Description du projet
Franchir la barrière de la vitesse pour étudier les photoprocessus subnanosecondaires dans l’ADN
L’objectif du projet SubNano, financé par l’UE, est d’accroître considérablement la vitesse de simulation de la dynamique des molécules photoexcitées pour permettre l’étude d’évènements subnanosecondaires (sub‑ns). La méthodologie sub‑ns sera utilisée pour étudier la dynamique non adiabatique se déroulant sur une longue échelle de temps des processus photoinduits dans les acides nucléiques, y compris la photostabilisation de l’ADN par des processus excitoniques, les marqueurs fluorescents et la formation d’anions transitoires dans la réparation de l’ADN. Les chercheurs étendront les simulations de la dynamique non adiabatique à la nouvelle échelle de temps en utilisant des algorithmes adaptatifs d’apprentissage automatique diabatique ainsi qu’une méthode mixte quantique‑classique corrigée du point zéro et corrigée vibratoirement. L’extension des études théoriques des processus photodynamiques dans le régime sub‑ns deviendra enfin une réalité.
Objectif
My goal in the SubNano project is to massively speed up the dynamics simulation of photoexcited molecules to allow addressing sub-nanosecond phenomena (that is, one thousand times above the current limits).
The sub-ns methodology will be employed to investigate the long timescale nonadiabatic dynamics of photoinduced processes in nucleic acids, including DNA photostabilization via excitonic processes, biological fluorescent markers, and transient anion formation in DNA repair.
To fulfill these goals, I will develop and implement a series of methods to extend nonadiabatic dynamics simulations into the new timescale, mainly based on a novel adaptive diabatic machine learning algorithm and a novel zero-point-corrected and vibronically-corrected mixed quantum-classical method.
The sub-ns methodology will be constrained to be general (any kind or size of molecule), black-box (minimum user intervention), modular (adaptable to any electronic structure theory), on-the-fly (no need of precomputed potential energy surfaces), and local (independent-trajectories).
It will be implemented into the Newton-X software platform, which I have been the main designer and developer. It will also be made available for all academic community through new releases of Newton-X.
For the last 25 years, theoretical investigations of photodynamical processes have been restricted to the ultrafast (picosecond) regime, selectively choosing problems in this domain. The extension into the sub-ns regime is finally feasible thanks to a large algorithmic infrastructure I have built over the last 13 years, paving the grounds to develop a new research area, atomistic nonadiabatic dynamics on the long timescale.
The success of the SubNano project will have an enormous impact on the research field, allowing to investigate outstanding interdisciplinary phenomena in chemistry, biology, and technology, which have been neglected due to a lack of methods.
Champ scientifique
- natural sciencescomputer and information sciencessoftware
- natural sciencesbiological sciencesbiochemistrybiomoleculesnucleic acids
- natural sciencesbiological sciencesgeneticsDNA
- natural scienceschemical sciencesphysical chemistryphotochemistry
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
Programme(s)
Thème(s)
Régime de financement
ERC-ADG - Advanced GrantInstitution d’accueil
13284 Marseille
France