Durant cette thèse, nous avons abordé l'étude de la réactivité en phase gazeuse des biomolécules. L’avènement des techniques d’ionisation douces telle que l’ionisation par éléctronébulisation, a rendu possible ces dernières années, la formation d'ions en phase gazeuse sans dégrader la biomolécule étudiée.La Dissociation Induite par Collision (CID) est un cas particulier de spectrométrie de masse en tandem, que nous avons utilisée durant ce travail. Le principe du CID est d'activer les modes rovibrationnelles d’un système moléculaire ionique par collision avec un gaz inerte, ce qui augmente la probabilité de fragmentation de l'ion. Bien qu'étant une technique très utile d'un point de vue analytique, la spectrométrie de masse en tandem ne donne pas d'informations sur les mécanismes des réactions se produisant dans la cellule de collision; afin d’obtenir ces informations, les simulations de dynamique chimiques apparaissent comme un outil satisfaisant. En effet, en utilisant la dynamique directe, nous évitons ainsi d'explorer la totalité de la surface d'énergie potentielle, qui devient compliquée lors de l’étude d’édifices moléculaires de grande taille. Etant donné que les simulations de dynamique chimiques sont limitées à de courtes échelles, de l’ordre de la dizaine de picosecondes, nous avons également employé la théorie unimoléculaire RRKM (Rice-Ramsperger-Kassel-Marcus) pour étudier la réactivité à des temps plus longs, en vue de comprendre les processus réactionnels se produisant à l’issue du processus de relaxation vibrationnelle intramoléculaire (IVR). Durant ce travail de thèse, nous avons choisi d'étudier comme système modèle de base nucléique la molécule d'uracile. Par ailleurs,nous avons aussi étudié la réactivité en phase gazeuse de sucres (cellobiose, maltose et gentiobiose), qui ont été au préalable dérivatisés afin de localiser la charge sur la molécule et ainsi simplifier l’étude théorique associée. / In the present thesis, we address the study of the reactivity of biomolecules in the gasphase.The advent of soft ionization techniques such as electrospray ionization, made possible, in the last years, the gentle formation of ions in the gas phase without breaking the molecule understudy.Collision Induced Dissociation (CID) is aparticular case of tandem mass spectrometrydynamics simulations are pointed like asatisfactory tool. Using direct dynamics weavoid exploring the whole potential energysurface, which becomes really complicatedwhen dealing with big molecules.Since chemical dynamics simulations arerestricted to the short time scale reactivity,typically ~10ps, we make use of the Rice–Ramsperger–Kassel–Marcus (RRKM)unimolecular theory to study the reactivity atUniversité Paris-SaclayEspace Technologique / Immeuble DiscoveryRoute de l’Orme aux Merisiers RD 128 / 91190 Saint-Aubin, Francethat we use in the present thesis. The aim of CIDis to activate the rovibrational modes of an ionicmolecular system by collisions with an inert gas,increasing the probability of the ion of beingfragmented.Despite being a really useful technique, tandemmass spectrometry does not give informationabout the mechanisms of the reactions takingplace in the collision cell; in order to obtain suchinformation, chemicallonger time scales to understand reaction pathsthat take place after intramolecular vibrationrelaxation (IVR).In the present thesis we have chosen to study asmodel system of nucleobase the uracil molecule.Furthermore, we also studied the gas-phase reactivity of carbohydrates (cellobiose, maltose and gentiobiose), which were preliminarily derivatized in order to simplify the charge localization, and consequently the theoretical study.

<p>Bridging the accuracy of ab initio (AI) QM/MM with the efficiency of semi-empirical<br> (SE) QM/MM methods has long been a goal in computational chemistry. This dissertation<br> presents four ∆-Machine learning schemes aimed at achieving this objective. Firstly, the in-<br> corporation of negative force observations into the Gaussian process regression (GPR) model,<br> resulting in GPR with derivative observations, demonstrates the remarkable capability to<br> attain high-quality potential energy surfaces, accurate Cartesian force descriptions, and reli-<br> able free energy profiles using a training set of just 80 points. Secondly, the adaptation of the<br> sparse streaming GPR algorithm showcases the potential of memory retention from previous<br> phasespace, enabling energy-only models to converge using simple descriptors while faith-<br> fully reproducing high-quality potential energy surfaces and accurate free energy profiles.<br> Thirdly, the utilization of GPR with atomic environmental vectors as input features proves<br> effective in enhancing both potential energy surface and free energy description. Further-<br> more, incorporating derivative information on solute atoms further improves the accuracy<br> of force predictions on molecular mechanical (MM) atoms, addressing discrepancies arising<br> from QM/MM interaction energies between the target and base levels of theory. Finally, a<br> comprehensive comparison of three distinct GPR schemes, namely GAP, GPR with an aver-<br> age kernel, and GPR with a system-specific sum kernel, is conducted to evaluate the impact<br> of permutational invariance and atomistic learning on the model’s quality. Additionally, this<br> dissertation introduces the adaptation of the GAP method to be compatible with the sparse<br> variational Gaussian processes scheme and the streaming sparse GPR scheme, enhancing<br> their efficiency and applicability. Through these four ∆-Machine learning schemes, this dis-<br> sertation makes significant contributions to the field of computational chemistry, advancing<br> the quest for accurate potential energy surfaces, reliable force descriptions, and informative<br> free energy profiles in QM/MM simulations.<br> </p>

Fourier transform infrared difference spectroscopy (FTIR DS) is widely used to study the structural details of electron transfer cofactors in photosynthetic protein complexes. In photosynthetic proteins quinones play an important role, functioning as a cofactor in light-driven electron transfer. In photosystem I (PS I) phylloquinone (PhQ) functions as an intermediary in electron transfer. To investigate the properties of PhQ that occupies the, so called, A1 binding site in PS I, time-resolved step-scan FTIR DS, with 5µs time resolution at 77K has been used. By replacing PhQ in the A1 binding site with specifically isotope labeled version, information on the vibrational frequencies associated specifically with the quinone in the binding site were obtained, which could be compared to the vibrational properties of quinone in solution or quinones in other protein binding sites. To further aid in assessing the origin of bands in the spectra, quantum mechanics /molecular mechanics (QM/MM) ONIOM type calculations were undertaken. ONIOM is an acronym for Our own N-layered Integrated molecular Orbital and molecular Mechanics. We find that the phytyl tail of PhQ does not play an important role in the orientation of PhQ in the A1 binding site. We also find that PhQ, in both neutral and reduced states, is strongly hydrogen bonded. To test and verify the applicability of our QM/MM approach, ONIOM calculations were also undertaken for ubiquinone and a variety of other quinones incorporated into the, so called, QA binding site in purple bacteria photosynthetic reaction centers. The calculated and experimental spectra agree well, demonstrating the utility and applicability of our ONIOM approach. Hydrogen bonding to the carbonyl groups of quinones in the QA binding site was shown to be relatively weak, and it was found that hydrogen bonding to neutral ubiquinone in purple bacterial reaction centers can be considered in purely electrostatic terms, contrary to the widely held belief that the hydrogen bonding amino acids should be treated quantum mechanically.

No description available.

Des études antérieures ont mis en évidence que le transfert de solutés neutres à travers des membranes est influencé par la présence d'ions en solution. Ainsi, la connaissance des interactions multiples à l'échelle nanométrique, entre le polymère, l'eau et les solutés (ions, espèces organiques) constituent un verrou pour l'amélioration des performances des procédés membranaires. Dans cette étude une approche multi-échelle fondamentale est proposée, combinant des outils théoriques et expérimentaux, afin d'obtenir les paramètres microscopiques et macroscopiques caractérisant les interactions étudiées pour différentes compositions ioniques. Plus précisément, il s'agit de comprendre comment les ions affectent le transfert d'un soluté organique. Dans un premier temps, certaines propriétés caractérisant l'hydratation des ions sont calculées et comparées aux flux de diffusions de sucres à travers des membranes de Nanofiltration et échangeuses d'ions obtenus pour différentes compositions ioniques. Dans un deuxième temps, des systèmes constitués d'une membrane échangeuse de cations (CMX) équilibrée avec différents cations ainsi que le glucose hydraté sont modélisés en utilisant une approche combinée Mécanique Quantique/ Mécanique Moléculaire. Cette approche a permis d'étudier la solubilité du sucre dans la matrice polymère ainsi que les interactions polymère-polymère comme l'énergie de cohésion. Enfin, l'influence des ions sur les caractéristiques physiques de la membrane CMX est étudiée en utilisant diverses méthodes expérimentales comme la détermination des angles de contacts et des spectres IR ou la mesure de la température de solidification par DSC. Les propriétés vibrationnelles sont également calculées dans le cadre de la théorie de la fonctionnelle de la densité (DFT). L'ensemble de ces données sont comparées avec les grandeurs de transport afin de valider les mécanismes moléculaires proposés. Ce travail montre que la nature des contre-ions de la membrane modifie l'énergie de cohésion entre les fragments de la membrane. Ainsi, l'énergie de cohésion influe sur la diffusion des composés organiques neutres à travers les membranes. / Previous works have shown that the transfer of neutral solutes through membranes is influenced by the presence of ions in solution. In the framework of process intensification, the knowledge of the molecular mechanisms involved is of fundamental importance to increase and predict the process performances. The aim of this Thesis is to use a combined quantum/molecular computational approach and experimental methodologies to better understand how ions can affect the solute flux. In the first part of the work, some properties of ions in solution are computed and compared with sugar fluxes through membranes for nanofiltration and electrodialysis. In the following, systems composed of Cation-exchange membrane equilibrated by different counter-ion and hydrated glucose are examined by Quantum Mechanics/Molecular Mechanics. This is done mainly to investigate the sugar solubility in the polymer matrix and diffusion related interactions like polymer chain-chain cohesion energy. In the last part, contact angle, differential scanning calorimetry and Infra-Red spectra are measured to characterize the physical properties of the membrane and possible influence of the counter-ion on cation exchange membrane. This work shows that the nature of the counter-ions modifies the cohesion energy between the membrane polymer fragments. In its turn, the cohesion energy affects the diffusion of neutral organic compounds through the membranes.

This dissertation details the use of computational methods to understand the effect that cancer-related mutations have on proteins that complex with nucleic acids. Firstly, we perform molecular dynamics (MD) simulations of various mutations in DNA polymerase κ (pol κ). Through an experimental collaboration, we classify the mutations as more or less active than the wild type complex, depending upon the incoming nucleotide triphosphate. From these classifications we use quantum mechanics/molecular mechanics (QM/MM) to explore the reaction mechanism. Preliminary analysis points to a novel method for nucleotide addition in pol κ. Secondly, we study the ten-eleven translocation 2 (TET2) enzyme in various contexts. We find that the identities of both the substrate and complementary strands (or lack thereof) are crucial for maintaining the complex structure. Separately, we find that point mutations within the protein can affect structural features throughout the complex, only at distal sites, or only within the active site. The mutation's position within the complex alone is not indicative of its impact. Thirdly, we share a new method that combines direct coupling analysis and MD to predict potential rescue mutations using poly(ADP-ribose) polymerase 1 as a model enzyme. Fourthly, we perform MD simulations of mutations in the protection of telomeres 1 (POT1) enzyme. The investigated variants modify the POT1-ssDNA complex dynamics and protein—DNA interactions. Fifthly, we investigate the incorporation of remdesivir and other nucleotide analogue prodrugs into the protein-RNA complex of severe acute respiratory syndrome-coronavirus 2 RNA-dependent RNA polymerase. We find evidence for destabilization throughout the complex and differences in inter-subunit communication for most of the incorporation patterns studied. Finally, we share a method for determining a minimum active region for QM/MM simulations. The method is validated using 4-oxalocrotonate, TET2, and DNA polymerase λ as test cases.

<p>In this thesis, we present novel computational methods and frameworks to address the challenges associated with the determination of free energy profiles for condensed-phase chemical reactions using combined quantum mechanical and molecular mechanical (QM/MM) approaches. We focus on overcoming issues related to force matching, molecular polarizability, and convergence of free energy profiles. First, we introduce a method called Reaction Path-Force Matching in Collective Variables (RP-FM-CV) that efficiently carries out ab initio QM/MM free energy simulations through mean force fitting. This method provides accurate and robust simulations of solution-phase chemical reactions by significantly reducing deviations on the collective variables forces, thereby bringing simulated free energy profiles closer to experimental and benchmark AI/MM results. Second, we explore the role of pairwise repulsive correcting potentials in generating converged free energy profiles for chemical reactions using QM/MM simulations. We develop a free energy correcting model that sheds light on the behavior of repulsive pairwise potentials with large force deviations in collective variables. Our findings contribute to a deeper understanding of force matching models, paving the way for more accurate predictions of free energy profiles in chemical reactions. Next, we address the underpolarization problem in semiempirical (SE) molecular orbital methods by introducing a hybrid framework called doubly polarized QM/MM (dp-QM/MM). This framework improves the response property of SE/MM methods through high-level molecular polarizability fitting using machine learning (ML)-derived corrective polarizabilities, referred to as chaperone polarizabilities. We demonstrate the effectiveness of the dp-QM/MM method in simulating the Menshutkin reaction in water, showing that ML chaperones significantly reduce the error in solute molecular polarizability, bringing simulated free energy profiles closer to experimental results. In summary, this thesis presents a series of novel methods and frameworks that improve the accuracy and reliability of free energy profile estimations in condensed-phase chemical reactions using QM/MM simulations. By addressing the challenges of force matching, molecular polarizability, and convergence, these advancements have the potential to impact various fields, including computational chemistry, materials science, and drug design.</p>

Indiana University-Purdue University Indianapolis (IUPUI) / In this thesis, we present novel computational methods and frameworks to address the challenges associated with the determination of free energy profiles for condensed-phase chemical reactions using combined quantum mechanical and molecular mechanical (QM/MM) approaches. We focus on overcoming issues related to force matching, molecular polarizability, and convergence of free energy profiles. First, we introduce a method called Reaction Path-Force Matching in Collective Variables (RP-FM-CV) that efficiently carries out ab initio QM/MM free energy simulations through mean force fitting. This method provides accurate and robust simulations of solution-phase chemical reactions by significantly reducing deviations on the collective variables forces, thereby bringing simulated free energy profiles closer to experimental and benchmark AI/MM results. Second, we explore the role of pairwise repulsive correcting potentials in generating converged free energy profiles for chemical reactions using QM/MM simulations. We develop a free energy correcting model that sheds light on the behavior of repulsive pairwise potentials with large force deviations in collective variables. Our findings contribute to a deeper understanding of force matching models, paving the way for more accurate predictions of free energy profiles in chemical reactions. Next, we address the underpolarization problem in semiempirical (SE) molecular orbital methods by introducing a hybrid framework called doubly polarized QM/MM (dp-QM/MM). This framework improves the response property of SE/MM methods through high-level molecular polarizability fitting using machine learning (ML)-derived corrective polarizabilities, referred to as chaperone polarizabilities. We demonstrate the effectiveness of the dp-QM/MM method in simulating the Menshutkin reaction in water, showing that ML chaperones significantly reduce the error in solute molecular polarizability, bringing simulated free energy profiles closer to experimental results. In summary, this thesis presents a series of novel methods and frameworks that improve the accuracy and reliability of free energy profile estimations in condensed-phase chemical reactions using QM/MM simulations. By addressing the challenges of force matching, molecular polarizability, and convergence, these advancements have the potential to impact various fields, including computational chemistry, materials science, and drug design.