Fix bugs in miassembler

[Ge94]

Draft PR for numerous bugs to be fixed in miassembler:

• adapted human reference genome name to genome name in codon
• adjusted modules name in modules.config

[jmattock5]

Solving the following bugs:

• pipeline failed when the assembly, after host cleaning, was empty
• ENA rejecting assemblies being uploaded that contained fewer than 2 contigs
• Scientific notation error when the jgi_summarize_contigs output was being used to calculate depth - due to long contigs who's length was for example 2.60658e+06

I have added a filtering step to short_reads_assembly_qc.nf, after the host cleaning, which counts the number of contigs and only generates cleaned_contigs.fa.gz if the threshold is met - I had to remove the extra metadata generated during this quality check for the rest of the pipeline to run. I've set the default as 2 contigs so this fix solves problem 1&2.

I think there was still an issue with which file was being published in PUBLISH_CLEANED_CONTIGS. It was filtered_contigs which I think meant any non-human host cleaning still wasn't being used. Please check that this logic makes sense in my changes!

I've added another if statement to short_reads_assembly_qc.nf so if no reference is specified it still uses the length and human filtered outputs. Again please check this logic.

The qc_failed_runs csv generated now includes samples that fail the assembly qc check I've added. We could split this into another csv if we wanted this to be separate from the read qc output.

[mberacochea]

Thank you, @jmattock5. It looks pretty good to me. I added some nitpicky comments; I've learned that having clear names makes it easier to follow the code, so I've included some suggestions to improve them.

[Ge94]

Good stuff @jmattock5 !

[jmattock5]

Thanks for your comments! I've made all the changes you suggested but for two, please look at my comments for justification

[mberacochea]

Nice work 🏅 @jmattock5