Skip to content

Latest commit

 

History

History
63 lines (42 loc) · 2.34 KB

gubbins_tutorial.md

File metadata and controls

63 lines (42 loc) · 2.34 KB

Generating and analysing alignments with Gubbins

Dataset

This dataset comprises a handful of genomes from S. pneumoniae PMEN3, a multidrug-resistant strain described in detail in D'Aeth et al. It can be downloaded from FigShare. This consists of one complete reference sequence (S. pneumoniae RMV4) and eight draft genomes. These can be extracted with the command:

tar xfz PMEN3_assemblies.tar.gz

Generating the alignment

The draft genomes can be aligned to the reference using SKA2. This is first installed through conda:

conda install -c bioconda ska2

An index file is then generated to name the isolates to be aligned:

for f in 6678*fa; do echo "${f%.fa}\t${f}"; done > PMEN3_isolates.list

The PMEN3_isolates.list should contain this text:

6678_3_10    6678_3_10.fa
6678_3_16    6678_3_16.fa
6678_3_19    6678_3_19.fa
6678_3_4    6678_3_4.fa
6678_3_5    6678_3_5.fa
6678_3_6    6678_3_6.fa
6678_3_7    6678_3_7.fa
6678_3_9    6678_3_9.fa

The alignment is then constructed using the Gubbins script generate_ska_alignment.py:

generate_ska_alignment.py --reference RMV4.fa --input PMEN3_isolates.list --out PMEN3.aln

Analysis with Gubbins

The resulting whole genome alignment can then be analysed using Gubbins:

run_gubbins.py --prefix PMEN3 PMEN3.aln

This should complete in under a minute. The results can be visualised using the website https://jameshadfield.github.io/phandango. Three files can be dragged and dropped onto the loading screen:

  • PMEN3.final_tree.tre

  • PMEN3.recombination_predictions.gff

  • RMV4.gff

The output should resemble this image: PMEN3 output

The tree is on the left, and the refence annotation is across the top. The red bars indicate recombinations reconstructed as occurring on internal branches, which are therefore shared by isolates through common ancestry. The blue bars indicate recombinations reconstructed as occurring on terminal branhces, which are unique to individual isolates. The graph at the bottom summarises the distribution of recombination across the genome, with peaks corresponding to loci encoding antigens (e.g. the cps locus and pspA) and resistance loci (e.g. murM).