Il est estimé qu'une fracture se produit toutes les trois secondes autour du monde, accompagné par un risque élevé d'invalidité ou même de mortalité. La connaissance des mécanismes de fractures dans une configuration de chargement représentatif d'une chute semble être d'un intérêt majeur pour le développement de méthodes dédiées à la prédiction du risque de fracture. La ténacité est un paramètre approprié lorsqu'on s'intéresse à ces mécanismes de fracture, elle détermine l'énergie nécessaire pour propager une fissure à travers l'architecture du tissu. L'objectif de cette étude est d'évaluer la ténacité de l'os cortical humain, considérant à la fois des conditions chargement quasi-statique et représentatif d'une chute sur sites anatomiques appariés. L'acquisition d'images en micro-tomographie ainsi qu'une mesure des cross-links ont été réalisées afin d'évaluer leur influence sur les mécanismes fracture du tissu. Les résultats ont montré que dans des conditions quasi-statiques, les différents sites anatomiques présentent des propriétés mécaniques différentes : le radius résiste mieux à une propagation de fissure. Dans des conditions de chute, il n'y a plus de différences entre ces sites, mais la ténacité décroit de façon significative par rapport au chargement standard. L'os cortical résiste mieux à une propagation de fissure dans des conditions quasi-statiques. Les analyses structurales et biochimiques ont montré des différences entre les sites anatomiques qui expliquent les différences mécaniques. Les caractéristiques architecturales du tissu sont déterminantes vis-à-vis des mécanismes de fracture dans des conditions quasi-statiques. Mais leur rôle lors d'une chute est moins évident. Ces résultats impliquent que la microstructure de l'os cortical n'est pas un déterminant majeur vis-à-vis du risque de fracture. De futures études doivent être réalisées afin de déterminer les paramètres décisifs dans des conditions représentatives d'une chute / A fracture is estimated every three seconds in the world, leading to an increased risk of impairment or even mortality. The biomechanical knowledge of bone fracture mechanisms in a fall configuration of loading is of great interests for the development of clinical method for the prediction of the risk of fracture. Toughness seems to be a good candidate to investigate this fracture process as it corresponds to the energy needed to propagate a crack through cortical bone complex microstructure. The aim of this study was thus to evaluate human cortical bone toughness parameter under both quasi-static and fall-like loading conditions paired anatomical locations. Micro-computed tomography images using synchrotron radiation and collagen cross-links maturation measurements were performed to investigate the influence of the tissue architecture on crack propagation. Results found showed that under quasi-static condition, the different anatomical locations present different mechanical behavior. Radius significantly better resist crack propagation than the other studied location. Considering a fall-like loading condition, no more difference is observed between the locations but a significant decreased is measured compare to the first configuration. Human cortical bone has a better capacity to resist crack propagation under a standard quasi-static loading condition. By investigating the tissue morphometric and biochemical parameters, we observed different organization from a location to another that explains the mechanical differences. The architectural features appear to be determinant for crack propagation mechanisms under quasi-static condition, but they play a lesser role under fall-like condition. These results imply that the tissue microstructure is not a determinant when dealing with the prediction of the risk of fracture. Further work has to be done to reach out which parameters are more determinants under a specific fall-like loading condition

Whole-exome and genome sequencing are widely used to diagnose individual patients. However, despite its success, this approach leaves many patients undiagnosed. This could be due to the need to discover more disease genes and variants or because disease phenotypes are novel and arise from a combination of variants of multiple known genes related to the disease. Recent rapid increases in available genomic, biomedical, and phenotypic data enable computational analyses, reducing the search space for disease-causing genes or variants and facilitating the prediction of causal variants. Therefore, artificial intelligence, data mining, machine learning, and deep learning are essential tools that have been used to identify biological interactions, including protein-protein interactions, gene-disease predictions, and variant--disease associations. Predicting these biological associations is a critical step in diagnosing patients with rare or complex diseases. In recent years, computational methods have emerged to improve gene-disease prioritization by incorporating phenotype information. These methods evaluate a patient's phenotype against a database of gene-phenotype associations to identify the closest match. However, inadequate knowledge of phenotypes linked with specific genes in humans and model organisms limits the effectiveness of the prediction. Information about gene product functions and anatomical locations of gene expression is accessible for many genes and can be associated with phenotypes through ontologies and machine-learning models. Incorporating this information can enhance gene-disease prioritization methods and more accurately identify potential disease-causing genes. This dissertation aims to address key limitations in gene-disease prediction and variant prioritization by developing computational methods that systematically relate human phenotypes that arise as a consequence of the loss or change of gene function to gene functions and anatomical and cellular locations of activity. To achieve this objective, this work focuses on crucial problems in the causative variant prioritization pipeline and presents novel computational methods that significantly improve prediction performance by leveraging large background knowledge data and integrating multiple techniques. Therefore, this dissertation presents novel approaches that utilize graph-based machine-learning techniques to leverage biomedical ontologies and linked biological data as background knowledge graphs. The methods employ representation learning with knowledge graphs and introduce generic models that address computational problems in gene-disease associations and variant prioritization. I demonstrate that my approach is capable of compensating for incomplete information in public databases and efficiently integrating with other biomedical data for similar prediction tasks. Moreover, my methods outperform other relevant approaches that rely on manually crafted features and laborious pre-processing. I systematically evaluate our methods and illustrate their potential applications for data analytics in biomedicine. Finally, I demonstrate how our prediction tools can be used in the clinic to assist geneticists in decision-making. In summary, this dissertation contributes to the development of more effective methods for predicting disease-causing variants and advancing precision medicine.