HoSeIn: A Workflow for Integrating Various Homology Search Results from Metagenomic and Metatranscriptomic Sequence Datasets

Abstract

Data generated by metagenomic and metatranscriptomic experiments is both enormous and inherently noisy. When using taxonomy-dependent alignment-based methods to classify and label reads, the first step consists in performing homology searches against sequence databases. To obtain the most information from the samples, nucleotide sequences are usually compared to various databases (nucleotide and protein) using local sequence aligners such as BLASTN and BLASTX. Nevertheless, the analysis and integration of these results can be problematic because the outputs from these searches usually show inconsistencies, which can be notorious when working with RNA-seq. Moreover, and to the best of our knowledge, existing tools do not criss-cross and integrate information from the different homology searches, but provide the results of each analysis separately. We developed the HoSeIn workflow to intersect the information from these homology searches, and then determine the taxonomic and functional profile of the sample using this integrated information. The workflow is based on the assumption that the sequences that correspond to a certain taxon are composed of:1)sequences that were assigned to the same taxon by both homology searches;2)sequences that were assigned to that taxon by one of the homology searches but returned no hits in the other one.