Another bioinformatic tool: PyANI

Overview

Objectives

• To calculate average nucleotide identity of a genome with related genomes.

In the regular lessons, we implemented three bioinformatic tools: blast for homology search, maaft for sequence alignment, and raxml for phylogenetic analysis.

In this section, we will discuss two more tools.

PyANI

PyANI is an open-source python-based tool for calculating Average Nucleotide Identity (ANI) between two or more sequences. When comparing two genomes, first syntenic regions are identified using tools such as mummer or blast. Then the nucleotide identity is calculated in the syntenic regions.

The source code for PyANI is available at widdowquinn/pyani. The documentation for basic usage is available here.

The first stage of PyANI is

PyANI v2 is available in Hipergator, but has to be loaded. The dependencies, mummer and blast+ will be loaded together with pyani.

$ ml pyani

Lmod is automatically replacing "python/3.8" with "pyani/0.2.10".

We will be using the genomes present in files/ani for computing ANI. The file UXhortspp.fasta contains genome of a unknown X. hortorum species. The other sequences are genome of some X. hortorum pathovars downloaded from NCBI.

Getting genome sequences from NCBI

PyANI has a script called genbank_get_genomes_by_taxon.py to download all genomes for a taxon from NCBI. For usage, check the documentation linked above.

The objective now is to perform pairwise comparisons of all reference genomes and calculate ANI. This can be performed with following command.

average_nucleotide_identity.py -i files/ani -o ani -m ANIm -g --gformat png,pdf

• average_nucleotide_identity.py is the name of the script
• -i is used to specify directory containing input genomes/sequences.
• -o is used to specify output directory. Note that the program will exit if this directory preexists.
• -m is used to specify mode for alignment of syntenic region. ANIm specifies mummer and ANIb specifies blast+.
• -g is used to generate graphic output, i.e., heatmap. --gformat specifies the graphic output formats.

$ ls ani

ANIm_alignment_coverage.pdf  ANIm_hadamard.pdf             ANIm_similarity_errors.pdf
ANIm_alignment_coverage.png  ANIm_hadamard.png             ANIm_similarity_errors.png
ANIm_alignment_coverage.tab  ANIm_hadamard.tab             ANIm_similarity_errors.tab
ANIm_alignment_lengths.pdf   ANIm_percentage_identity.pdf  nucmer_output.tar.gz
ANIm_alignment_lengths.png   ANIm_percentage_identity.png
ANIm_alignment_lengths.tab   ANIm_percentage_identity.tab

You can now transfer ANIm_percentage_identity.png to your computer to view the heatmap.

You can also get numeric ANI values from the ANIm_percentage_identity.tab file.

$ awk 'NR==1{print "-"$0; next}{for (i=2; i<=NF; i++) {$i=substr($i,1,5)}; print $0}' ani8/ANIm_percentage_identity.tab | column -t

-             Xpopuli  Xhgardneri  Xhcynarae  Xhhederae  Xhvitians  Xhunknown  Xhcarotae  Xhtaraxaci  Xhpelargonii
Xpopuli       1.0      0.913       0.913      0.912      0.912      0.913      0.913      0.912       0.913
Xhgardneri    0.913    1.0         0.993      0.960      0.982      0.999      0.961      0.974       0.958
Xhcynarae     0.913    0.993       1.0        0.960      0.984      0.993      0.961      0.975       0.958
Xhhederae     0.912    0.960       0.960      1.0        0.960      0.960      0.965      0.955       0.961
Xhvitians     0.912    0.982       0.984      0.960      1.0        0.982      0.961      0.976       0.958
Xhunknown     0.913    0.999       0.993      0.960      0.982      1.0        0.961      0.974       0.958
Xhcarotae     0.913    0.961       0.961      0.965      0.961      0.961      1.0        0.956       0.961
Xhtaraxaci    0.912    0.974       0.975      0.955      0.976      0.974      0.956      1.0         0.954
Xhpelargonii  0.913    0.958       0.958      0.961      0.958      0.958      0.961      0.954       1.0

Based on ANI, the unknown strain seems closest to X. hortorum pv. gardneri.