This repository contains the code for my masters project par in its initial form, written in literate programming style. For a more recent version I split this in two. The library for indexing the input and for matching can be found at github.com/dadidange/esaMatcher. Currently, I am working on some adaptions at the code to compute the alignments. At some point, this will be published at github.com/dadidange/par.

Right now, this repository here should work best.

To run it from scratch type make. This will overwrite most of the files.

For this to work, some dependencies need to be installed:

• noweb $ sudo apt install noweb
• latex
• go
• libdivsufsort $ sudo apt install libdivsufsort-dev

The working directory should now contain the executable par and the corresponding code in src. The directory ./texfiles contains the document pardoc.pdf for documentation. The documentation is rather a monologue that needs to be reworked.

Running from scratch is not necessary, code and documentation are already provided.
To just compile the code, type go build in the src/par directory. libdivsufsort and go are still required for this.

Par expects FASTA files as input. For example, it can be applied to the 4 covid genomes in ./samples/.

./par samples/*.fa > covid.maf

Results in a multiple alignment in MAF-format.

mview is a great tool to display alignments a little better than plain text. Mview is available here: https://github.com/desmid/mview/ and could be executed like this:

mview -in maf -html head -css on -coloring mismatch covid.maf > covid.html