Otimizações qualitativas e quantitativas nas fases de leitura e análise em pipelines metagenômicos

Abstract

Metagenomic sequencing technologies are advancing rapidly and the size of output data from high-throughput genetic sequencing has increased substantially over the years. Our optimízations and performance evaluations are focused in some of the most critical and time-consuming steps of a metagenomic analysís: pre-processing, taxonomic classification assignment and post-processing of classification results. Optimizations and functions were implemented and introduced in a new architecture, PANGEA+, based on the PANGEA metagenomic pipeline. The main improvements of the present tool are: support of new input file formats and NCBI taxonomy database, new species classification methods, consensus analysis, implementation of distributed memory (MPI) for species classification step, and low complexity optimizations for the post-processing of classification results. The evaluation of the new architecture, shows remarkable improvements in many features and, mainly, in the species classification accuracy and performance.