This dissertation addresses problems arising in freight routing and scheduling where full truckload (FTL) and less-than-truckload (LTL) carriers are used to serve transportation needs. Each of the problems investigated in this dissertation tries to optimize/maximize consolidation to decrease system transportation costs by (1) carefully choosing the timing and path of freight and/or (2) introducing consolidation points. Approaches are proposed that enable effective planning and operation of freight routing and scheduling for large-scale transportation networks. Chapter 2 presents solution approaches for a shipper pickup and delivery planning problem faced by many large retailers to move freight from suppliers to distribution centers. Each shipment is moved either direct via a LTL carrier or possibly consolidated with other shipments and moved by one or two FTL routes. When using a FTL carrier, the shipper takes advantage of contracted lane rates that establish prices per mile for a truck operated between two locations that are significantly less than the comparable LTL price for shipping a full truckload. Consolidated FTL routes may each visit multiple shipment origins (supplier locations) and/or destinations (distribution center locations). Additionally, FTL routes may move shipments through a single crossdock facility en route. The challenge in this planning problem is to exploit as much as possible negotiated truckload lane rates and to judiciously make use of routes through crossdock facilities to consolidate shipments. The primary contributions of this section are that (1) an interesting new problem variant is introduced to the field of transportation and logistics that is important in practice and (2) the solution approach demonstrates that exploiting knowledge of the problem and solution structure to cleverly select subsets of path variables for evaluation during each iteration of an integer programming based local search heuristic is effective on path-based routing models. Chapter 3 evaluates how to route each customer shipment through a sequence of transfer terminals in a LTL carrier network. At each terminal stop, a shipment is unloaded from an inbound trailer and reloaded onto an outbound trailer. A load plan determines the specific sequence of terminal transfers to be used for freight moving between each origin and destination. The design of the load plan determines the linehaul transportation and handling costs required to serve customers. We develop an improved very large-scale neighborhood search heuristic for solving an integer programming model for load plan design. The main contributions of this section include (1) the investigation of the pros and cons of optimizing system-wide into a single destination versus optimizing freight for all destinations in a small region, and (2) a solution approach that can find load plans with costs 6 to 7\% lower than those used in practice, and can find 2.5 to 5\% additional cost savings using the same time budget when compared to an approach optimizing system-wide into a single destination. Chapter 4 addresses a strategic planning problem that extends the load plan design problem to consider terminal roles. We investigate two-stage approaches that first identify the set of transfer terminals and then develop the corresponding load plan. Computational results compare the terminals chosen as transfer facilities from the proposed integer programming based local search method with a traditional hub location formulation and a simple facility location formulation to depict the benefits gained from modeling additional information. The key contributions of this section are (1) the introduction of a new hub location problem variant incorporating freight dispatch timing and trailer transportation cost characteristics found in the LTL trucking industry and (2) a solution approach utilizing IP-based local search that demonstrates the importance of incorporating freight dispatch timing. Finally, Chapter 5 summarizes the main conclusions from this dissertation and discusses directions for further research.

Les fonctions de la majorité des protéines sont subordonnées à l'interaction avec un ou plusieurs partenaires : acides nucléiques, autres protéines... La plupart de ces interactions sont transitoires, difficiles à détecter expérimentalement et leurs structures sont souvent impossible à obtenir. C'est pourquoi la prédiction in silico de l'existence de ces interactions et la structure du complexe résultant ont été l'objet de nombreuses études depuis plus d'une décennie maintenant. Pour autant les protéines sont des objets complexes et les méthodes informatiques classiques sont trop "gourmandes" en temps pour l'exploration à grande échelle de l'interactome des différents organismes. Dans ce contexte de développement d'une méthode de docking protéine-protéine haut débit nous présenterons ici l'implémentation d'une nouvelle méthode d'amarrage, celle-ci est basée sur : l'utilisation de deux types de formalismes : Les tessellations de Voronoï et Laguerre permettant la manipulation de modèles géométriques simplifiés permettant une bonne modélisation des complexes et des temps de calcul plus raisonnable qu'en représentation atomique. L'utilisation et l'optimisation d'algorithmes d'apprentissage (algorithmes génétiques) permettant d'isoler les conformations les plus pertinentes entre deux partenaires protéiques. Une méthode d'évaluation basée sur le clustering de méta-attributs calculés au niveau de l'interface permettant de trier au mieux ce sous-ensemble de conformations candidates.

Compatible interactions between male gametophytes (pollen) and the female reproductive organ (pistil) are essential for fertilization in flowering plants. Recognition at a molecular level allows “compatible” pollen grains to adhere/germinate on the stigma while pollen grains from unrelated plant species are largely ignored. The exocyst is a large eight subunit complex that is primarily involved in polarized secretion or regulated exocytosis in eukaryotic cells where it functions to tether vesicles to the plasma membrane. Recent research has implicated one of the Exo70 family members, Exo70A1, in compatible pollen-pistil interactions in Arabidopsis and Brassica. The loss of Exo70A1 in Arabidopsis Col-0 stigmas leads to the rejection of compatible pollen producing a “female sterile” phenotype. Through my research I have demonstrated that, driven by a stigma-specific promoter, an RFP:Exo70A1 fusion protein rescues this defect in exo70A1-1 mutant and Exo70A1 is found to be localized to the plasma membrane at flower opening.

Compatible interactions between male gametophytes (pollen) and the female reproductive organ (pistil) are essential for fertilization in flowering plants. Recognition at a molecular level allows “compatible” pollen grains to adhere/germinate on the stigma while pollen grains from unrelated plant species are largely ignored. The exocyst is a large eight subunit complex that is primarily involved in polarized secretion or regulated exocytosis in eukaryotic cells where it functions to tether vesicles to the plasma membrane. Recent research has implicated one of the Exo70 family members, Exo70A1, in compatible pollen-pistil interactions in Arabidopsis and Brassica. The loss of Exo70A1 in Arabidopsis Col-0 stigmas leads to the rejection of compatible pollen producing a “female sterile” phenotype. Through my research I have demonstrated that, driven by a stigma-specific promoter, an RFP:Exo70A1 fusion protein rescues this defect in exo70A1-1 mutant and Exo70A1 is found to be localized to the plasma membrane at flower opening.

La presente tesis "Modelling of 5-HT2A and 5-HT2C receptors and of their complexes with actual and potential antipsychotic drugs" tiene como objetivo de profundizar los conocimientos actuales sobre el mecanismo de acción de los fármacos antipsicóticos. En este proyecto de larga duración, se han construidos modelos computacionales de los receptores 5-HT2A y 5-HT2C, utilizando un nuevo protocolo de modelización basado sobre los datos experimentales de otras proteínas GPCR de la misma familia. Las estructuras 3D se han validado e utilizado en estudios de acoplamiento ligando-receptor, simulaciones de dinámica molecular, y estudios 3D-QSAR con el ligando natural (serotonina), un agonista inverso bien conocido (ketanserina) y una serie de butyrofenonas con afinidad para ambos subtipos receptoriales. Las metodologías directas e indirectas utilizadas, han permitido de comprender mejor los elementos claves que gobiernan el acoplamiento ligando - receptor, mediante la identificación de los residuos más involucrados en esta interacción, el rol de la quiralidad de los ligandos y también las posiciones alternativas de acoplamiento que algunos ligandos pueden asumir en el sitio de unión de los receptores.Los resultados son coherentes con los datos experimentales y su interpretación ha proporcionado información valiosa, difícilmente obtenible con la simple inspección visual de las estructuras de los ligandos y de los receptores. / This thesis "Modelling of 5-HT2A and 5-HT2C receptors and of their complexes with actual and potential antipsychotic drugs" has the objective of investigate the mechanism of action of antipsychotic drugs. During the development of this project, computational models of 5-HT2A and 5-HT2C receptors have been built, by means of a new modeling protocol based on experimental data from other GPCR of the same family. 3D structures have been validated by means of docking, molecular dynamic simulations and 3D-QSAR studies, using the natural ligand (serotonin), a well known inverse agonist (ketanserin) and a series of butyrophenones with affinity for both receptor subtypes. Direct and indirect methodologies have been applied, allowing a better comprehension of the key elements governing the ligand-receptor docking, thanks to the identification of the most important residues that stabilize such interaction, role of chirality and alternative positions within the binding site. The results are coherent with experimental data and its interpretation provided valuable information, not available at a simple visual inspection of ligand - receptor structures.

Die Dissertation beschreibt Arbeiten zur Herstellung neuer Cetrorelixkomplexe, zur Kinetik der dynamischen Liberation, zur Struktur von Aggregaten und Komplexen von Cetrorelix und zur Berechnung von Komplexeigenschaften mit Molecular Modeling. / Presented are results about new complexes of cetrorelix, the kinetics of dynamical liberations, the structure of Cetrorelix aggregates and complexes and the computation of properties of complexes by molecular modeling.

Darbo tikslas nustatyti skirstymo baltymo GAB1 ryšį su anti-apoptoziniu PI3K/Akt bei mitogeniniu Ras/MAPK signalo perdavimo keliais ir įvertinti GAB1 įtaką šių kelių sąveikai EGFR signalo perdavimo tinkle. Darbo uždaviniai: įvertinti epitelinių ląstelių endogeninio GAB1 veiksmingumą EGF signalo perdavimo metu; nustatyti sąveikos pobūdį tarp PI3K/Akt ir Ras/MAPK kelių EGF signalo metu; kiekybiškai įvertinti GAB1 svarbą EGF signalo perdavimui per PI3K/Akt ir Ras/MAPK kelių in vivo, rezultatus lyginant su matematinio modelio prognozėmis in silico; nustatyti GAB1 veiksmingumo ir jo reguliacijos grįžtamaisiais ryšiais įtaką PI3K-MAPK sąveikos stiprumui priklausomai nuo EGF dozės ir laiko; ištirti GAB1 svarbą EGFR ir insulino receptoriaus signalo perdavimo tinklų sąveikai; modifikuoti Westerno pernašos metodą palyginamajai kiekybinei ir kokybinei baltymų analizei. Darbo išvados: stimuliavus EGFR, skirstymo baltymas GAB1 tampa veiksmingu; EGF lemia reciprokinę PI3K/Akt ir Ras/MAPK signalo perdavimo kelių sąveiką per GAB1; GAB1 yra pagrindinis teigiamo atgalinio ryšio elementas PI3K kelyje, padedąs pagreitinti, stiprinti ir išlaikyti MEK/ERK kinazių atsaką; PI3K-MAPK sąveikos stiprumas kinta laike ir yra atvirkščiai proporcingas EGF signalo stiprumui; GAB1 reikalingas sinergistiškai stiprinti insulinu ir mažomis EGF dozėmis stimuliuojamų ląstelių Ras/MAPK atsaką; sukurtas „Multi-juostelių“ imunopernašos metodas yra tinkamas palyginamajai kiekybinei ir kokybinei baltymų analizei... [toliau žr. visą tekstą] / The aim of the thesis was to determine a connection of endogenous docking protein GAB1 with anti-apoptotic PI3/Akt and Ras/MAPK signaling pathways and to estimate GAB1 contribution to their crosstalk in EGFR signaling network. The tasks: to evaluate GAB1 efficacy upon EGFR stimulation; to examine the nature of crosstalk between PI3K/Akt and Ras/MAPK pathways; to evaluate the contribution of GAB1 to EGF signaling via PI3K/Akt and Ras/MAPK pathways in vivo comparing the results with prognosis in silico; to estimate the EGF dose- and time-dependent impact of GAB1 efficacy and its feedback regulation on the strength of PI3K-MAPK interaction; to investigate the role of GAB1 for crosstalk of EGFR and insulin receptor signaling networks; to modify Western blotting procedure for comparative quantitative and qualitative protein analysis. The conclusions: the docking protein GAB1 is functional upon EGFR stimulation; PI3K/Akt and Ras/MAPK signaling pathways crosstalk reciprocally via GAB1 in response to EGF; GAB1 is major positive feedback element in PI3K pathway amplifying and sustaining MEK/ERK response to EGF; the strength of PI3K-MAPK interaction depends on time and is inversely proportional to EGF signal strength; GAB1 is required to synergistically potentate the Ras/MAPK response to tandem cell treatment with insulin and low EGF doses; the developed Multistrip immunoblotting method is suitable for comparative quantitative and qualitative protein analysis. In comparison with a... [to full text]

La transglutaminase microbienne (Microbial transglutaminase : MTG) est fortement exploitée dans l’industrie textile et alimentaire afin de modifier l’apparence et la texture de divers produits. Elle catalyse la formation de liaisons iso-peptidiques entre des protéines par l’entremise d’une réaction de transfert d’acyle entre le groupement γ-carboxamide d’une glutamine provenant d’un substrat donneur d’acyle, et le groupement ε-amino d’une lysine provenant d’un substrat accepteur d’acyle. La MTG est tolérante à un large éventail de conditions réactionnelles, ce qui rend propice le développement de cette enzyme en tant que biocatalyseur. Ayant pour but le développement de la MTG en tant qu’alternative plus soutenable à la synthèse d’amides, nous avons étudié la réactivité d’une gamme de substrats donneurs et accepteurs non-naturels. Des composés chimiquement diversifiés, de faible masse moléculaire, ont été testés en tant que substrats accepteurs alternatifs. Il fut démontré que la MTG accepte une large gamme de composés à cet effet. Nous avons démontré, pour la première fois, que des acides aminés non-ramifiés et courts, tels la glycine, peuvent servir de substrat accepteur. Les α-acides aminés estérifiés Thr, Ser, Cys et Trp, mais pas Ile, sont également réactifs. En étendant la recherche à des composés non-naturels, il fut observé qu’un cycle aromatique est bénéfique pour la réactivité, bien que les substituants réduisent l’activité. Fait notable, des amines de faible masse moléculaire, portant les groupements de forte densité électronique azidure ou alcyne, sont très réactives. La MTG catalyse donc efficacement la modification de peptides qui pourront ensuite être modifiés ou marqués par la chimie ‘click’. Ainsi, la MTG accepte une variété de substrats accepteurs naturels et non-naturels, élargissant la portée de modification des peptides contenant la glutamine. Afin de sonder le potentiel biocatalytique de la MTG par rapport aux substrats donneurs, des analogues plus petits du peptide modèle Z-Gln-Gly furent testés; aucun n’a réagi. Nous avons toutefois démontré, pour la première fois, la faible réactivité d’esters en tant que substrats donneurs de la MTG. L’éventuelle amélioration de cette réactivité permettrait de faire de la MTG un biocatalyseur plus général pour la synthèse d’amides. Mots clés: Lien amide, biocatalyse, biotransformation, transglutaminase, arrimage moléculaire, criblage de substrats, ingénierie de substrats. / Microbial transglutaminase (MTG) is used extensively in the food and textile industry to alter the appearance and texture of products. MTG catalyses the formation of isopeptide linkages between proteins by an acyl transfer reaction between the γ-carboxamide group of a glutamine ‘acyl-donor’ substrate, and the ε-amino group of a lysine ‘acyl-acceptor’ substrate. MTG is tolerant to a broad range of reaction conditions and is therefore suitable for further development as a biocatalyst. Toward developing MTG as a “green” alternative for amide synthesis, we have investigated a range of non-native donor and acceptor substrates to probe the scope of MTG reactivity. Small, chemically varied compounds were tested as alternative acyl-acceptor substrates. We observed a broad acceptor specificity. We show, for the first time, that very short-chain alkyl-based amino acids such as glycine can serve as acceptor substrates. The esterified α-amino acids Thr, Ser, Cys and Trp – but not Ile – also show reactivity. Extending the search to non-natural compounds, an aromatic ring was observed to be beneficial for reactivity, although ring substituents reduced reactivity. Overall, bonding of the amine to a less hindered carbon increases reactivity. Importantly, very small amines carrying either the electron-rich azide or the alkyne groups required for click chemistry were highly reactive as acceptor substrates, providing a ready route to minimally modified, ‘clickable’ peptides. These results demonstrate that MTG is tolerant to a variety of chemically varied natural and non-natural acceptor substrates, which broadens the scope for modification of glutamine-containing peptides. To further probe the biocatalytic potential of MTG in terms of the donor substrate, smaller analogues of the model substrate Z-Gln-Gly were tested. We did not find product formation with substrates smaller than the model substrate. We observed, for the first time, trace esterase activity with MTG. Future improvement of this activity would render MTG a more attractive, general biocatalyst for amide bond formation.

Proteins and the formation of large protein complexes are essential parts of living organisms. Proteins are present in all aspects of life processes, performing a multitude of various functions ranging from being structural components of cells, to facilitating the passage of certain molecules between various regions of cells. The 'protein docking problem' refers to the computational method of predicting the appropriate matching pair of a protein (receptor) with respect to another protein (ligand), when attempting to bind to one another to form a stable complex. Research shows that matching the three-dimensional (3D) geometric structures of candidate proteins plays a key role in determining a so-called docking pair, which is one of the key aspects of the Computer Aided Drug Design process. However, the active sites which are responsible for binding do not always present a rigid-body shape matching problem. Rather, they may undergo sufficient deformation when docking occurs, which complicates the problem of finding a match. To address this issue, we present an isometry-invariant and topologically robust partial shape matching method for finding complementary protein binding sites, which we call the ProtoDock algorithm. The ProtoDock algorithm comes in two variations. The first version performs a partial shape complementarity matching by initially segmenting the underlying protein object mesh into smaller portions using a spectral mesh segmentation approach. The Heat Kernel Signature (HKS), the underlying basis of our shape descriptor, is subsequently computed for the obtained segments. A final descriptor vector is constructed from the Heat Kernel Signatures and used as the basis for the segment matching. The three different descriptor methods employed are, the accepted Bag of Features (BoF) technique, and our two novel approaches, Closest Medoid Set (CMS) and Medoid Set Average (MSA). The second variation of our ProtoDock algorithm aims to perform the partial matching by utilizing the pointwise HKS descriptors. The use of the pointwise HKS is mainly motivated by the suggestion that, at adequate times, the Heat Kernel Signature of a point on a surface sufficiently describes its neighbourhood. Hence, the HKS of a point may serve as the representative descriptor of its given region of which it forms a part. We propose three (3) sampling methods---Uniform, Random, and Segment-based Random sampling---for selecting these points for the partial matching. Random and Segment-based Random sampling both prove superior to the Uniform sampling method. Our experimental results, run against the Protein-Protein Benchmark 4.0, demonstrate the viability of our approach, in that, it successfully returns known binding segments for known pairing proteins. Furthermore, our ProtoDock-1 algorithm still still yields good results for low resolution protein meshes. This results in even faster processing and matching times with sufficiently reduced computational requirements when obtaining the HKS.

L’ADN, support de l’information génétique, est constamment altéré par des agents physiques ou chimiques d’origines endogènes (métabolisme) et exogènes (UV, radiations ionisantes, produits chimiques) dont les effets sont génotoxiques. Ces modifications structurales délétères de l’ADN sont éliminées par de nombreux mécanismes de réparation. Parmi eux, le système de réparation par excision de bases (BER) est initié par les ADN glycosylases qui reconnaissent et éliminent les bases endommagées. Dans certaines stratégies anti-cancéreuses, l’utilisation de la chimiothérapie et la radiothérapie ont pour but la destruction des cellules cancéreuses en altérant leur ADN. Dans ce contexte, les ADN glycosylases réparent l’ADN des cellules traitées et induisent une résistance non désirée au traitement, faisant de ces enzymes des cibles thérapeutiques intéressantes. Le but de ces travaux est d’approfondir la compréhension des mécanismes de réparation des ADN glycosylases de la superfamille structurale Fpg/Nei grâce à la modélisation moléculaire et de pouvoir identifier et concevoir des inhibiteurs de ces enzymes. Les simulations de dynamique moléculaire (DM) nous ont permis d’étudier la « Lesion Capping Loop » (LCL) et de l’associer à la stabilisation de la base endommagée positionnée dans le site actif. Nous avons également étudié les chemins de sortie possibles de la base après coupure par l’enzyme et l’implication de la boucle LCL dans ce phénomène grâce à des simulations de DM ciblée (TMD-1). De plus, les simulations de DM couplées à un protocole d’amarrage moléculaire « aveugle » nous ont permis d’identifier 2 sites de fixations possibles majoritaires pour des petites molécules potentiellement inhibitrices. Un de ces sites correspondant au site actif de hNEIL1 a fait l’objet d’un criblage virtuel d’une partie de la base de molécules Ambinter. Ceci nous a permis d’identifier des molécules potentiellement inhibitrices dont les effets seront prochainement testés in vitro dans l’équipe sur la protéine humaine hNeil1. / The DNA, genetic information support, is frequently damaged by physical or chemical agents from endogenous (cell metabolism) and exogenous (UV, ionizing radiations, chemicals) factors whose effects are genotoxic. These deleterious DNA structural alterations are removed by many DNA repair mechanisms. Among them, the base excision repair (BER) is initiated by DNA glycosylases which recognize and remove damaged bases. In some anti-cancer strategies, the use of chemo- and radiotherapy is aimed to cancerous cells destruction by altering their DNA. In that specific context, DNA glycosylases repair the DNA of treated cells and induce unwanted resistance to treatments, making these enzymes interesting therapeutic targets. The purpose of this work is to deepen the repair mechanism knowledge of Fpg/Nei structural superfamily of DNA glycosylases using molecular modeling and designing inhibitors of these enzymes. Molecular dynamic simulations allowed us to study the « Lesion Capping Loop » (LCL) and to associate its role to substrate stabilization in the enzyme active site. We also studied some possible excision’s product release pathways and LCL implication in this phenomena by targeted molecular dynamic simulations (TMD-1). Furthermore, molecular dynamic simulations coupled to a blind molecular docking protocol allowed us to identify 2 possible main binding sites of potential inhibitiors. One of these binding sites corresponding to the hNEIL1 active site has been the object of a virtual screening of the Greenpharma database. This allowed us to identify potential inhibitors whom effects will be soon tested in vitro on the humain protein hNEIL1.