Evaluation of robustness of common microbiome taxonomic classifiers. Here we present a set of tools and softwares to perform a systematic comparison of commonly used fungal ITS classifiers and databases with the aim to improve taxonomic classification of environmental sequence reads.
Branch: master
Clone or download
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
CONSTAX updates to supplemental for manuscript revision (#9) Nov 20, 2017
.gitignore
CONSTAX.tar.gz updates to supplemental for manuscript revision (#9) Nov 20, 2017
LICENSE Initial commit Dec 31, 2016
README.md Update README.md Nov 20, 2017
Supplemental_File1_CONSTAX_tutorial.Rmd Update Supplemental_File1_CONSTAX_tutorial.Rmd Nov 20, 2017
Supplemental_File3_otu_processing.sh

README.md

'CONSTAX: a tool for improved taxonomic resolution of environmental fungal ITS sequences'

Table of Contents

Authors

Abstract

One of the most crucial steps in high-throughput sequence-based microbiome studies is the taxonomic assignment of sequences belonging to operational taxonomic units. Without taxonomic classification, functional and biological information of microbial communities cannot be inferred or interpreted. The internal transcribed spacer (ITS) region of the ribosomal DNA is the conventional marker region for fungal community studies. While bioinformatics pipelines that cluster reads into operational taxonomic units (OTUs) have received much attention in the literature, less attention has been given to the taxonomic classification of these sequences, upon which biological inference is dependent. Here we compare how three common fungal OTU taxonomic assignment tools (RDP Classifier, UTAX, and SINTAX) handle ITS fungal sequence data. The classification power, defined as the proportion of assigned operational taxonomic units at a given taxonomic rank, varied among the classifiers. Classifiers were generally consistent (assignment of the same taxonomy to a given operational taxonomic unit) across datasets and ranks; a small number of operational taxonomic units were assigned unique classifications across programs. We developed CONSTAX (CONSensus TAXonomy), a Python tool that compares taxonomic classifications of the three programs and merges them into an improved consensus taxonomy. This tool also produces summary classification outputs that are useful for downstream analyses. Our results demonstrate that independent taxonomy assignment tools classify unique members of the fungal community, and greater classification power is realized by generating consensus taxonomy of available classifiers with CONSTAX.

Study Purpose

We set out to test whether the most commonly used taxonomic classifiers generate similar profiles of the fungal community. To do so we used published MiSeq ITS datasets and compared the standalone command line Ribosomal Database Project Classifier (Wang et al. 2007; Cole et al. 2013), UTAX (Edgar 2010; 2013), and SINTAX (Edgar 2016) Classifiers.

Pipeline

The analysis requires four general steps to go from raw sequences to a consesus taxonomy:

  1. Sequence QC and OTU-picking
  2. Database formatting and training
  3. Taxonomy assignment
  4. Post-taxonomy-assignment processing, filtering, analyzing

Steps 2 through 4 are automated through constax.sh and implemented in the CONSTAX tool.

System Requirements

Supplemental Files

  • Supplementary File 1. CONSTAX.zip - download and unzip to use CONSTAX
  • Supplementary File 2. CONSTAX Tutorial
  • Supplementary File 3. otu_processing.sh - Code for sequence QC and OTU-picking.