veg / bcell-phylo Goto Github PK

Some of the germline sequences are very long, causing the resulting profile alignments to contain thousands of uninformative gaps at the beginning of the sequence. This inflates file size, slowing down loading. For instance, V-30-3 does not contain any non-germline characters until site 7967 (see below), and is 16MB. The largest I've seen thus far 189MB.

A solution could be as simple as running the gap trimmer on the profile aligments.

Probably caused by an off-by-one error; see the bottom of the following image.

The output for rule: unaligned_to_amino_acids, has to be muted for the different rearrangements or else it the rule kicks an error. Need to figure out how to use the wildcards to pick up all the rearrangements for each germline gene

It will be much simpler and more comprehensive to annotate from IgSCUEAL output. The portion of pipeline that uses BLAST should be removed.

Add the ability to toggle amino acid vs. nucleotide alignments.

At the moment, IgSCUEAL was ran to create the initial starting data for the pipeline. This is not good for reproducibility, and explicit invocations will help efforts to productize these tools for Galaxy.

rule gene_unaligned_fasta: need to change the wildcards to loop over the files instead of listing them all out in the rule

the regex in full_fafsta_to_v-gene_separate_fasta.py needs to be modified so that it only grabs germline genes 1-6 and NOT 7

the README on the master branch needs to be updated to reflect our current pipeline

reduce the environment file to only the dependencies that we actually need. it is a waste of time to read in all the dependencies

The data directory can become a bit polluted after running several analyses. It would be nice to have some sort of script to clean it, so that it reflects it's state at the beginning of the pipeline.