sm_assemblies

Analysis of protein coding exons in single molecule assemblies

dependencies

• Snakemake
• BioPython
• conda
• splign (https://www.ncbi.nlm.nih.gov/sutils/splign/splign.cgi?textpage=downloads)
• Python 3.5
• BLAT
• wget
• samtools

clone this repo

git clone https://github.com/WatsonLab/sm_assemblies.git
cd sm_assemblies

download genomes

/bin/bash scripts/download.sh
gunzip *.fasta.gz
mkdir genomes
mv *.fasta genomes

run

snakemake --use-conda

parse splign output

Perl script alnparse.pl can be used to summarise the splign output.

The way the pipeline stores splign results is in a "one query, one subject" file i.e. "ENST00000052569.10.OCVW01001666.1.aln" - this would be all of the splign hits from ENST00000052569.10 (query) against OCVW01001666.1 (subject).

To summarise the best hit from a single alignment file, run the script like this:

perl scripts/alnparse.pl ENST00000052569.10.OCVW01001666.1.aln

However, often we want to consider the hits against multiple subjects in order to find the best hit. In this case, we run it like this:

perl scripts/alnparse.pl <(cat ENST00000052569.10.*.aln)

This will find the best hit from alignments of ENST00000052569.10 against all subjects it has been aligned against.

The output is tab-delimited:

• query name
• hit name
• query start
• length of alignment
• number of mismatch events
• number of bases in mismatch events
• number of insertion events
• number of bases in insertion events
• number of deletion events
• number of bases in deletion events
• protein sequence of aligned bases