Thomas Young Centre Soirée: 'Scalable polarizable molecular dynamics using Tinker-HP', Prof Jean-Philip Piquemal & 'Can we derive many-body non-additive polarization energies from 1-body properties and 2-body energies only?, Dr Alston Misquitta
Date: 17 October 2019 Time: 17:00 - 19:00
Our next Thomas Young Centre Soirée begins with a first talk at 17:00 covering research advancements in force-fields, intermolecular electronic structure methods, and simulation codes. Join us as we welcome Professor of Theoretical Chemistry at the Sorbonne Université - Jean-Philip Piquemal to our next soirée.
Scalable polarizable molecular dynamics using Tinker-HP (Jean-Philip Piquemal, Sorbonne Université, Paris, France)
Tinker-HP is a CPU based, double precision, massively parallel package dedicated to long
polarizable molecular dynamics simulations and to polarizable QM/MM. Tinker-HP is an
evolution of the popular Tinker package (http://tinker-hp.ip2ct.upmc.fr/) that conserves it
simplicity of use but brings new capabilities allowing performing very long molecular dynamics simulations on modern supercomputers that use thousands of cores. Indeed, this project gathers scientists from various fields including Chemistry, Applied Mathematics and Biomolecular Engineering and redefined completely the algorithmic of usual polarizable simulations package.The Tinker-HP approach offers various strategies using domain decomposition techniques forperiodic boundary conditions in the framework of the (N)log(N) Smooth Particle Mesh Ewald or using polarizable continuum simulations through the new generation ddCosmo approach. Tinker-HP proposes a high performance scalable computing environment for polarizable force fields giving access to large systems up to millions of atoms. I will present the performances and scalability of the software in the context of the AMOEBA force field and show the incoming new
features including the advanced SIBFA polarizable molecular mechanics approach and the density based GEM force field as well as newly available "fully polarizable" QM/MM capabilities. Various benchmarks and examples on biomolecular systems will be provided on several architectures showing that the approach is competitive with GPUs for small and medium size systems but allows addressing larger molecules on modern supercomputers. As the present implementation is clearly devoted to petascale applications, the applicability of such an approach to future exascale machines will be exposed and future directions of Tinker-HP discussed including the new GPUs-based implementation.
1) Tinker-HP: a Massively Parallel Molecular Dynamics Package for Multiscale Simulations of Large Complex Systems with Advanced Polarizable Force Fields. L. Lagardère, L.-H. Jolly, F. Lipparini, F. Aviat, B. Stamm, Z. F. Jing, M. Harger, H. Torabifard, G. A. Cisneros, M. J. Schnieders, N. Gresh, Y. Maday, P. Ren, J. W. Ponder, J.-P. Piquemal, Chem. Sci., 2018, 9, 956-972 (Open Access)
2) Towards Large Scale Hybrid QM/MM Dynamics of Complex Systems with Advanced Point Dipole Polarizable Embeddings. D. Loco, L. Lagardère, G. A. Cisneros, G. Scalmani, M. Frisch, F. Lipparini, B. Mennucci, J.-P. Piquemal, Chem. Sci., 2019, 10, 7200-7211 (Open Access)
3) Raising the Performance of the Tinker-HP Molecular Modeling Package [Article v1.0]. L. H. Jolly, A. Duran, L. Lagardère, J. W. Ponder, P. Y. Ren, J.-P. Piquemal, LiveCoMS, 2019, en ligne (Open Access) DOI: 10.33011/livecoms.1.2.10409
Jean-Philip Piquemal bio:
Jean-Philip Piquemal has been trained as a Quantum Chemist at Université Pierre et Marie Curie (UPMC, Paris). After his PhD (2004) and a 2-year postdoctorate as an NIH fellow at the National Institute of Environmental Health Sciences (NIEHS, USA), he became Assistant Professor at UPMC in 2006. He defended his Research Habilitation (HDR) in 2009 to become Full Professor in 2011. In 2016 he was nominated as Junior Member of the Institut Universitaire de France (Research Chair 2016-2021). He is currently Distinguished Professor (Exceptional Class, PRCE1) in Theoretical Chemistry at Sorbonne Université (SU) and Director of the Laboratoire de Chimie Théorique (LCT, UMR 7616 SU/CNRS). He is PI of the ERC EMC2 initiative (2019-2025).
His research is devoted to theoretical chemistry and includes methodological/software developments in multiscale quantum chemistry for large systems, new generation polarizable force fields and quantum chemical topology. This work is performed in strong interdisciplinary interactions with Applied Mathematics and High-Performance Computing.
Our second talk is 'Can we derive many-body non-additive polarization energies from
1-body properties and 2-body energies only?' with Dr Alston J. Misquitta
Force-fields are normally fitted to a range of data, which could be experimental or theoretical, or a mixture of both. As a rule, many-body intermolecular interaction models
are fitted to a range of energies computed on dimers, trimers and often even larger clusters of the interacting molecules. This is done so as to get a good representation of the many-body interactions in the system, but while it is possible to make such extensive computations on small systems like water, it is not generally feasible on larger systems. So
we raise the question: Can we derive the many-body polarization energies from molecular
properties and a dimer energy evaluations only?
Using very recent work on water models  we demonstrate that this is indeed possible.
Central to this scheme are the distributed molecular multipoles computed using the basis-space implementation of the iterated stockholder atoms (BS-ISA) algorithm , the distributed molecular polarizabilities computed using the ISA-Pol algorithm , and charge-delocalization energies defined using regularized-SAPT(DFT) . With these ISA-based molecular properties and a limited number of SAPT(DFT) and regularized-SAPT(DFT) interaction energies computed on dimers only, we can now construct accurate many-body interaction models that correctly describe many-body polarization effects. Using a series of water models of varying complexity we demonstrate how these DIFF (derived inter-molecular force fields) models are very close to optimal and have a remarkable predictive power for both the energerics and geometries of large water clusters. All of the techniques described in this talk are available in the CamCASP code  and more information can be found on the CamCASP wiki.
 “Many-body non-additive polarization energies for water clusters from first principles
polarization models”, Rory A. J. Gilmore, Martin T. Dove, and Alston J. Misquitta (in
 BS-ISA: “Distributed Multipoles from a Robust Basis-Space Implementation of the
Iterated Stockholder Atoms Procedure”, A. J. Misquitta and A. J. Stone and F. Fazeli,
J. Chem. Theory Comput. 10, 5405 (2014).
 ISA-Pol: “ISA-Pol: Distributed Polarizabilities and Dispersion Models from a Basis-
Space Implementation of the Iterated Stockholder Atoms Procedure”, Theor Chem Acc
(2018) 137: 153. https://doi.org/10.1007/s00214-018-2371-4.
 Reg-SAPT(DFT): “Charge-transfer from Regularized Symmetry-Adapted Perturba-
tion Theory”, Alston J Misquitta, J. Chem. Theory Comput. 9, 5313 (2013).
 CamCASP: Cambridge package for Calculation of Anisotropic Site Properties, Alston
J Misquitta and Anthony J Stone, https://app.ph.qmul.ac.uk/wiki/ajm:camcasp:start
Alston J. Misquitta bio:
My primary expertise is in the field of intermolecular interactions. Here I have made major advances in the fundamental electronic structure methods that are used. One of these is the symmetry-adapted perturbation theory based on density-functional theory, or SAPT(DFT), which is arguably one of the most competitive electronic structure methods for computing the interactions between molecular systems of moderate size. I have also developed advanced methods for computing molecular properties in distributed form: these include the ISA-DMA multipole moments, ISA-Pol frequency-dependent polarizabilities and dispersion models, and WSM polarizabiliity and dispersion models. All of these methods, as well as key methods to combine them to generate intermolecular force-fields, are implemented in the CamCASP program which has been written by me and my long-term collaborator, Prof. Anthony Stone (Cambridge).
The talks will be followed by supper.
|Location:||G.O. Jones Building - Room 6.10, Mile End Campus, Queen Mary University of London|