| Resumo: | Genome-scale metabolic models are based on the well-known stoichiometry of biochemical reactions. Usually, the main objective of a reconstruction is the in silica simulation of the phenotypic behaviour of a microorganism, under different environmental and genetic conditions, thus representing an important tool in Metabolic Engineering. The first step of this type of reconstruction is the genome annotation, which is an essential step, since precursory data can be retrieved for the model reconstruction. There are currently more than 4.000 fully sequenced genomes, with more than 700 being drafted right now. Hence tools such as BLAST, one of the most widely used bioinformatics tools, are being used to establish sequence similarities between genomes. Whereas the reconstruction of the metabolic network of a given organism is becoming a widespread procedure, there are still many improvements needed in the current methodologies and a clear lack of computational tools for many of the steps. Especially, for less characterized organisms, for which reconstruction will rely more intensely in similarity information, dedicated methodologies are required to speed up the process and increase model accuracy. We therefore propose Merlin for the reconstruction of genome-scale metabolic models of less characterized organisms. Merlin performs similarity searches for any organism that has its genome sequenced, and allows a semi-automatic dynamic (re-)annotation of the genome. The semiautomated (re-)annotation is supported by a specific algorithm developed for scoring of the BLAST hits results. The genome of the yeast Kluyveromyces lactis has been used used as a case study for this method, providing information for the first stage of the reconstruction of this eukaryote. Given the input of 5085 gene sequences, Merlin identified more than 4200 distinct organisms and almost 394.000 genes with sequence similarities to the K. /actis genome. This information, after user appraisal, will be used to assemble a metabolic model with the reactions catalysed by the enzymes encoded in the genome. Such model, in the SBML format, can be used as a first raw approach to the study of the K. lactis metabolism.
|
|---|