Desenvolvimento de ferramentas para a análise de metagenomas

Abstract

Metagenomics is the study of the total genetic content of an environmental sample, the environment being any from a soil sample to a human biopsy or any source of microorganisms. Metagenomic data can be generated by the whole-genome shotgun (WGS) method, in which all DNA fragments gathered from the sample are sequenced. The process of analyzing shotgun metagenomic data in bioinformatics comprehends many diverse steps, usually consisting of sequence pre-processing, taxonomic classification, functional annotation and assembly of these fragmented sequences - called reads - into contiguous sequences - or contigs. To this end, you must first select a set of tools that can, from the raw data, generate interpretable and insightful results. However, the selection of these tools remains a challenge, especially since it’s an influencing factor in the accuracy, formatting and the type of identifiers present in the final result - which therefore influences the biological discovery itself. In this work, we propose MEDUSA, a modular workflow that comprises a variety of tools, capable of executing read pre-processing, taxonomic classification, sequence alignment against a reference database, read assembly and functional annotation of genes. MEDUSA’s development was realized through a comparison of multiple tools through performance benchmarks and also the development of new tools, those being annotate, that performs functional annotation of alignment hits, and MicroView, which processes taxonomic classification results into an interactive report, containing quality control and biodiversity metrics. Through the selection and development of these tools, we provide to the field of metagenomics a new highly customizable and easy to set up methodology, which equates to or surpasses current alternatives.