Jean-Philip Piquemal @piquemal@mastodon-pa.sorbonne-universite.fr

#compchem #MachineLearning #HPC Out in Chemical Science. "Scalable Hybrid Deep Neural Networks/Polarizable Potentials Biomolecular Simulations including long-range effects", introducing the Deep-HP #NeuralNetworks multi-#GPU-accelerated platform (part of Tinker-HP). Check the paper out (#openaccess) !!! https://doi.org/10.1039/D2SC04815A

#compchem Simulating more than 10 μs, we compute charged/uncharged ligands solvation free energies in 4 solvents, & absolute binding free energies of host–guest complexes from SAMPL challenges. ANI-2X/AMOEBA average errors are within chemical accuracy at force-field cost.
---
RT @jppiquem
#compchem #MachineLearning #HPC Out @ChemicalScience. "Scalable Hybrid Deep Neural Networks/Polarizable Potentials Biomolecular Simulations including lo…
https://twitter.com/jppiquem/status/1643163320503918592

#compchem🚨#preprint🚨:Force-Field-Enhanced Neural Network Interactions: from Local Equivariant Embedding to Atom-in-Molecule properties & long-range effects. Great work by @Thomas__Ple@twitter.com introducing the hybrid physically-driven #machinelearning FENNIX model. Funded by the ERC (project EMC2). Supercomputer time by GENCI. https://arxiv.org/abs/2301.08734

I did not introduce myself after setling up in #Mastodon. I am Professor of Chemistry at Sorbonne Université in Paris. Technically, I am a quantum chemist but I learned a while back ago that molecules do move over time...(incredible!). I am therefore interested in Multiscale Quantum Chemistry and Molecular Simulation. My group's research is performed in strong interdisciplinary interactions with Applied #Maths & #HPC.
#compchem #supercomputing #quantumcomputing #machinelearning