You are here : Home > The Institute > News > High Performance Computing and molecular modeling

News | Article | Computing | Simulation & modelling | Supercomputers | Structural biology

The european PPI4HPC project echoes high-performance computing at CEA for the modelisation of complex molecular assemblages


​A project at the interface between quantum chemistry and molecular dynamics for the study of complex molecular assemblies, led by Michel Masella (Joliot) and Luigi Genovese (IRIG), makes it possible, for applications in biology/health, to take full advantage of the new supercomputers of the Public Procurement of Innovations for High-Performance Computing (PPI4HPC, H2020) project hosted at the Très Grand Centre de Calcul (TGCC) of the CEA-Bruyères-le-Châtel. Highlight from PPI4HPC.

Published on 14 September 2021

​The aim of the project "Advanced protein/protein or protein/ligand binding energy estimator", recently highlighted on the PPI4HPC website and led by Michel Masella (Nuclear Envelope, Telomeres and DNA Repair team, I2BC) and Luigi Genovese (Modelling and Exploration of Materials Laboratory, IRIG) is to make extensive use of a new numerical simulation tool that will allow a better understanding of the microscopic interactions that are responsible for the stability of large molecular assemblies such as antibody/antigen complexes. In this project, which received a Sanofi iTech Award in November 2020, two powerful simulation approaches are coupled : a molecular dynamics approach to exhaustively sample the potential energy surface associated with a molecular complex (POLARIS(MD) code) and a quantum chemistry code (BIGDFT) allowing, thanks to its mathematical formalism based on wavelets, to simulate molecular systems at the scale of several thousand atoms.

Such a hybrid molecular dynamics/quantum chemistry approach has been successfully applied to macromolecular systems in biology, notably in the context of the Covid-19 pandemic, allowing to identify the active fragments of a SARS-CoV-2 protein interacting with an inhibitory molecule and to provide the key parameters for their modeling. This work should be the subject of a forthcoming publication (Learn more).

Contacts: Michel Masella (michel.masella@cea.fr) – Luigi Genovese (luigi.genovese@cea.fr)

- This news (link) was published on 31/08/2021 on the PPI4HPC (Public Procurement of Innovations for High Performance Computing) website, co-financed by Horizon 2020.

- The TGCC (Très Grand Centre de Calcul at CEA-Bruyères-le-Châtel) is an infrastructure for very high performance scientific computing and Bigdata, capable of hosting exaflop-scale supercomputers. In the context of the PPI4HPC project, the researchers benefited from the allocation of 15 million core hours on the AMD partition of the Joliot-Curie supercomputer at the TGCC. This partition, an extension acquired by GENCI, has enabled record performance to be achieved, necessary for operations as demanding as the modeling of molecular assemblies.

- GENCI : In charge of providing high-performance computing and massive data processing resources, GENCI's mission, at the national and European levels, is to promote the use of high performance computing associated with Artificial Intelligence for the benefit of academic and industrial research communities.


Top page