search

UNM-CARC / QuickBytes

Development of short tutorials for UNM's Center for Advanced Research Computing
18 stars 16 forks source link

Michaelamann metabarcoding #107

Closed michaelamann closed 2 years ago

ethangyllenhaal commented 2 years ago

General comment, for anything other than setting up directories and conda environments, assume they are running it from a script (e.g. no "cd $src"). Assume directories and conda environments are done before scripts, so no need for variables like $src to be set off right away. This will often mean any conda stuff is activated and run on a node.

When downloading data, the folder name after unzipping is "MiSeq_SOP". Also, need to specify to make fastq directory and move fastqs there.mv and change "gzip >fa" to "gzip *fastq".

Qiime: Conda environment year doesn't match comment The denoising stats are named inconsistently, they're output as "stats-dada2.qza" then attempted to be called as "denoising-stats.qza" in metadata tabulate

usearch: Probably add a line to re-zip all the fastqs after merging.

Didn't try Mothur yet on MM's request.

For all, it is worth giving approx runtimes. It's ~5 mins for qiime, ~1 min for usearch. Could be worth just saying "all take less than X minutes", just something to let people know they can queue the test on debug. May also be worth saying how long full datasets could take when you write it up.

gmfricke commented 2 years ago

@ethangyllenhaal please add your review