pankmer 0.11.7

Generate a PanGenome given a set of genomes

Primary contact: Anthony Aylward, [e-mail]

PanKmer

k-mer based and reference-free pangenome analysis. See the quickstart below, or read the documentation.

Installation

With pip

pip install pankmer

In a conda environment

First create an environment that includes all dependencies:

conda create -c bioconda -c conda-forge -n pankmer python==3.10 \
  biopython==1.79 cython setuptools seaborn urllib3 wheel python-newick \
  pyfaidx gff2bed upsetplot

If running on OSX, a few additional packages will be required:

conda activate pankmer
conda install -c conda-forge clang_osx-64 clangxx_osx-64 gfortran_osx-64

Then install PanKmer with pip:

conda activate pankmer
pip install pankmer

Check installation

Check that the installation was successful by running:

pankmer --version

Tutorial

Download example dataset

The download_example subcommand will download a small example dataset of Chr19 sequences from S. polyrhiza.

pankmer download_example -d .

After running this command the directory PanKmer_example_Sp_Chr19/ will be present in the working directory. It contains FASTA files representing Chr19 from three genomes, and GFF files giving their gene annotations.

ls PanKmer_example_Sp_Chr19/*

PanKmer_example_Sp_Chr19/README.md

PanKmer_example_Sp_Chr19/Sp_Chr19_features:
Sp7498_HiC_Chr19.gff.gz Sp9509_oxford_v3_Chr19.gff3.gz Sp9512_a02_genes_Chr19.gff3.gz

PanKmer_example_Sp_Chr19/Sp_Chr19_genomes:
Sp7498_HiC_Chr19.fasta.gz Sp9509_oxford_v3_Chr19.fasta.gz Sp9512_a02_genome_Chr19.fasta.gz

To get started, navigate to the downloaded directory.

cd PanKmer_example_Sp_Chr19/

Build a k-mer index

The k-mer index is a table tracking presence or absence of k-mers in the set of input genomes. To build an index, use the index subcommand and provide a directory containing the input genomes.

pankmer index -g Sp_Chr19_genomes/ -o Sp_Chr19_index.tar

After completion, the index will be present as a tar file Sp_Chr19_index.tar.

tar -tvf Sp_Chr19_index.tar

Sp_Chr19_index/
Sp_Chr19_index/kmers.b.gz
Sp_Chr19_index/metadata.json
Sp_Chr19_index/scores.b.gz

Note

The input genomes argument proided with the -g flag can be a directory, a tar archive, or a comma-separated list of FASTA files.

If the output argument provided with the -o flag ends with .tar, then the index will be written as a tar archive. Otherwise it will be written as a directory.

Create an adjacency matrix

A useful application of the k-mer index is to generate an adjacency matrix. This is a table of k-mer similarity values for each pair of genomes in the index. We can generate one using the adj-matrix subcommand, which will produce a CSV file containing the matrix.

pankmer adj-matrix -i Sp_Chr19_index.tar -o Sp_Chr19_adj_matrix.csv

Note

The input index argument proided with the -i flag can be tar archive or a directory.

Plot a clustered heatmap

To visualize the adjacency matrix, we can plot a clustered heatmap of the adjacency values. In this case we use the Jaccard similarity metric for pairwise comparisons between genomes:

pankmer clustermap -i Sp_Chr19_adj_matrix.csv \
  -o Sp_Chr19_adj_matrix.svg \
  --metric jaccard \
  --width 6.5 \
  --height 6.5

Generate a gene variability heatmap

Generate a heatmap showing variability of genes across genomes. The following command uses the --n-features option to limit analysis to the first two genes from each input GFF file. The resulting image shows the level of variability observed across genes from each genome.

pankmer reg_heatmap -i Sp_Chr19_index/ \
  -r Sp_Chr19_genomes/Sp7498_HiC_Chr19.fasta.gz Sp_Chr19_genomes/Sp9509_oxford_v3_Chr19.fasta.gz Sp_Chr19_genomes/Sp9512_a02_genome_Chr19.fasta.gz \
  -f Sp_Chr19_features/Sp7498_HiC_Chr19.gff.gz Sp_Chr19_features/Sp9509_oxford_v3_Chr19.gff3.gz Sp_Chr19_features/Sp9512_a02_genes_Chr19.gff3.gz \
  -o Sp_Chr19_gene_var.png \
  --n-features 2 \
  --height 3