dsl2: metagenomics uncollapsed paired end #1098
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Warning Newer version of the nf-core template is available. Your pipeline is using an old version of the nf-core template: 3.1.2. For more documentation on how to update your pipeline, please see the nf-core documentation and Synchronisation documentation. |
|
dsl-2 metagenomics-pairedend
ilight1542
commented
Feb 21, 2025
|
Ready for review: Basic Overview: We can now retain PE information from unmerged input fastq files (when PE samples going into metagenomics will be separately input into metaphlan, kraken2, and krakenuniq. Malt does not accept PE reads so it is not allowed with the param combo. Also fixed a small error in the parsing of metagenomics input data from bamfiltering --> metagnomics. |
8 tasks
merszym
requested changes
Feb 28, 2025
subworkflows/local/bamfiltering.nf
Outdated
| } | ||
|
|
||
| // Routing for metagenomic screening -> first accounting for paired-end mapping, then merged mapping, then no metagenomics | ||
| if ( ( params.run_metagenomics && params.metagenomics_input == 'unmapped' ) && params.preprocessing_skippairmerging ) { | ||
| ch_fastq_for_metagenomics = CAT_FASTQ_UNMAPPED.out.reads | ||
| ch_fastq_for_metagenomics = CAT_FASTQ_UNMAPPED.out.reads.mix(ch_paired_fastq_for_cat) | ||
| } else if ( ( params.run_metagenomics && ( params.metagenomics_input == 'mapped' || params.metagenomics_input == 'all' ) ) && params.preprocessing_skippairmerging ) { |
There was a problem hiding this comment.
if input now is 'all', the unmapped reads are not taken into account.
I think that needs one more if-condition, with something like
if metagenomics_input == 'all':
CAT_FASTQ_MAPPED.out.reads.mix(CAT_FASTQ_UNMAPPED.out.reads).mix(ch_paired_fastq_for_cat)
but I might miss something here?
There was a problem hiding this comment.
It looks like the unmapped reads are not filtered out? then it makes sense...
There was a problem hiding this comment.
the splitting of all vs only mapped reads is dealt with at the modules.config level.
withName: SAMTOOLS_FASTQ_MAPPED { tag = { "${meta.reference}|${meta.sample_id}_${meta.library_id}" } ext.args = [ params.metagenomics_input == 'all' ? '' : '-F 4',
We could split the module/import into an additional SAMTOOLS_FASTQ_ALL, for clarity for anyone looking into this section of the pipeline... @jfy133 what are your thoughts?
There was a problem hiding this comment.
I adjusted this whole section to imporve clarity. please see latest commit.
essentially the splitting of the fastq and cat into unmapped and mapped versions is unnecessary. all parameters can be parsed fine within modules config.
i confirmed that the md5sum of the final files sent to the ch_fastq_for_metagenomics channel are consistent across all the parameter combos that we were there.
|
|
||
| So: needs to be fixed up higher (eg in bamfiltering.nf, likely with a new adjustment to the SAMTOOLS_FASTQ_UNMAPPED, SAMTOOLS_FASTQ_MAPPED, and SAMTOOLS_VIEW_BAM_FILTERING modules ) | ||
|
|
||
| ISSUE FOUND: while the outputting of PE reads is OK in bamfiltering.nf (fastq_mapped & fastq_unmapped) when overlap merging is not done cat_fastq weirdly merges singletons to one PE file and other to the other PE file, so then everything gets fucked up |
There was a problem hiding this comment.
That sounds like a hackathon-thing to do
There was a problem hiding this comment.
let me double check this behavior and add more context of what we might need to do in a hackathon... i this was more for my own notes reference but was not edited prior to opening the PR.
…tering --> fastq for metagenomics. improves clarity (IMO)
ilight1542
commented
Mar 7, 2025
| ch_fastq_for_metagenomics = SAMTOOLS_FASTQ_MAPPED.out.other | ||
| } else if ( !params.run_metagenomics ) { | ||
| // No length/quality filtering applies to metagenomic bam files (could be extension) | ||
| // All bam -> fastq filtering options (-F 4, -f 4 or none) will be dealt with within modules.config |
There was a problem hiding this comment.
@TCLamnidis please take a quick look -- i restructured everything into a single call, since the modules.config is easily robust enough to deal with the splitting. and it allows for much cleaner code in this section which should help going forward
There was a problem hiding this comment.
or @jfy133 who i think wrote the initial implementation...
There was a problem hiding this comment.
I think that makes it much easier! Nice one :)
merszym
approved these changes
Mar 14, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
PR for metagenomics paired-end merging off, retaining that info for metagenomic tools that have PE modes.
This is a QOL improvement for tools that can use paired end information (kraken2, krakenuniq), and now allows for variable inputs into these tools (by splitting inputs into either PE or SE data and running separate instances).
Main issues resolved:
--preprocessing_skippairmergingused.PR checklist
scrape_software_versions.pynf-core lint .).nextflow run . -profile test,docker).docs/usage.mdis updated.docs/output.mdis updated.CHANGELOG.mdis updated.README.mdis updated (including new tool citations and authors/contributors).