Chimie numérique

Résumé

La chimie numérique ou chimie informatique, parfois aussi chimie computationnelle, est une branche de la chimie et de la physico-chimie qui utilise les lois de la chimie théorique exploitées dans des programmes informatiques spécifiques afin de calculer structures et propriétés d'objets chimiques tels que les molécules, les solides, les agrégats atomiques (ou clusters), les surfaces, etc., en appliquant autant que possible ces programmes à des problèmes chimiques réels. La frontière entre la simulation effectuée et système réel est définie par le niveau de précision requis et/ou la complexité des systèmes étudiés et les théories employées lors de la modélisation. Les propriétés recherchées peuvent être la structure (géométrie, relations entre constituants), l'énergie totale, l'énergie d'interaction, les charges, dipôles et moments multipolaires, fréquences vibrationnelles, réactivité ou autres quantités spectroscopiques, sections efficaces pour les collisions, etc. Le domaine le plus

Source officielle

https://fr.wikipedia.org/wiki/Chimie_num%C3%A9rique

À propos de ce résultat

Cette page est générée automatiquement et peut contenir des informations qui ne sont pas correctes, complètes, à jour ou pertinentes par rapport à votre recherche. Il en va de même pour toutes les autres pages de ce site. Veillez à vérifier les informations auprès des sources officielles de l'EPFL.

Publications associées (100)

Publications associées

Chargement

Personnes associées (84)

Personnes associées

Chargement

Unités associées (48)

Unités associées

Chargement

Concepts associés (87)

Concepts associés

Chargement

Cours associés (58)

Cours associés

Chargement

Séances de cours associées (84)

Séances de cours associées

Chargement

Towards a Composite Benchmark Method for the Aqueous Solvation Free Energies of Organic Compounds

Daniel David Sadowsky

Water is ubiquitous in the biosphere, and the difficulty of rigorously treating the effects of aqueous solvation currently limits the widespread use of computational methods in many areas of environmental chemistry. We propose an approach to benchmark calculations of aqueous free energies of solvation based on molecular dynamics simulation which is derived from first principles, statistical mechanically complete, and systematically improvable. The approach partitions the solvation free energy into the free energies of solvent density change, solvent cavity organization, and solvent-solute interaction, each of which is computed from first principles. A composite approach is considered for the calculation of the contributions of quantum nuclear effects to solvation free energies and the contributions of electron correlation and relativistic effects to solvent-solute interactions.

2015

Chargement

Over the last two decades, data-powered machine learning (ML) tools have profoundly transformed numerous scientific fields. In computational chemistry, machine learning applications have permitted faster predictions of chemical properties and provided powerful analytical tools, facilitating the exploration of the chemical space. The original work presented in this thesis leverages the paradigm-shifting influence of ML and focuses on bridging the divide between unsupervised and supervised learning with the overarching objective of improving the predictive power of similarity-based machine learning algorithms such as kernel regression.Despite their widespread use in chemistry, current implementations of kernel regression suffer from biased definitions of similarity between chemical environments. This problem originates from the rigidity of current numerical approaches for encoding molecular information, based on expert-crafted representations. Moreover, it is amplified by the incorrect (yet generalized) assumption that increasing the amount of information encoded in molecular representations unequivocally improves the evaluation of molecular similarity. As a result, the performance of kernel models can be sub-optimal reducing their broad applicability. To overcome such limitations, we introduce a series of statistical tools and methodologies based on supervised dimensionality reduction and metric learning capable of filtering and adapting the features of common molecular representations. This allows tailoring the notion of "molecular similarity" in order to optimize the prediction of specific chemical targets.Using examples such as the exploration of the free-energy landscape of oligopeptides or the prediction of subtle properties associated with the outcome of chemical reactions (for example, enantiomeric excess), we demonstrate how the methods proposed in this thesis unlock the optimal performance of kernel regression and, more generally, of any similarity-based algorithm. Overall, the work within is part of a larger, more comprehensive effort aimed at extending the capabilities of computational modeling to increasingly complex chemical situations by exploiting the latest advances in statistical learning.

EPFL2022

Chargement

In the last years, it has been demonstrated a link between the overload of metal ions inside nervous system cells and the onset of severe neurodegenerative diseases. This prompted the investigation of the structural and functional properties of transporters responsible for the uptake and regulation of metal ions inside cells. In particular, members of the Nramp/SLC11 ("natural resistance-associated macrophage proteins"/"solute carrier 11") family of transporters tightly regulate the influx of divalent transition-metal ions across cellular membranes contributing to cellular homeostasis. In these secondary active transporters, the translocation of divalent transition metal ions from the extracellular matrix to the cellular cytoplasm is coupled to proton transport (symport). The co-transport is realized using an alternate access mechanism in which the protein is alternately open towards the extracellular matrix, ("outward facing conformation", OFC) or towards the cellular cytoplasm ("inward facing conformation", IFC).Although a wide range of functional and structural studies provided fundamental insights on the proton-coupled transport, most mechanistic details are still unknown. As an example, the coordination sphere of the metal ion in the inward-facing conformation of the protein has only been partially resolved by crystallographic experiments and later attempts to provide comprehensive descriptions led to contradictory models. Furthermore, several experimental evidence proved that proton uniport in the absence of the metal ion, is also possible, suggesting an unusual symporter behaviour for Nramps, in which the two substrates are not tightly coupled. Despite the recent advances in the comprehension of Nramps proton transport, the underlying mechanistic details remain still ambiguous, as demonstrated by inconsistent mechanistic models proposed during the last years.In this thesis, computational methods have been used in order to provide new molecular level insights in the proton-coupled divalent transition metal ions transport performed by Nramps. In particular, a combination of classical molecular dynamics (MD) and QM/MM ab initio MD methods have been applied in order to refine the partially resolved coordination sphere of Mn2+ in the active site of Staphylococcus Capitis Divalent Metal-ion Transporter (ScaDMT), representing the only available crystal structure of a substrate-bound member of the family in an IFC. This QM/MM multiscale approach, using an accurate quantum treatment of the active site and a classical treatment of the surroundings, allowed for state-of-the-art calculations of the entire system that led to the unequivocal identification of the metal's coordination sphere. Furthermore, MD simulations have been performed for both the IFC ScaDMT and Eremococcus Coleocola DMT (EcoDMT), whose crystal structure has been obtained in OFC. Since experimental evidences proved that proton transport can occur even in the absence of the metal ion substrate, we performed simulations of both proteins in their apo forms. These simulations allowed us to characterize four amino acidic residues identified as likely key players in the proton transport process and to determine which role they assume in the process. Finally, we were able to identify the likely primary proton acceptor and proposed a potential route followed by protons in their influx toward the intracellular cytoplasm.

EPFL2022

Chargement

EPFL2022