Skip to content

comparative pipeline

This workflow undertakes a comparative anaysis of all the sequences included in the analysis. You can use reference based alignments with snippy or you can use reference free approaches with mash and ska2. 

flowchart LR
sequence --> sequence_assessment --> report
sequence --> speciation --> report
sequence --> variant_detection --> distances --> cluster --> report
variant_detection --> alignment --> tree_generation --> report

Please take note - it will not run any assembly based tools, like MLST or AMR. If these are required use the full pipeline. There are three tools available in bohra that can be used for this purpose 1. snippy (default) - a reference based approach, detecting SNPS compared to a reference genome. This option can ONLY take paired-end fastq files as input and requires the provision of a reference genome. 2. ska2 - this is a reference free approach, inferring SNPS based on splits in kmers. You can use paired-end fastq and/or de novo assemblies as input. 3. mash - this is a quick reference free approach which allows for approximation of genetic distances between sequences. Like ska2 you can provide paired-end fastq and/or de novo assemblies as input.

Further customisation includes selection of the data which is used to generate the tree (for snippy and ska2)

  1. distances - you can generate a distance based tree, where the branch lengths are simply the SNP distances between sequences
  2. alignment - this will generate a maximum liklihood phylogentic tree, using the GTR model of phylogenetic inference.

Additionally - you can select the tree builder to use. VeryFastTree is the default tree builder - as it is very quick. But if required you can also use IQtree.

  1. IQtree
  2. VeryFastTree (default)

For example

snippy (alignment and iqtree) and cluster at a threshold of <= 5 and <= 25 

bohra run comparative -i input_file.tsv -j my_snippy_pipeline -ref <path_to_reference.fa(gbk)> --tree_builder iqtree --cluster_threshold 6,26
ska2 (distance and veryfastree) and cluster at a threshold of <= 5 and <= 25, using complete linkage 
bohra run comparative -i input_file.tsv -j my_ska2_pipeline --comparative_tool ska2 --cluster_method complete --cluster_threshold 6,26