Update isoseq modules and migrate to nf-test

[sguizard]

Modules migrated to nf-test:

• isoseq/cluster
• isoseq/refine
• pbccs
• pbbam/pbmerge
• lima

Software update:

• isoseq/cluster: v.3.8.1 -> 4.0.0
• isoseq/refine: v.3.8.1 -> 4.0.0
• pbbam/pbmerge: 2.1.0 -> 2.4.0
• lima: 2.7.1 -> 2.9.0

[sguizard]

The conda pbbam/pbmerge is failing for not matching snapshots, but singularity and docker ones succeed. Should it be ignored?

[famosab]

Looks good :) I left some comments. For the next time: I think its best practise to do one module per PR so it stays easier for the reviewers.

[famosab]

The conda pbbam/pbmerge is failing for not matching snapshots, but singularity and docker ones succeed. Should it be ignored?

I would go for comparing the file contents rather than the checksums in that case but not ignoring conda alltogether. So go for something like

# checking output vcf files
{ assert path(process.out.candidate_small_indels_vcf.get(0).get(1)).linesGzip.contains("##fileformat=VCFv4.1") }

# checking output files with same headers
{ assert snapshot(file(process.out.stats.get(0).get(1)).readLines()[0..5]).match() }

[SPPearce]

@sguizard , are you able to finish this PR off and get these changes merged in? Please ask if you need help.
Conda often gives md5sums for bam files that are different to docker/singularity, so you need to check for existence of the file rather than the md5sum.

[famosab]

From the anaconda registry it seems to me that there was a version jump from 2.1 to 2.4 so 2.3 and 2.5 do not exist there. And for pbcopper there is no version 2.4 on anaconda. Or am I missing something?

[famosab]

Second comment: It seems to me that pbbam was archived and the individual tools were transferred to pbtk. I think this should be reflected in the nf-core modules?

I would suggest to remove all changes made to pbbam/pbmerge from this PR and tackle the proper update on a separate branch with a separate PR. Because all the other modules have been ported properly and then it would be less crowded. What do you think @SPPearce @sguizard?

[SPPearce]

Second comment: It seems to me that pbbam was archived and the individual tools were transferred to pbtk. I think this should be reflected in the nf-core modules?

I would suggest to remove all changes made to pbbam/pbmerge from this PR and tackle the proper update on a separate branch with a separate PR. Because all the other modules have been ported properly and then it would be less crowded. What do you think @SPPearce @sguizard?

That sounds sensible, let's do that then.

[famosab]

Remove the ppbam related files from this commit and follow up on #5601

[sguizard]

I removed pbbam/pbmerge and all tests are green 🎉
Thanks for the support.

[famosab]

@SPPearce Do you think we should remove the module completely here (as indicated in the issue #5601)? Then I will retract my comments :D

Edit: I asked in the #modules slack channel

[SPPearce]

As discussed on the slack channel, I would suggest that we leave the pbmerge module untouched in this PR, and open a separate PR to deprecate it (rather than delete it).

[famosab]

LGTM

You just need to run prettier on the pytest yml I think

[SPPearce]

Good work @sguizard 🎉

[sguizard]

Thanks for the help!