Add umicollapse as an alternative to umi-tools #1369
Conversation
TODO: - [ ] Benchmark paired ends mode - [ ] Include in multiqc
There was a problem hiding this comment.
Excellent work, clean and well-structured! I think, you found very smart solutions to not duplicate too much of the code.
Of course, we should wait for the lead developers of the pipeline to review as well prior to merging, but in general you have my support for this addition! Ideally, your fix for umicollapse would also be included in the form of an updated tool version.
Please don't forget the Changelog later and also to add your name to the contributors (well deserved)!
| @@ -30,6 +30,7 @@ params { | |||
| with_umi = false | |||
| skip_umi_extract = false | |||
| umitools_extract_method = 'string' | |||
| umi_dedup_tool = 'umicollapse' | |||
| umitools_grouping_method = 'directional' | |||
There was a problem hiding this comment.
Reading that makes we wonder if we should be conservative regarding the parameter names or make them more generic, e.g. renaming umitools_grouping_method to something like umi_grouping_method / umi_dedup_grouping_method etc.
There was a problem hiding this comment.
Yes, I thought about that too. Depends on your policy for introducing breaking changes between versions. If the authors are OK, I can rename these parameters to be more generic.
| @@ -30,6 +30,7 @@ params { | |||
| with_umi = false | |||
| skip_umi_extract = false | |||
| umitools_extract_method = 'string' | |||
| umi_dedup_tool = 'umicollapse' | |||
| umitools_grouping_method = 'directional' | |||
There was a problem hiding this comment.
Yes, I thought about that too. Depends on your policy for introducing breaking changes between versions. If the authors are OK, I can rename these parameters to be more generic.
Signed-off-by: Siddhartha Bagaria <[email protected]>
|
I checked the output of UMICollapse, and it definitely does not follow the assumptions stated in prepare_for_rsem.py. So |
|
@MatthiasZepper all remaining tasks are now done. Please take another look. |
Unfortunately, the UMICollapse module has not yet been upgraded... |
|
@siddharthab: The module is finally upgraded, so you can replace the local path with the up-to-date module. |
Thanks. Can you check if everything looks good now? I just ran |
Co-authored-by: Jonathan Manning <[email protected]>
siddharthab
force-pushed
the
umicollapse
branch
from
f4d6b0d
to
05513ee
Compare
@pinin4fjords, done. |
|
@nf-core-bot fix linting |
|
Sorry, good in principle now for me (maybe some factoring out we can do later). Just need to get tests to run + pass. |
Well, it appears the issue is on my end. Will fix asap! Sorry! The UMICollapse tests are failing, because they try importing the BWA-Mem module, which isn’t included in this pipeline. As obviously everything is available in the modules directory, the tests pass in that context. Considering that I had copied the tests from Turns out that I had coicidentially copied them as a template during a very brief period on March 4 before Maxime updated them again, which explains the discrepancy... Long story short, I will have to fix the UMICollapse test in the modules directory to be self-reliant. |
Checklist:
PREPARE_FOR_SALMONis needed