scholarly journals An Efficient GaMD Multi-Level Enhanced Sampling Strategy for Polarizable Force Fields Simulations of Large Biological Systems

2021 ◽  
Author(s):  
Fréderic Célerse ◽  
Theo Jaffrelot-Inizan ◽  
Louis Lagardère ◽  
Olivier Adjoua ◽  
Pierre Monmarché ◽  
...  
Keyword(s):  
Adaptive Sampling ◽  
Sampling Strategy ◽  
Umbrella Sampling ◽  
Potential Of Mean Force ◽  
Enhanced Sampling ◽  
Polarizable Force Field ◽  
Accelerated Molecular Dynamics ◽  
Polarizable Force Fields ◽  
Solute Interactions ◽  
Multi Level

We introduce a novel multi-level enhanced sampling strategy grounded on Gaussian accelerated Molecular Dynamics (GaMD). First, we propose a GaMD multi-GPUs -accelerated implementation within Tinker-HP. For the specific use with the flexible AMOEBA polarizable force field (PFF), we introduce the new "dual–water" GaMD mode. By adding harmonic boosts to the water stretching and bonding terms, it accelerates the solvent-solute interactions while enabling speedups with fast multiple–timestep integrators. To further reduce time-to-solution, we couple GaMD to Umbrella Sampling (US). The GaMD—US/dual–water approach is tested on the 1D Potential of Mean Force (PMF) of the CD2–CD58 system (168000 atoms) allowing the AMOEBA PMF to converge within 1 kcal/mol of the experimental value. Finally, Adaptive Sampling (AS) is added enabling AS–GaMD capabilities but also the introduction of the new Adaptive Sampling–US–GaMD (ASUS–GaMD) scheme. The highly parallel ASUS–GaMD setup decreases time to convergence by respectively 10 and 20 compared to GaMD–US and US.

scholarly journals An Efficient GaMD Multi-Level Enhanced Sampling Strategy: Application to Polarizable Force Fields Simulations of Large Biological Systems

2021 ◽  
Author(s):  
Fréderic Célerse ◽  
Theo Jaffrelot-Inizan ◽  
Louis Lagardère ◽  
Olivier Adjoua ◽  
Pierre Monmarché ◽  
...  
Keyword(s):  
Molecular Dynamics ◽  
Adaptive Sampling ◽  
Sampling Strategy ◽  
Umbrella Sampling ◽  
Potential Of Mean Force ◽  
Enhanced Sampling ◽  
Polarizable Force Field ◽  
Accelerated Molecular Dynamics ◽  
Polarizable Force Fields ◽  
Multi Level

We detail a novel multi-level enhanced sampling strategy grounded on Gaussian accelerated Molecular Dynamics (GaMD). First, we propose a GaMD multi-GPUs-accelerated implementation within the Tinker-HP molecular dynamics package. We then introduce the new "dual-water" mode and its use with the flexible AMOEBA polarizable force field. By adding harmonic boosts to the water stretching and bonding terms, it accelerates the solvent-solute interactions while enabling speedups thanks to the use of fast multiple--timestep integrators. To further reduce time-to-solution, we couple GaMD to Umbrella Sampling (US). The GaMD—US/dual-water approach is tested on the 1D Potential of Mean Force (PMF) of the CD2-CD58 system (168000 atoms) allowing the AMOEBA PMF to converge within 1 kcal/mol of the experimental value. Finally, Adaptive Sampling (AS) is added enabling AS-GaMD capabilities but also the introduction of the new Adaptive Sampling--US--GaMD (ASUS--GaMD) scheme. The highly parallel ASUS--GaMD setup decreases time to convergence by respectively 10 and 20 compared to GaMD--US and US.

An Efficient GaMD Multi-Level Enhanced Sampling Strategy: Application to Polarizable Force Fields Simulations of Large Biological Systems

2021 ◽  
Author(s):  
Fréderic Célerse ◽  
Theo Jaffrelot-Inizan ◽  
Louis Lagardère ◽  
Olivier Adjoua ◽  
Pierre Monmarché ◽  
...  
Keyword(s):  
Molecular Dynamics ◽  
Adaptive Sampling ◽  
Sampling Strategy ◽  
Umbrella Sampling ◽  
Potential Of Mean Force ◽  
Enhanced Sampling ◽  
Polarizable Force Field ◽  
Accelerated Molecular Dynamics ◽  
Polarizable Force Fields ◽  
Multi Level

We detail a novel multi-level enhanced sampling strategy grounded on Gaussian accelerated Molecular Dynamics (GaMD). First, we propose a GaMD multi-GPUs-accelerated implementation within the Tinker-HP molecular dynamics package. We then introduce the new "dual-water" mode and its use with the flexible AMOEBA polarizable force field. By adding harmonic boosts to the water stretching and bonding terms, it accelerates the solvent-solute interactions while enabling speedups thanks to the use of fast multiple--timestep integrators. To further reduce time-to-solution, we couple GaMD to Umbrella Sampling (US). The GaMD—US/dual-water approach is tested on the 1D Potential of Mean Force (PMF) of the CD2-CD58 system (168000 atoms) allowing the AMOEBA PMF to converge within 1 kcal/mol of the experimental value. Finally, Adaptive Sampling (AS) is added enabling AS-GaMD capabilities but also the introduction of the new Adaptive Sampling--US--GaMD (ASUS--GaMD) scheme. The highly parallel ASUS--GaMD setup decreases time to convergence by respectively 10 and 20 compared to GaMD--US and US.

scholarly journals An Efficient GaMD Multi-Level Enhanced Sampling Strategy: Application to Polarizable Force Fields Simulations of Large Biological Systems

2021 ◽  
Author(s):  
Fréderic Célerse ◽  
Theo Jaffrelot-Inizan ◽  
Louis Lagardère ◽  
Olivier Adjoua ◽  
Pierre Monmarché ◽  
...  
Keyword(s):  
Molecular Dynamics ◽  
Adaptive Sampling ◽  
Force Fields ◽  
Biological Systems ◽  
Sampling Strategy ◽  
Umbrella Sampling ◽  
Potential Of Mean Force ◽  
Enhanced Sampling ◽  
Polarizable Force Fields ◽  
Multi Level

We introduce a novel multi-level enhanced sampling strategy grounded on Gaussian accelerated Molecular Dynamics (GaMD). First, we propose a GaMD multi-GPUs-accelerated implementation within the Tinker-HP molecular dynamics package. We introduce the new "dual-water" mode and its use with the flexible AMOEBA polarizable force field.By adding harmonic boosts to the water stretching and bonding terms, it accelerates the solvent-solute interactions while enabling speedups thanks to the use of fast multiple--timestep integrators. To further reduce time-to-solution, we couple GaMD to Umbrella Sampling (US). The GaMD—US/dual--water approach is tested on the 1D Potential of Mean Force (PMF) of the solvated CD2--CD58 system (168000 atoms) allowing the AMOEBA PMF to converge within 1 kcal/mol of the experimental value. Finally, Adaptive Sampling (AS) is added enabling AS-GaMD capabilities but also the introduction of the new Adaptive Sampling--US--GaMD (ASUS-GaMD) scheme. The highly parallel ASUS--GaMD setup decreases time to convergence by respectively 10 and 20 times compared to GaMD-US and US. Overall, beside the acceleration of PMF computations, Tinker-HP now allows for the simultaneous use of Adaptive Sampling and GaMD-"dual water" enhanced sampling approaches increasing the applicability of polarizable force fields to large scale simulations of biological systems.

scholarly journals High-resolution mining of the SARS-CoV-2 main protease conformational space: supercomputer-driven unsupervised adaptive sampling

2021 ◽  
Author(s):  
Théo Jaffrelot Inizan ◽  
Frédéric Célerse ◽  
Olivier Adjoua ◽  
Dina El Ahdab ◽  
Luc-Henri Jolly ◽  
...  
Keyword(s):  
Molecular Dynamics ◽  
High Resolution ◽  
Adaptive Sampling ◽  
Force Fields ◽  
Sampling Strategy ◽  
Md Simulations ◽  
Conformational Space ◽  
Many Body ◽  
Polarizable Force Fields ◽  
Main Protease

We provide an unsupervised adaptive sampling strategy capable of producing μs-timescale molecular dynamics (MD) simulations of large biosystems using many-body polarizable force fields (PFFs).

Capturing the Flexibility of a Protein-Ligand Complex with Enhanced Sampling Techniques: Binding Free Energies from Umbrella Sampling, REST2, and Bias Exchange

2019 ◽  
Author(s):  
Sebastian Wingbermühle ◽  
Lars V. Schäfer
Keyword(s):  
Degrees Of Freedom ◽  
Large Scale ◽  
Statistical Error ◽  
Umbrella Sampling ◽  
Potential Of Mean Force ◽  
Sampling Techniques ◽  
Enhanced Sampling ◽  
Ligand Complex ◽  
Mhc I ◽  
Binding Groove

The performance of the three popular enhanced sampling techniques Umbrella Sampling (US), Replica Exchange with Solute Tempering 2 (REST2), and Bias Exchange (BE) is tested on Major Histocompatibility Complex I (MHC I) binding an antigenic peptide. The configurational flexibility of peptide-MHC I complexes (pMHC I) is key to their immunological function and must therefore be captured thoroughly by the sampling techniques used to yield accurate thermodynamics of pMHC I. Here, we calculate the Potential of Mean Force (PMF) for the dissociation of the peptide N-terminus from the MHC I binding groove. We carefully analyze the statistical error of the resulting PMF and the sampling of the orthogonal degrees of freedom to assess how well the three sampling techniques used can handle large-scale configurational flexibility.<br>

scholarly journals Exploration of the selective binding mechanism of protein kinase Aurora A selectivity via a comprehensive molecular modeling study

PeerJ ◽  
2019 ◽  
Vol 7 ◽  
pp. e7832 ◽  
Author(s):  
Zhe Zhang ◽  
Yafei Xu ◽  
Jian Wu ◽  
Ying Shen ◽  
Hao Cheng ◽  
...  
Keyword(s):  
Protein Dynamics ◽  
Umbrella Sampling ◽  
Potential Of Mean Force ◽  
Aurora A ◽  
Enhanced Sampling ◽  
Selective Inhibitors ◽  
Altered Expression ◽  
Promising Therapeutic Target ◽  
Calculation Results ◽  
Underlying Mechanisms

Background The kinase of Aurora A has been regarded as a promising therapeutic target due to its altered expression in various human cancers. However, given the high similarity of the active binding site of Aurora A to other kinases, designing highly selective inhibitors towards Aurora A remains a challenge. Recently, two potential small-molecule inhibitors named AT9283 and Danusertib were reported to exhibit significant selectivity to Aurora A, but not to Gleevec. It was argued that protein dynamics is crucial for drug selectivity to Aurora A. However, little computational research has been conducted to shed light on the underlying mechanisms. Methods In this study, MM/GBSA calculations based on conventional molecular dynamics (cMD) simulations and enhanced sampling simulations including Gaussian accelerated MD (GaMD) simulations and umbrella sampling were carried out to illustrate the selectivity of inhibitors to Aurora A. Results The calculation results from cMD simulation showed that the binding specificity is primarily controlled by conformational change of the kinase hinge. The protein dynamics and energetic differences were further supported by the GaMD simulations. Umbrella sampling further proved that AT9283 and Danusertib have similar potential of mean force (PMF) profiles toward Aurora A in terms of PMF depth. Compared with AT9283 and Danusertib, Gleevec has much lower PMF depth, indicating that Gleevec is more easily dissociated from Aurora A than AT9283 and Danusertib. These results not only show the selective determinants of Aurora A, but also provide valuable clues for the further development of novel potent Aurora A selective inhibitors.

A multi-level splitting algorithm based on differential evolution

2018 ◽  
Vol 09 (02) ◽  
pp. 1850021
Author(s):  
Liping Wang ◽  
Wenhui Fan
Keyword(s):  
Differential Evolution ◽  
Rare Event ◽  
Sampling Strategy ◽  
Splitting Algorithm ◽  
Algorithm Efficiency ◽  
Event Simulation ◽  
Sample Paths ◽  
Level Splitting ◽  
Key Factor ◽  
Multi Level

Multi-level splitting algorithm is a famous rare event simulation (RES) method which reaches rare set through splitting samples during simulation. Since choosing sample paths is a key factor of the method, this paper embeds differential evolution in multi-level splitting mechanism to improve the sampling strategy and precision, so as to improve the algorithm efficiency. Examples of rare event probability estimation demonstrate that the new proposed algorithm performs well in convergence rate and precision for a set of benchmark cases.

Sign in / Sign up

Export Citation Format

Share Document