Development of bioinformatics resources for the integrative analysis of Next Generation omics data

Abstract

The advances in high-throughput sequencing techniques and the technological development accompanying them have favoured the development and popularisation of a new range of genomic research disciplines, collectively known as the omics. These technologies are capable of simultaneously measuring thousands of molecules which are essential for life, including DNA, RNA, proteins, and metabolites. Historically, classical genomic research has followed a reductionist approach by studying the structure, regulation, and function of these biological units independently. However, despite being a powerful analytical tool, the reductionist method cannot explain many of the biological phenomena that take place in living systems. This is because these biological events are not represented by the sum of their components, rather, only the interacting dynamics of the different omics elements can explain their complexity. In recent years Systems Biology has established itself as a multidisciplinary area of research which tries to model the dynamic behaviour of biological systems by holistically studying the interactions between the different omics disciplines; it combines simultaneous measurements of different types of molecules and integrates multiple sources of information in order to identify changing components in a coordinated way and under controlled study conditions. Thus, Systems Biology is an interdisciplinary area that requires biologists, mathematicians, biochemists, and other researchers to work closely together, and in which computer sciences plays a fundamental role because of the volume and complexity of the data handled. This thesis addresses the problem of data management, integration, and analysis in multi-omics studies. More specifically, this research focused on two of the most characteristic computational challenges in Systems Biology: the development of integrated databases and the problem of integrative visualisation. Therefore, the first part of this work was devoted to designing and creating a bioinformatics resource for managing multi-omics experiments. The resulting platform, known as STATegra EMS, offers a complete set of tools that facilitate the storage and organisation of the large datasets generated during omics experiments, and also provides tools for data annotation in the later stages of processing and analysis of the information. The development of this platform required overcoming problems created by the heterogeneity, volume, and high variability of the data. Thus, as part of the solution to these problems, detailed metadata can be recorded within STATegra EMS, allowing dataset discrimination and successful data integration. To aid this process, the platform also offers a collaborative and easy-to-use web interface that combines modern web technologies and well-known community standards to represent the different components of the integrated experiments. The second part of this thesis examines the current situation and challenges in integrative data visualisation in multi-omic experiments, and presents the PaintOmics 3 web tool which was developed to address these issues. Since the capacity of the human brain for visual processing is highly evolved, integrative visualisation combined with data analysis techniques is probably one of the most powerful tools for interpreting and validating results in Systems Biology. PaintOmics 3 provides a comprehensive framework for performing biological function enrichment analyses in experiments with multiple conditions and data types; it combines powerful tools for integrative data visualisation on KEGG molecular-interaction diagrams, biological-process interaction-networks, and statistical analyses. Moreover, unlike similar tools, PaintOmics 3 is interactive and easy to use, and stands out for its flexibility and the variety of omics data types it accepts, which include epigenomics data based on genomic regions, proteomics data, and miRNA-study data.