Ferramentas e serviços online para a análise da origem cladística de genes e vias metabólicas

• Ano de defesa: 2014
• Autor(a) principal: Henrique Velloso Ferreira Melo
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Bioinformatica; Genes; Bioinformática
• Link de acesso: http://hdl.handle.net/1843/BUOS-B7UN89

The origins of genes have always fascinated scientists, and the list of proposed mechanisms for the origin of new genes has progressively increased. The exploitation of such mechanisms would greatly benefit from a description of the scenario underlying the rise of genes throughout evolution. This work creates tools for the description of this scenario andassesses the emergence of genes on large scale, not in chronological time, but in a cladistic context. The generated tools, all available online, can be grouped into three products: (i) a set of tools and services named BioTools Service (ii) the Genesis website and (iii) a Web services platform called BOWS. BioTools Service is a suite of scripts and services targeted to process taxonomy data and infer the origin of genes. It consists of four byproducts: TaxSimple, GetAllChildren, LCA and GeneOriginUEKO. TaxSimple flattens the hierarchical taxonomicdata from NCBI in order to facilitate taxonomic queries. GetAllChildren is a Web service that returns all leaf nodes descendants of a given clade in the tree of life. LCA (Lowest Common Ancestor) is a script that returns the common ancestor of two or more given nodes. GeneOriginUEKO can infer the origin of genes using UEKO clusters of orthologous (anenriched version of the KEGG ORTHOLOGY database). Graphs generated based on the results show wave patterns of human gene emergence with peaks in certain clades, such as Euteleostomi and Eutheria. It was also observed that some human metabolic pathways were increasingly gaining elements throughout evolution, as seen for the JAK-STAT signaling pathway. However, it was found that horizontal transfer events could cause the LCA to incorrectly move toward the root of the tree of life. To work around this issue, the multiple LCA was developed, which isolates horizontal transfers with the aid of complete genomes and thereby infer multiple origins of a gene. The Genesis website, built based on this algorithm, provides a graphical interface to the main results of the gene multiple origins analysis and also permits performing of analyses with custom input data. At least this work presents the BOWS platform, a system that centralizes and provides universal and secure access to bioinformatics tools installed on remote computational clusters. BOWS can be installed in any lab and is already being used successfully in Biodados Lab at UFMG