https://nf-co.re logo
Join Slack
Powered by
I have a pull request in progress for cellranger 1...
# scrnaseq
m
I have a pull request in progress for cellranger 10.1.0 update, one of the 
cellranger multi
tests is failing both locally and in Github Actions with
Copy code
Test Process CELLRANGER_MULTI

  Test [af57fe63] 'cellranger - multi - 10k - PBMC' Assertion failed:
...
     >   2026-10-01 15:01:08 [runtime] (failed)          ID.subsampled_sc5p_v2_hs_PBMC_10k.SC_MULTI_CS.SC_MULTI_CORE.MULTI_GEM_WELL_PROCESSOR.COUNT_GEM_WELL_PROCESSOR._BASIC_SC_RNA_COUNTER._MATRIX_COMPUTER.MAKE_SHARD
      >
      >   [error] Pipestance failed. Error log at:
      >   subsampled_sc5p_v2_hs_PBMC_10k/SC_MULTI_CS/SC_MULTI_CORE/MULTI_GEM_WELL_PROCESSOR/COUNT_GEM_WELL_PROCESSOR/_BASIC_SC_RNA_COUNTER/_MATRIX_COMPUTER/MAKE_SHARD/fork0/chnk0-u00a9be75b4/_errors
      >
      >   Log message:
      >   There were no reads to process. Please check whether the read lengths in the input fastqs satisfy the minumum read length requirements for the chemistry.
github.com/nf-core/modules/…/110431458175?pr=13012
f
I can maybe give it a check tomorrow.
s
Hi! Has there been any progress on this since Felipe’s message? If it’s still unresolved, I’d be interested in helping check the test FASTQ read lengths and chemistry settings to narrow down the failure.
n
I just pushed some changes that are supposed to fix it
s
Thanks, Nico! Could you point me to the commit or PR with the changes? I’d like to understand the fix and see whether there’s any additional testing I can help with.
n
github.com/nf-core/modules/pull/13012/…/d4e8e9b956b9… There are some things that still need attention though
s
Thanks, Nico! My local resources are limited, so I can’t run the full Cell Ranger tests. Is there a small documentation, configuration, or existing-log review task I could help with?
n
Little note, you don't need to send every message to the channel, this just adds noise for the ~800 people in this channel
I will send a more extensive response in the afternoon
👍 1
So basically what I did: • Adjust the read chemistry, because apparently the chemistry was previously wrong, but cellranger ignored this before 10.1, now it filters out reads that don't match the specified chemistry (which is all in this case) and then it fails because there are no reads left. Fixed by specifying the chemistry that matches the read geometry • Memory limit adjustments: this is just a matter of CI infrastructure, prevents an infinite wait • Updated snapshot version captures to 10.1.0 If you want to contribute, feel free to go through the open issues in the #CHN5BV5DW repo, and pick something that you are interested in. Before starting to work on it, you can tell what you picked here, so I can maybe give some guidance
👍 1
😎 1
m
Thank you, @Nico Trummer, I'll follow up on the nf-core/modules pull request!
...and it's already been merged, thank you!
Open in Slack
PreviousNext
Powered by