Atomistic or ab initio molecular dynamics simulations are widely used to predict thermodynamics and kinetics relate them structure. A common approach go beyond the time- length-scales accessible with such computationally expensive is definition of coarse-grained models. Existing coarse-graining approaches define an effective interaction potential match defined properties high-resolution models experimental data. In this paper, we reformulate as a supervised machine learning problem. We use...

Machine learning (ML) is transforming all areas of science. The complex and time-consuming calculations in molecular simulations are particularly suitable for a machine revolution have already been profoundly impacted by the application existing ML methods. Here we review recent methods simulation, with particular focus on (deep) neural networks prediction quantum-mechanical energies forces, coarse-grained dynamics, extraction free energy surfaces kinetics generative network approaches to...

The definition of reaction coordinates for the characterization a protein-folding has long been controversial issue, even "simple" case in which one single free-energy barrier separates folded and unfolded ensemble. We propose general approach to this problem obtain few collective by using nonlinear dimensionality reduction. validate usefulness method characterizing folding landscape associated with coarse-grained protein model src homology 3 as sampled molecular dynamics simulations....

The associative memory, water mediated, structure and energy model (AWSEM) is a coarse-grained protein force field. AWSEM contains physically motivated terms, such as hydrogen bonding, well bioinformatically based local biasing term, which efficiently takes into account many-body effects that are modulated by the sequence. When combined with appropriate or global alignments to choose memories, can be used perform de novo prediction. Herein we present prediction results for particular choice...

We present a multiscale method for the determination of collective reaction coordinates macromolecular dynamics based on two recently developed mathematical techniques: diffusion map and local intrinsic dimensionality large datasets. Our accounts variation molecular configuration space, resulting global are correlated with time scales motion. To illustrate approach, we results model systems: all-atom alanine dipeptide coarse-grained src homology 3 protein domain. provide clear physical...

The long-timescale dynamics of macromolecular systems can be oftentimes viewed as a reaction connecting metastable states the system. In past decade, various approaches have been developed to discover collective motions associated with this dynamics. corresponding variables are used in many applications, e.g., understand mechanism, quantify system's free energy landscape, enhance sampling path, and determine rate. review we focus on number key developments field, providing an overview...

Characterizing macromolecular kinetics from molecular dynamics (MD) simulations requires a distance metric that can distinguish slowly interconverting states. Here, we build upon diffusion map theory and define kinetic for irreducible Markov processes quantifies how conformations interconvert. The be computed given model approximates the eigenvalues eigenvectors (reaction coordinates) of MD operator. employ time-lagged independent component analysis (TICA). TICA components scaled to provide...

With the rapid increase of available data for complex systems, there is great interest in extraction physically relevant information from massive datasets. Recently, a framework called Sparse Identification Nonlinear Dynamics (SINDy) has been introduced to identify governing equations dynamical systems simulation data. In this study, we extend SINDy stochastic which are frequently used model biophysical processes. We prove asymptotic correctness infinite limit, both original and projected...

Molecular dynamics simulations provide a mechanistic description of molecules by relying on empirical potentials. The quality and transferability such potentials can be improved leveraging data-driven models derived with machine learning approaches. Here, we present TorchMD, framework for molecular mixed classical All force computations including bond, angle, dihedral, Lennard-Jones, Coulomb interactions are expressed as PyTorch arrays operations. Moreover, TorchMD enables simulating neural...

Coarse graining enables the investigation of molecular dynamics for larger systems and at longer timescales than is possible atomic resolution. However, a coarse model must be formulated such that conclusions we draw from it are consistent with would finer level detail. It has been proven force matching scheme defines thermodynamically coarse-grained an atomistic system in variational limit. Wang et al. [ACS Cent. Sci. 5, 755 (2019)] demonstrated existence limit use supervised machine...

Abstract A generalized understanding of protein dynamics is an unsolved scientific problem, the solution which critical to interpretation structure-function relationships that govern essential biological processes. Here, we approach this problem by constructing coarse-grained molecular potentials based on artificial neural networks and grounded in statistical mechanics. For training, build a unique dataset unbiased all-atom simulations approximately 9 ms for twelve different proteins with...

Coarse-grained (CG) molecular dynamics enables the study of biological processes at temporal and spatial scales that would be intractable an atomistic resolution. However, accurately learning a CG force field remains challenge. In this work, we leverage connections between score-based generative models, fields, to learn without requiring any inputs during training. Specifically, train diffusion model on protein structures from simulations, show its score function approximates can directly...

Abstract Protein-ligand docking is an established tool in drug discovery and development to narrow down potential therapeutics for experimental testing. However, a high-quality protein structure required often the treated as fully or partially rigid. Here we develop AI system that can predict flexible all-atom of protein-ligand complexes directly from sequence information. We find classical methods are still superior, but depend upon having crystal structures target protein. In addition...

The prediction of protein folding rates and mechanisms is currently great interest in the community. A close comparison between theory experiment this area promising to advance our understanding physical-chemical principles governing process. delicate interplay entropic energetic/enthalpic factors free energy regulates details complex reaction. In article, we propose use topological descriptors quantify amount heterogeneity configurational entropy contribution energy. We apply procedure a...

The overall structure of the transition-state and intermediate ensembles observed experimentally for dihydrofolate reductase interleukin-1β can be obtained by using simplified models that have almost no energetic frustration. predictive power these suggests that, even very large proteins with completely different folding mechanisms functions, real protein sequences are sufficiently well designed, much structural heterogeneity in intermediates is determined topological effects.

Significance Notch signaling pathway plays crucial roles in cell-fate determination during embryonic development and cancer progression. According to the current paradigm, Notch–Delta leads complementary selection between two neighboring cells where one acts as Sender or Receiver. However, this picture is not complete because an additional ligand, Jagged, involved signaling. We devise a specific theoretical framework decipher functional role of Jagged. find that asymmetry modulations Delta...

An adaptive sampling algorithm is proposed to rapidly reconstruct free-energy landscapes of macromolecular systems.

Accurate mechanistic description of structural changes in biomolecules is an increasingly important topic and chemical biology. Markov models have emerged as a powerful way to approximate the molecular kinetics large while keeping full resolution divide-and-conquer fashion. However, accuracy these limited by that force fields used generate underlying dynamics (MD) simulation data. Whereas quality classical MD has improved significantly recent years, remaining errors Boltzmann weights are...