Gauge your phage: Benchmarking of bacteriophage identification tools in metagenomic sequencing data

Abstract

Background: As the relevance of bacteriophages in shaping diversity in microbial ecosystems is becoming increasingly clear, the prediction of phage sequences in metagenomic datasets has become a topic of considerable interest, which has led to the development of many novel bioinformatic tools. A comprehensive comparative analysis of these tools has so far not been performed. Methods: We benchmarked ten state-of-the-art phage identification tools. We used artificial contigs generated from complete RefSeq genomes representing phages, plasmids, and chromosomes, and a previously sequenced mock community containing four phage strains to evaluate the precision, recall and F1-scores of the tools. In addition, a set of previously simulated viromes was used to assess diversity bias in each tool's output. Results: DeepVirFinder performed best across the datasets of artificial contigs and the mock community, with the highest F1-scores (0.98 and 0.61 respectively). Generally, machine learning-based tools performed better on the artificial contigs, while reference and machine learning based tool performed comparably on the mock community. Most tools produced a viral genome set that had similar alpha and beta diversity patterns to the original population with the notable exception of Seeker, whose metrics differed significantly from the diversity of the underlying data. Conclusions: This study provides key metrics used to assess performance of phage detection tools, offers a framework for further comparison of additional viral discovery tools, and discusses optimal strategies for using these tools.

Perspective on taxonomic classification of uncultivated viruses

“…Different viral strains or subtypes can be distinguished through viromics, opening up new possibilities to distinguish evolutionary and ecological dynamics of uncultivated viruses in infected hosts and in natural biomes. Recent benchmarking studies based on simulated or mock community data provide information on the advantages and disadvantages of different computational tools for identifying and classifying viruses [24][25][26]. Viruses that are relatively closely related to known ones can be identified by direct sequence-similarity searches of whole genomes or taxon-specific hallmark genes, which are also used for meaningful phylogenies and taxonomic classification.…”

Prediction of prophages and their host ranges in pathogenic and commensal Neisseria species

“…In this study, we used prediction tools that search for viruses. However, because VirSorter2 has been reported to have difficulty distinguishing plasmids from viral sequences (49,52,53), we wanted to address the possibility that predictions from any of the tools may resemble other types of Neisseria MGEs. Specifically, we compared our predictions to known Neisseria plasmids and the Gonococcal Genetic Island (GGI).…”

MetaPhage: an automated pipeline for analyzing, annotating, and classifying bacteriophages in metagenomics sequencing data

“…Several tools exist to perform each step, but while pre-processing, filtering and assembly can rely on gold standard softwares and workflows, viral identification and classification may vary greatly between different tools. Many of the virus mining tools were implemented in the last decade, and differ on the approach used, such as prophage detection, homology, machine learning, random forest or hybrid approaches, which is a combination of some of the aforementioned (Ho et al, 2021). Furthermore, one major drawback of virus metagenomic analysis is the variety of tools used in each step, which needs to be installed separately, may require different data formats and an extended set of dependencies, which may clash in their versions.…”