Network methods for integrative Omics and pathway analysis
Tutor / director / avaluadorSánchez Pla, Àlex
Tipus de documentProjecte Final de Màster Oficial
Condicions d'accésAccés restringit per decisió de l'autor
Omics data analysis is more accessible nowadays and it will become even more accessible in the future. A broad range of analysis can be done individually on each type of Omics, leading to conclusions on the factor of interest. Yet, due to human s variability, these results are not always concordant. Understanding the biological implications after combining different types of Omics together is of great interest and may reveal new results, only visible with an integrated approach. Different approaches to integrative Omics data analysis exist. Multivariate techniques, which provide dimension reduction approaches and a great numerical flexibility to Omics data, but fail to ease interpretation of the results; machine learning techniques, well-known for biomarker discovery; and network analysis approaches, which are yet in development. The aim of this thesis is to review methodologies developed in the field of network analysis related to integrative Omics data analysis and for pathway analysis. A state-of-the art review has been done, describing approaches for network-based integrative data analysis and their limitations. One of these approaches has also been tested in two case studies. On the other hand, a comparison of tools for pathway analysis in metabolomics is also performed. Even though different statistical approaches can be used to analyse Omics data with an integrative approach, and network-based integrative analysis being in a very juvenile stage yet, it may be the most suitable approach to take the logical step beyond statistics, leading to a more comprehensive approach for biologists.