

“…Our group recently proposed a probabilistic search framework, FeLTr, which employs a geometric projection layer to ensure structural diversity during conformational search [24], [25], [28], [29]. Conformations sampled during the search are projected onto a low-dimensional geometric space where they can be efficiently grouped based on geometric similarity.…”

“…A different search template has recently been proposed by the Shehu lab [24], [25]. Instead of exploring the conformational space through independent trajectories and relying on analysis aposteriori, the search framework in [24] conducts a tree-based exploration of the protein conformational space.…”

“…Instead of exploring the conformational space through independent trajectories and relying on analysis aposteriori, the search framework in [24] conducts a tree-based exploration of the protein conformational space. Analysis over lower-dimensional embeddings of the explored conformational space and its energy surface is conducted during the exploration in order to guide the tree towards low-energy regions of the energy surface and away from over-populated regions of the conformational space.…”

“…EC is incorporated into our recently published FeLTr framework which uses a geometric projection layer to ensure structural diversity during a conformational search [24], [25], [28], [29]. Section III shows that, for many target proteins, EC is as effective as our previous approach at biasing sampling towards nearnative conformations and is able to significantly improve the average lRMSD sampled for several proteins.…”

Rigid‐CLL: Avoiding constant‐distance computations in cell linked‐lists algorithms

“…Larger rigid clusters appear when running rigid-body molecular dynamics simulations [8,18], or when considering simplified models tailored to simulate large-scale conformational transitions in biological macromolecules [7,20]. Moreover, there is a growing interest in Monte Carlo algorithms applied to protein structure prediction and protein design that, at every step, replace some protein segments with one selected from a database [4,31]. In this case, a rigid subset is replaced by another and, thus, large portions of the molecule are considered as rigid bodies.…”