Bulletin of the American Physical Society
APS March Meeting 2019
Volume 64, Number 2
Monday–Friday, March 4–8, 2019; Boston, Massachusetts
Session B21: Exploring Free Energy Landscapes in Biology and Materials Science IFocus
|
Hide Abstracts |
Sponsoring Units: DCOMP DBIO DMP GSOFT Chair: Pratyush Tiwary, University of Maryland, College Park Room: BCEC 157B |
Monday, March 4, 2019 11:15AM - 11:51AM |
B21.00001: Incorporating experimental data into long timescale simulations of macromolecules Invited Speaker: Cecilia Clementi Recent breakthroughs in experimental technologies and in high-performance computing have enabled unprecedented measurements and simulations of complex biophysical systems such as macromolecules. However, experiments provide only a partial view of macromolecular processes and are limited in their temporal and spatial resolution. On the other hand, atomistic simulations are still not able to sample the conformation space of large complexes, thus leaving significant gaps in our ability to study molecular processes at a biophysically relevant scale. We present our efforts to bridge these gaps, by using experimental data as a starting point in a computational modeling approach. We use models at different resolutions and "anchor" them to experimental measurements, to provide quantitatively accurate representations of systems of interest, and address open biophysical questions. |
Monday, March 4, 2019 11:51AM - 12:03PM |
B21.00002: Complex landscape of the epithelial-mesenchymal transition Francesc Font Clos, Stefano Zapperi, Caterina La Porta Cells can change their phenotype from epithelial to mesenchymal during development and in cancer progression, where this transition is often associated with metastasis and poor disease prognosis. We simulate a Boolean network model of the pathway regulating the epithelial–mesenchymal transition, reconstruct its complex phenotypic landscape and validate it by analyzing single cell transcriptomic data. The statistical features of the landscape are similar to those found in the free energy of glassy materials and suggest that highly aggressive hybrid epithelial/mesenchymal cell phenotypes are located in metastable regions that can easily switch under external and internal perturbations. We thus propose a general mapping strategy for gene regulatory networks that can be used to visualize the ever increasing number of gene expression data obtained from single cells and tissues. |
Monday, March 4, 2019 12:03PM - 12:15PM |
B21.00003: Peptide Aggregation in Solution and Interaction with Inorganic Surfaces: A Metadynamics Investigation Janani Sampath, Jim Pfaendtner Silaffin peptides that regulate the process of biomineralization in organisms like diatoms are interesting as their ability to generate intricate mineral morphologies at ambient conditions could be harnessed to produce precise materials from non-biological molecules. A phenomenon central to this process is peptide aggregation – prior studies show that different aggregates are associated with different mineral morphologies. We employ all-atom molecular dynamics (MD) simulations to study the aggregation of R5 peptide, a 19 amino acid residue of silaffin that precipitates silica in vitro. In order to access relevant time scales, we enhance our sampling by using a recent variation of the metadynamics approach, metadynamic metainference, which guides average simulation behavior towards experimental structures obtained from NMR. Another relevant aspect of biomineralization is the interaction of peptides with mineral surfaces and the effect that amino acid modification has on peptide-surface binding. By implementing MD simulations enhanced by parallel-bias metadynamics, we generate complex free energy landscapes, which helps elucidate peptide-mineral structure-property relationships. |
Monday, March 4, 2019 12:15PM - 12:27PM |
B21.00004: All-Atom Structure-Based Model of RNA with Explicit Electrostatics and Explicit Treatment of Mobile Ions Ailun Wang, Mariana Levi, Udayan Mohanty, Paul Whitford Structure-based models (SBM) allow milliseconds effective timescale simulations of large biomolecular assemblies. In an SBM, the prevalent potential energy minima are defined to correspond to native structures. Here, we have developed an all-atom SBM in which non-hydrogen atoms and diffuse ions (K+, Cl-, Mg2+) electrostatics are described explicitly. The effective desolvation potentials for ion-ion and ion-RNA interactions are introduced such that solvation is treated implicitly. Excluded volume term and long range Coulomb interaction are also included in the forcefield. To calibrate the model, we simulate RNA Helix 44 and compare various observable quantities with explicit-solvent simulations. Using an iterative refinement protocol, the parameters are optimized to reproduce the potential of mean force obtained from two-microsecond explicit-solvent simulations, until the parameters converge iteratively. This transferable forcefield for ion-ion and ion-RNA interactions are then used in SBM of large assemblies. In an initial application of this model, we show how diffuse ions and the excess ion atmosphere can modulate large-scale conformation dynamics in the ribosome. This opens the door to investigate the role of ions in a wide range of RNA-protein biomolecular assemblies. |
Monday, March 4, 2019 12:27PM - 12:39PM |
B21.00005: Helix Mediated Aggregation in Poly-Glutamine Tracts Mehmet Sayar In protein aggregation related neurodegenerative diseases, a disease specific host protein misfolds and adopts a metastable, aggregation prone conformation. For Huntington's disease, the htt exon I domain is the smallest physiological N-terminal fragment that forms disease related poly-glutamine rich fibrils. Exon I consists of a glutamine tract, flanked by a 17 amino acid long fragment (httNT) and a proline rich fragment. The presence of httNT fragment has been shown to significantly accelerate the formation of poly-glutamine rich fibrils. In this study, we test the helix mediated aggregation mechanism of glutamine tracts as proposed by the Wetzel group by using molecular dynamics simulations. Analysis of httNT fragment in bulk water suggests only a weak helix propensity. However, when several fragments come into contact they adopt the alpha-helix conformation and form bundles. When the httNT fragment is flanked to the poly-glutamine tract, these bundles lead to a dramatic increase in glutamine concentration. By using a coarse grained model, we study the beta-sheet formation and show that, In agreement with the Wetzel group's hypothesis, beta-sheet propensity significantly increases in these bundles compared to the isolated poly-glutamine tracts. |
Monday, March 4, 2019 12:39PM - 12:51PM |
B21.00006: Boltzmann Generators: Deep Learning of Thermodynamics and Efficient Monte Carlo Frank Noe We introduce Boltzmann Generators, a machine learning method to efficiently sample Boltzmann distributions of complex multibody physics systems that give rise naturally to thermodynamic properties of the system, such as free energies, entropies and temperature dependence. |
Monday, March 4, 2019 12:51PM - 1:03PM |
B21.00007: Nonlinear and Hierarchical Discovery of Slow Molecular Modes using Sequential Learning with Natural Constraints Wei Chen, Hythem Sidky, Andrew L Ferguson Discovering the slow modes governing molecular dynamics can unveil new mechanistic understanding and provide collective variables along which to direct enhanced sampling. Time-lagged independent component analysis (tICA) is a well-developed method that discovers linear combination of molecular features as slow modes. The linearity of tICA, however, hampers its capacity to discover nonlinear modes. Nonlinear feature engineering can prove profitable but is typically reliant on human intuition or expensive preprocessing. Kernel tICA integrates tICA and kernel trick to effect nonlinear discovery, but is computationally expensive and selection and tuning of the kernel limits its generalizability. Time-lagged autoencoders and variational dynamics encoders are neural networks that can identify the slowest mode but are unable to resolve higher-order modes. In this work, we introduce a sequential learning method that we term hierarchical dynamics encoder (HDE) as a novel neural network that sequentially learns hierarchical nonlinear slow CVs. Each CV is orthogonal to all previously learned CVs, and orthogonality is imposed naturally without regularization. We demonstrate HDEs for several toy systems where the true slow modes are known, and in simulations of peptides and proteins. |
Monday, March 4, 2019 1:03PM - 1:15PM |
B21.00008: Learning reaction coordinate on-the-fly for sampling complex biomolecular systems with SGOOP and metadynamics Debabrata Pramanik, Adam Kells, Zachary Smith, Pratyush Tiwary Understanding functioning and stabilizing/destabilizing forces of biomolecules such as protein-ligand and protein-DNA are highly desirable due to implication in basic biology as well as diseases. Experiments can be useful in measuring various thermodynamic quantities, but they cannot, at least directly, provide microscopic details and kinetics, pathways etc. Here, we complement experiments with all-atom molecular dynamics (MD) simulations. Unfortunately MD is limited by a huge time scale problem. We attempt to solve it through developing sampling methods based on statistical mechanics and demonstrate progress on model systems and in ambitious systems such as protein-ligand interactions and transcription factor-DNA interactions. One of the specific issues we will address is the need to know beforehand an accurate reaction coordinate (RC), which is a challenge for any biased simulation. Recently, we have shown how to construct a 1-dimensional RC by a method called “spectral gap optimization of order parameters (SGOOP)”. Here we will show how to extend its scope by introducing a simple but powerful extension based on the notion of conditional probability factorization for systems with inherent complexity and where a 1-d RC is not enough to accurately capture the energy landscape. |
Monday, March 4, 2019 1:15PM - 1:27PM |
B21.00009: Understanding Binding Dynamics in Host-Guest Systems with Advanced Sampling Simulations Anne Leonhard, Jonathan Whitmer Host-guest complexes are useful for applications in biomaterials science and engineering, including drug delivery, separations, transport regulation, and novel hydrogels. Highly-specific binding interactions are a key feature of such systems, and as such knowledge of systems' binding affinity, free energy landscapes, and binding pathways is critical. Here, we propose that advanced sampling techniques, such as the Adaptive Biasing Force and Artificial Neural Network methods, are promising ways to explore free energy landscapes of host-guest systems due to their ability to quickly sample the full phase space. We apply these methods to calculation of binding affinities between an array of small molecules and curcurbit[n]uril hosts. These binding affinities calculated with advanced sampling methods are compared to previously-published experimentally- and computationally-obtained values and obtain exceptional agreement with each. We also discuss how such methods might be generalized to more complex systems of interest in the biomolecular and materials simulations communities. |
Monday, March 4, 2019 1:27PM - 1:39PM |
B21.00010: ζ-Glycine: Insight into the Mechanism of a Polymorphic Phase Transition Craig L. Bull, Giles Flowitt-Hill, Stefano de Gironcoli, Emine Kucukbenli, Simon Parsons, Cong-Huy Pham, Helen Playford, Matthew G. Tucker Glycine is the simplest amino acid and a great test-bed for studying polymorphism in molecular crystals. With five phases having been structurally characterized at atmospheric or high pressure, it is a well-studied system both experimentally and theoretically. Yet a sixth form, the ζ phase, was discovered over a decade ago as a short-lived intermediate upon decompression of a high-pressure phase. However, its structure has remained unsolved despite several theoretical investigations to glycine polymorphism. We have recently reported [1] the structure of the elusive ζ phase, which was resolved thanks to the collaboration between crystal structure prediction procedure based on fully ab initio total energy calculations combined with a genetic algorithm, and neutron powder diffraction data obtained on a sample trapped at 100 K. Here we will discuss the approach that led to the successful theoretical prediction of this phase, the lessons that can be learnt and applied to similar systems, and mention remaining challenges for a predictive ab initio crystal structure prediction process. |
Monday, March 4, 2019 1:39PM - 1:51PM |
B21.00011: A Climbing Multi-String Method to Map Free-Energy Saddles and Minima Gourav Shrivastav, Cameron Abrams Finding stationary points, minima or saddles, on hyperdimensional free energy surface is crucial to gain insights about the possible transition events and to compute the associated transition rates. Though minima can be easily found, the tasks of locating saddles and measuring their energies relative to their associated minima remain challenging, especially in high-dimensional spaces. We propose here a modified climbing string method to locate multiple saddles by initiating multiple strings in one go. For a given minimum, strings propagate by gradient flow in the path space, with one end fixed and other end climbing across the free energy profile to locate a saddle. The convergence of multiple strings to a common saddle is avoided by checking repulsive forces between the climbing ends. The presence of multiple images along the string helps to ensure the direct connectivity between the minimum and saddles and to compute the free energy along the path. Hence, the climbing multi-string method can be used to map the network of directly connected stationary points in the free energy hypersurface. We demonstrate this method to locate saddle points in two-dimensional and four-dimensional free energy surface of alanine dipeptide and alanine tripeptide, respectively. |
Monday, March 4, 2019 1:51PM - 2:03PM |
B21.00012: Conformational free energy surface of cyclooctane from metadynamics in the collective variable space of autoencoder neural network Bumjoon Seo, Seulwoo Kim, Minhwan Lee, Youn-Woo Lee, Won Bo Lee For rare event problems in which the important free energy basins are separated by large barriers, enhanced sampling methods provide the means to perform simulations within tractable timescales. Metadynamics simulation, which is one of the widely used enhanced sampling methods, requires the definition of a set of collective variables for accumulating the bias potentials. An important aspect of the collective variables is their dimensionality because the efficiency of the method decreases exponentially with the dimensionality. We present here a methodology of incorporating the codes from an autoencoder neural network as the collective variables for metadynamics simulations. This dimensionality reduction of an eight-dimensional space of dihedral angles into a three-dimensional space of features enables the computation of the conformational free energy surface of cyclooctane. |
Monday, March 4, 2019 2:03PM - 2:15PM |
B21.00013: Disordered peptide chains in a coarse-grained model Marek Cieplak, Lukasz Mioduszewski We construct a one-bead-per-residue coarse-grained dynamical model |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700