LAST: Latent Space Assisted Adaptive Sampling for Protein Trajectories

Abstract

Molecular dynamics (MD) simulation is widely used to study protein conformations and dynamics. However, conventional simulation suffers from being trapped in some local energy minima that are hard to escape. Thus, most computational time is spent sampling in the already visited regions. This leads to an inefficient sampling process and further hinders the exploration of protein movements in affordable simulation time.The advancement of deep learning provides new opportunities for protein sampling.Variational autoencoders are a class of deep learning models to learn a low-dimensional representation (referred to as the latent space) that can capture the key features of the input data. Based on this characteristic, we proposed a new adaptive sampling method, latent space assisted adaptive sampling for protein trajectories (LAST), to accelerate the exploration of protein conformational space. This method comprises cycles of (i) variational autoencoders training, (ii) seed structure selection on the latent space and (iii) conformational sampling through additional MD simulations. The proposed approach is validated through the sampling of four structures of two protein systems: two metastable states of E. Coli adenosine kinase (ADK) and two native states of Vivid (VVD). In all four conformations, seed structures were shown to lie on the boundary of conformation distributions. Moreover, large conformational changes were observed in a shorter simulation time when compared with conventional MD (cMD) simulations in both systems. In metastable ADK simulations, LAST explored two transition paths toward two stable states while cMD became trapped in an energy basin. In VVD light state simulations, LAST was three times faster than cMD simulation with a similar conformational space.

PASSer2.0: Accurate Prediction of Protein Allosteric Sites Through Automated Machine Learning

“…The past decade has witnessed the rapid development of machine learning in chemistry and biology (Zhang et al 2020;Chen L. et al 2021;Tian et al 2020;Tian et al 2021b;Tian et al 2022). ML methods have been shown to be superior in the classification of protein allosteric pockets.…”

Assessments of Variational Autoencoder in Protein Conformation Exploration

“…[12][13][14] Many enhanced sampling methods have been developed to address this issue, which in general fall into two categories: collective variable based (such as metadynamics 14,15 and variationally enhanced sampling 16,17 and collective variable free methods (such as replica exchange molecular dynamics 18,19 and integrated tempering sampling 20,21 ). 12,22,23 The deep-learning autoencoders models 24 present a powerful nonlinear dimensionality reduction technique to mine data-driven collective variables from MD trajectories [25][26][27][28][29][30][31][32] . This technique furnishes explicit and differentiable expressions for the highly abstract and differentiable collective variables, making it the ideal candidate for integration with enhanced sampling techniques to accelerate the exploration in the proteins configurational space.…”