initial df scripts

[erinyoung]

@jwarnn , I mentioned last week that I was going to try and set something up for Donut Falls in a similar way to Grandeur.

THIS IS HOW FAR I'VE GOTTEN. I know that it currently doesn't work.

It's also as far as I'm going to go for now. You are more than welcome to tackle this challenge. I think I may give this some additional attention in October-ish. Maybe.

monitor_ngs.py is meant to look for new GridIon runs and copy over the files to /Volumes/BioNGS_1. Some of these will be runs for Donut Falls, some are for @poojasgupta 's projects (esp. with TaxTriage). Python's tree function, however, either takes forever or freezes, so I don't really recommend this script.

This is what I've been doing to copy files over once I know a run is done.

run=UT-P2S01293-240711
rsync -rvh  /Volumes/NGS/Output/GridION/${run} /Volumes/BioNGS_1/ --exclude *fail*

I'm sticking with rsync for now.

A nanopore sequencing run is complete (generally) when the sequencing_summary_*_*.txt is created. Generated fastq files are not named after any user supplied id. Instead, they are divided by barcode. Each barcode should have multiple fastq files that have to be combined together for further analysis/processing into one per sample.

I like to name these combined files into 'sample.fastq.gz', where sample is the lims id.

Then we get to the hard part: For Donut Falls, we have to find the corresponding Illumina files for each sample - which are in a variety of locations.

This is how I've been doing it in bash: https://github.com/UPHL-BioNGS/Donut_Falls/blob/main/bin/uphl_sample_sheet.sh , and my initial attempts at doing this in python (not working and incomplete) are found in aws_samplesheet_df_create.py.