Previous module:
Next module:

Taxonomic analysis of MAGs

Objectives

By the end of this tutorial, students will:

  1. Understand the importance and use of the Genome Taxonomy Database (GTDB) and GTDB-Tk.
  2. Set up and run GTDB-Tk to classify metagenome-assembled genomes (MAGs).
  3. Interpret the results from GTDB-Tk output files.
  4. Explore and utilize the GTDB website for taxonomy research and data exploration.

Introduction to GTDB and GTDB-Tk

The Genome Taxonomy Database (GTDB)

GTDB is a comprehensive resource that provides a standardized, phylogenetically consistent taxonomy for bacterial and archaeal genomes. It is the result of significant efforts to unify and update the classification of microbial genomes, addressing the inconsistencies present in traditional taxonomies. The GTDB has been instrumental in helping researchers better understand the evolutionary relationships among prokaryotes and refine classifications based on genome-based data.

  • Developer and Credits: Phil Hugenholtz and colleagues at the Australian Centre for Ecogenomics deserve immense credit for creating and maintaining the GTDB. Their contributions have been pivotal in reshaping microbial taxonomy.
  • Explore GTDB: Students are encouraged to visit the GTDB website to browse genome data, search classifications, and explore phylogenetic trees. This hands-on exploration will provide insights into how GTDB organizes genomic data.

What is GTDB-Tk?

GTDB-Tk (Genome Taxonomy Database Toolkit) is a software tool designed to classify genomes against the GTDB. It aligns genome data, identifies marker genes, and places genomes within a reference tree, offering reliable taxonomic classifications.

  • How it Works: GTDB-Tk identifies key marker genes in input genomes, aligns these markers, and places the genome into a reference tree to deduce the most probable taxonomic classification.
  • Key Concepts:
    • Marker Genes: GTDB-Tk uses a curated set of marker genes (e.g., bac120 for bacteria, ar53 for archaea) essential for taxonomic placement.
    • Phylogenetic Placement: The tool integrates input data with the GTDB reference to place genomes on the phylogenetic tree accurately.

Credit: Phil Hugenholtz and the GTDB team have significantly impacted microbial taxonomy, providing researchers worldwide with valuable resources and tools.

For the conceptual background on GTDB, watch this video.

Setting Up GTDB-Tk

Step 1: Install GTDB-Tk

Installation Command:

mamba create -n gtdbtk-2.4.0 -c conda-forge -c bioconda gtdbtk=2.4.0

Activate the Environment:

conda activate gtdbtk-2.4.0

Step 2: Access GTDB-Tk Reference Data

GTDB-Tk requires external reference data (~110 GB). For this course, the data has already been downloaded and stored here:

export GTDBTK_DATA_PATH=/home/jovyan/shared-team/gtdbtk_data/release220

Ensure that the GTDBTK_DATA_PATH is correctly set to avoid data path errors.

Running GTDB-Tk

Note: GTDB-Tk is rather brittle, so do not be surprised if you encounter problems. In such case, use Google and ChatGPT to find solutions or workarounds.

Step 1: Gene Calling (Identify)

Run the identify Command:

gtdbtk identify --genome_dir ./. --out_dir identify --extension fasta --cpus 16

Expected Output: You should see progress logs indicating the identification of marker genes, which will look like:

[2024-11-13 08:49:44] INFO: GTDB-Tk v2.1.1
[2024-11-13 08:49:44] INFO: Identifying markers in 3 genomes with 16 threads.
[2024-11-13 08:49:45] TASK: Running Prodigal V2.6.3 to identify genes.

Results: The identified marker genes and intermediate files will be stored in the identify directory:

ls /tmp/gtdbtk/identify/identify/intermediate_results/marker_genes/maxbin_bins.001/

Step 2: Aligning Genomes (Align)

Run the align Command:

gtdbtk align --identify_dir identify --out_dir align --cpus 16

Expected Output: You will see logs related to the alignment of identified markers and concatenation of sequences:

[2024-11-13 09:27:37] INFO: Aligning markers in 3 genomes with 16 CPUs.
[2024-11-13 09:31:20] INFO: Masked bacterial alignment from 41,084 to 5,035 AAs.

Results: Alignment results, including gtdbtk.bac120.msa.fasta.gz, will be saved in the align directory.

Step 3: Classifying Genomes (Classify)

Run the classify Command:

gtdbtk classify --genome_dir ./. --align_dir align --out_dir classify -x fasta --cpus 16 --skip_ani_screen

Expected Output: Logs showing the placement of genomes into the reference tree and classification output:

[2024-11-13 09:40:00] INFO: Classifying genomes based on the aligned data.

Results: The main output files are gtdbtk.bac120.summary.tsv and gtdbtk.ar53.summary.tsv, detailing genome classifications.

Exploring GTDB and GTDB-Tk Results

Hands-On Task:

  1. Run GTDB-Tk over a sample MAG provided in the course data.
  2. Explore the GTDB website to learn more about how microbial taxonomy is structured and how your classified genome fits within it.

Discussion Points:

  • Why is GTDB’s approach to taxonomy more consistent than traditional methods?
  • What do the markers and classifications reveal about your MAG?
  • How does the GTDB tree structure provide insight into evolutionary relationships?
Previous submodule:
Next submodule: