Terra workflow exiting with "invalid return code 1"

245 views
Skip to first unread message

Jackson Smith

unread,
Aug 3, 2022, 12:42:37 AM8/3/22
to Cumulus Support
Hello, 

I am attempting to use the cumulus workflow on some CITE-seq data that has an ADT component. When I run the workflow, it gets through mkfastq without problems, but when moving into cellranger count I get the following error:

Job cellranger_workflow.generate_count_config:NA:1 exited with return code 1 which has not been declared as a valid return code. See 'continueOnReturnCode' runtime attribute for more details.

The stdout file is empty, but when I check the stderr file, I see 
"Detected multiple feature barcode or target panel files for sample mouse1_adt!"

For context, my sample sheet references two distinct flowcells that samples are pooled across. So, there are two rows for which the Sample field is mouse1_adt, and these rows are identical save for their flowcells. like so:


Lane, Sample, Index, Reference, Flowcell, Chemistry, DataType, FeatureBarcodeFile
mouse1_adt, TTCAGTGG, mm10-2020-A, flowcell1, SC3Pv3, adt, feature_barcode_file
...
mouse1_adt, TTCAGTGG, mm10-2020-A, flowcell2, SC3Pv3, adt, feature_barcode_file

It seems that this error is being thrown because counting ADT reads does not appear to support pooling across flowcells, while counting mRNA reads does so. How do you advice I proceed? Happy to provide any further context needed


Yiming Yang

unread,
Aug 3, 2022, 2:50:53 AM8/3/22
to Jackson Smith, Cumulus Support
Hello Jackson,

Thank you for reporting this issue! We'll add this support to the count step on ADT samples, and will get back to you ASAP.


Sincerely,
Yiming

--
You received this message because you are subscribed to the Google Groups "Cumulus Support" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cumulus-suppo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cumulus-support/bdb86ad4-cc3e-4ce8-976c-be9c50648ea2n%40googlegroups.com.


--
γνθι σεαυτόν.

Yiming Yang

unread,
Aug 3, 2022, 3:09:06 AM8/3/22
to Yiming Yang, Jackson Smith, Cumulus Support
After double checking the error message, it looks like you have some ADT sample in your sample sheet with more than one feature barcode files associated. Namely, like the following example:

Sample,DataType,FeatureBarcodeFile,...
mouse1_adt,adt,fb_file1,...
mouse1_adt,adt,fb_file2,...

where "fb_file1" and "fb_file2" have different filenames.

Can you double check your sample sheet for this potential issue?


Sincerely,
Yiming

Jackson Smith

unread,
Aug 3, 2022, 7:28:36 AM8/3/22
to Cumulus Support
Hi Yiming,

Thank you so much for the quick response! After double checking the sample sheet, all of the adt samples are pointing to the same feature barcode file with the same path in the google bucket. Is there something else I can check too?

Best,
Jackson

Bo Li

unread,
Aug 3, 2022, 12:09:52 PM8/3/22
to Jackson Smith, Cumulus Support
Hi Jackson,

Can you share your workspace with us? We need to look into the data.

Best,
Bo

Bo Li, Ph.D.
Senior Scientist, Data Science and Statistical Computing
Genentech, A Member of the Roche Group



Jackson Smith

unread,
Aug 3, 2022, 1:37:20 PM8/3/22
to Cumulus Support
Hi Bo, 

Thank you and Yiming for your help. I have shared the workspace with bl...@mgh.harvard.edu, please let me know if this is incorrect

Best,
Jackson

Yiming Yang

unread,
Aug 3, 2022, 6:14:30 PM8/3/22
to Jackson Smith, Cumulus Support, Bo Li
Hi Jackson,

After checking your input sample sheet, I see the following issues:

1. You missed "gs://" prefix for all the URLs of "FeatureBarcodeFile" column values.
2. Some ADT samples have the FeatureBarcodeFile missing. For example, "7840_adt" has two rows:

Lane,Sample,Index,Reference,Flowcell,Chemistry,DataType,FeatureBarcodeFile
*,7840_adt,TTCAGTGG,mm10-2020-A,gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/data_new/220607_SL-NVC_0933_AHC2CMDMXY,SC3Pv3,adt,fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new
*,7840_adt,TTCAGTGG,mm10-2020-A,gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/data_new/220607_SL-NVC_0934_BHM3G2DSX3,SC3Pv3,adt,
... ...

However, we require all ADT samples have their corresponding FeatureBarcodeFile set.


Sincerely,
Yiming



--
γνθι σεαυτόν.

Jackson Smith

unread,
Aug 3, 2022, 7:08:10 PM8/3/22
to Cumulus Support
thank you very much! 

Jackson Smith

unread,
Aug 3, 2022, 9:29:00 PM8/3/22
to Cumulus Support
Hi Again, 

thank you for identifying that issue with my sample sheet, that worked perfectly for getting beyond that issue with initiating cellranger count. The workflow runs without issue beyond that. However, an error comes up in all shards when the workflow moves forward into cumulus-adt. 

The error is as follows:

Job cumulus_adt.run_generate_count_matrix_ADTs:NA:1 exited with return code 1 which has not been declared as a valid return code. See 'continueOnReturnCode' runtime attribute for more details.

The stdout file reads

gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_0
gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_1
Load feature barcodes.
Barcode PK exceeds the length limit 21!
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/1532_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_0
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/1532_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_1
generate_count_matrix_ADTs /cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz /cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv 1532_adt_0,1532_adt_1 1532_adt -p 4 --max-mismatch-feature 2 --feature antibody --max-mismatch-cell 0 --umi-length 12

The stderr file reads

Traceback (most recent call last): File "<stdin>", line 46, in <module> File "/usr/lib/python3.9/subprocess.py", line 373, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '['generate_count_matrix_ADTs', '/cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz', '/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv', '1532_adt_0,1532_adt_1', '1532_adt', '-p', '4', '--max-mismatch-feature', '2', '--feature', 'antibody', '--max-mismatch-cell', '0', '--umi-length', '12']' returned non-zero exit status 255.

I notice that exit status 255 is a problem with ssh connection... I assume it must not have to do with gcloud authentication since there was no trouble accessing other files in the google bucket. Is there some other problem? Or maybe it isn't related to ssh after all?

Best,
Jackson

Jackson Smith

unread,
Aug 4, 2022, 10:15:07 AM8/4/22
to Cumulus Support
one other thing to note: this error is not specific to the 1532_adt sample-- the error occurred throughout all shards and applies to every adt sample, I just copy-pasted the above as an example. Do we know what "Barcode PK exceeds the length limit 21!" is referring to?

Best,
Jackson

Bo Li

unread,
Aug 4, 2022, 11:22:27 AM8/4/22
to Jackson Smith, Cumulus Support
Hi Jackson,

Can you let us know if your feature barcode length > 21nt? 

Thanks,
Bo

Bo Li, Ph.D.
Senior Scientist, Data Science and Statistical Computing
Genentech, A Member of the Roche Group


Jackson Smith

unread,
Aug 4, 2022, 12:05:27 PM8/4/22
to Cumulus Support
Hi Bo,

All of our feature barcodes are 15nt long. For instance, 
GATTCCTTTACGAGC barcode corresponds to feature CD34

Bo Li

unread,
Aug 4, 2022, 11:53:13 PM8/4/22
to Jackson Smith, Cumulus Support
Hi Jackson,

I took a look at your workspace.

It seems that your adt_tags.csv file is corrupted. When I opened it using Sublime, I got random characters (see sublime.png). When I opened it using Excel, I got an alert (excel.png). It seems that you renamed an xlsx file with .csv suffix?

Hope it helps,
Bo

Bo Li, Ph.D.
Senior Scientist, Data Science and Statistical Computing
Genentech, A Member of the Roche Group


sublime.png
excel.png

Jackson Smith

unread,
Aug 5, 2022, 9:56:57 AM8/5/22
to Cumulus Support
Hi, 

Thank you for spotting that issue with the corrupted file. I fixed this and reran, but still run into the following issues-- 

Error message:


Job cumulus_adt.run_generate_count_matrix_ADTs:NA:1 exited with return code 1 which has not been declared as a valid return code. See 'continueOnReturnCode' runtime attribute for more details.

from Std Err:

generate_count_matrix_ADTs: barcode_utils.hpp:150: void parse_one_line(const string&, int&, int&, HashType&, std::vector<std::__cxx11::basic_string<char> >&, int, bool): Assertion `barcode_len == index_seq.length()' failed. Traceback (most recent call last): File "<stdin>", line 46, in <module> File "/usr/lib/python3.9/subprocess.py", line 373, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '['generate_count_matrix_ADTs', '/cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz', '/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv', '1532_adt_0,1532_adt_1', '1532_adt', '-p', '4', '--max-mismatch-feature', '2', '--feature', 'antibody', '--max-mismatch-cell', '0', '--umi-length', '12']' died with <Signals.SIGABRT: 6>.

from Std Out:

gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_0 gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_1 strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/1532_adt/ strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_0 strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/1532_adt/ strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/1532_adt_S*_L*_*_001.fastq.gz 1532_adt_1 generate_count_matrix_ADTs /cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz /cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv 1532_adt_0,1532_adt_1 1532_adt -p 4 --max-mismatch-feature 2 --feature antibody --max-mismatch-cell 0 --umi-length 12

At a glance this seems like the same error as before the corrupted file, what do you think?

Jackson

Jackson Smith

unread,
Aug 11, 2022, 3:44:05 PM8/11/22
to Cumulus Support
Hello again, 

I am still getting this issue currently, do you know of any way to resolve / did you have a chance to look into it?

Thank you very much!
Jackson

Bo Li

unread,
Aug 18, 2022, 2:58:06 PM8/18/22
to Jackson Smith, Mia Petljak, Cumulus Support, Yiming Yang
Hi Jackson,

Sorry for my late reply.
Your feature barcode CSV file still has issues:

1) I detected 3 chars with negative values before the first line. 
2) There is no new line after the last line. 

Can you try to run the pipeline using the attached CSV file? It corrected the above two issues.

P.S., thanks to your use case, we have discovered a bug in our pipeline, which we just fixed recently. Please update to the latest Cumulus WDL from Dockstore. Yiming can provide more details.

Best,
Bo

Bo Li, Ph.D.
Senior Scientist, Data Science and Statistical Computing
Genentech, A Member of the Roche Group


adt_tags_2.csv

Yiming Yang

unread,
Aug 18, 2022, 3:10:01 PM8/18/22
to Bo Li, Jackson Smith, Mia Petljak, Cumulus Support, Yiming Yang
Hi Jackson,

As Bo has mentioned, please synchronize your workflow with the "master" version of our Cellranger workflow. You should expect the default value of "cumulus_feature_barcoding_version" input is "0.11.0". After that, please retry with your analysis.


Sincerely,
Yiming



--
γνθι σεαυτόν.

Mia Petljak

unread,
Aug 18, 2022, 5:17:07 PM8/18/22
to Yiming Yang, Bo Li, Jackson Smith, Cumulus Support, Yiming Yang
Thank you Bo and Yiming ! Do we need to rerun the mRNA data too, or is the change in the new workflow only relating to the ADT?

Thanks
Mia 

Yiming Yang

unread,
Aug 18, 2022, 5:31:14 PM8/18/22
to Mia Petljak, Bo Li, Jackson Smith, Cumulus Support, Yiming Yang
It's only related to ADT. So no need to rerun the mRNA samples.


Sincerely,
Yiming
--
γνθι σεαυτόν.

Mia Petljak

unread,
Aug 18, 2022, 5:37:12 PM8/18/22
to Yiming Yang, Bo Li, Jackson Smith, Cumulus Support, Yiming Yang
Got it, thank you ! 

Mia

Jackson Smith

unread,
Aug 19, 2022, 9:24:08 AM8/19/22
to Mia Petljak, Yiming Yang, Bo Li, Cumulus Support, Yiming Yang
Thank you Bo and Yiming for the help. The run looks like it worked in 26/27 shards, very exciting! However, there was an error in shard 16. 

The stdout file says 
gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_0
gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_1
Load feature barcodes.
/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv is parsed. n_barcodes = 11, and barcode_len = 15.
In the index, 0 out of 19151 items are ambigious, percentage = 0.00%.
ntotA = 9, ntotBC = 0.
Error: Detected less than 10 feature barcodes in the first 10000 reads! Maybe you should consider to reverse complement your barcodes?
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_0
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_1
generate_count_matrix_ADTs /cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz /cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv 7839_adt_0,7839_adt_1 7839_adt -p 4 --max-mismatch-feature 2 --feature antibody --max-mismatch-cell 0 --umi-length 12
And the stderr file says 
Traceback (most recent call last):
  File "<stdin>", line 46, in <module>
  File "/usr/lib/python3.9/subprocess.py", line 373, in check_call
    raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['generate_count_matrix_ADTs', '/cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz', '/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv', '7839_adt_0,7839_adt_1', '7839_adt', '-p', '4', '--max-mismatch-feature', '2', '--feature', 'antibody', '--max-mismatch-cell', '0', '--umi-length', '12']' returned non-zero exit status 255.
Have you guys seen this before and was there a resolution? It seems like it may just be related to the structure of our raw data not having very many reads for this one specific shard? 

Best,
Jackson


Bo Li

unread,
Sep 25, 2022, 5:38:47 PM9/25/22
to Jackson Smith, Mia Petljak, Yiming Yang, Cumulus Support, Yiming Yang
Hi Jackson and Mia,

Have you resolved this issue (for the one sample that failed)?

Thanks,
Bo

Bo Li, Ph.D.
Senior Scientist, Data Science and Statistical Computing
Genentech, A Member of the Roche Group


Jackson Smith

unread,
Sep 26, 2022, 7:20:05 AM9/26/22
to Bo Li, Cumulus Support, Mia Petljak, Yiming Yang, Yiming Yang
Hi Bo and Yiming,

I hope you have been well, thank you very much for the follow up. Unfortunately, we have not been able to resolve this issue. As a refresher, it was an issue with only one of the mice with the following error messages:

The stdout file says 
gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_0
gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_1
Load feature barcodes.
/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv is parsed. n_barcodes = 11, and barcode_len = 15.
In the index, 0 out of 19151 items are ambigious, percentage = 0.00%.
ntotA = 9, ntotBC = 0.
Error: Detected less than 10 feature barcodes in the first 10000 reads! Maybe you should consider to reverse complement your barcodes?
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_0
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_1
generate_count_matrix_ADTs /cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz /cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv 7839_adt_0,7839_adt_1 7839_adt -p 4 --max-mismatch-feature 2 --feature antibody --max-mismatch-cell 0 --umi-length 12
And the stderr file says 
Traceback (most recent call last): File "<stdin>", line 46, in <module> File "/usr/lib/python3.9/
subprocess.py", line 373, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '['generate_count_matrix_ADTs', '/cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz', '/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv', '7839_adt_0,7839_adt_1', '7839_adt', '-p', '4', '--max-mismatch-feature', '2', '--feature', 'antibody', '--max-mismatch-cell', '0', '--umi-length', '12']' returned non-zero exit status 255.

Do you have any advice on how best to proceed? Hopefully it’s another easy fix!
Best,
Jackson

Mia Petljak

unread,
Oct 4, 2022, 10:40:53 AM10/4/22
to Jackson Smith, Bo Li, Cumulus Support, Yiming Yang, Yiming Yang
Hi Bo and Yiming, 

A kind/quick reminder about below - we are still having troubles resolving the issue that Jackson described and would be really appreciative of your input on how to take Cumulus forward on this

Thank you and all the best,
Mia

Jackson Smith

unread,
Oct 20, 2022, 12:08:28 PM10/20/22
to Bo Li, Yiming Yang, Mia Petljak, Cumulus Support
Hi Bo and Yiming,

I hope you have been well. I just wanted to send along a kind reminder that we are still struggling with an issue on a recent cumulus run. We are using CITE-seq to produce ADT count matrices for 27 mice, and we were able to get a successful run for 26/27 of the mice. Unfortunately, we have not been able to resolve the issue with the last mouse. 


The most recent run worked successfully on 26/27 shards, but left the following error messages for the remaining shard:

The stdout file says 
gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_0
gsutil -q -o GSUtil:parallel_composite_upload_threshold=150M -m cp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_1
Load feature barcodes.
/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv is parsed. n_barcodes = 11, and barcode_len = 15.
In the index, 0 out of 19151 items are ambigious, percentage = 0.00%.
ntotA = 9, ntotBC = 0.
Error: Detected less than 10 feature barcodes in the first 10000 reads! Maybe you should consider to reverse complement your barcodes?
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0933_AHC2CMDMXY_fastqs/fastq_path/HC2CMDMXY/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_0
strato exists --backend gcp gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt/
strato cp --backend gcp -m gs://fc-113b2ab4-b775-4321-96f6-5f28729d0963/output_new/220607_SL-NVC_0934_BHM3G2DSX3_fastqs/fastq_path/HM3G2DSX3/7839_adt_S*_L*_*_001.fastq.gz 7839_adt_1
generate_count_matrix_ADTs /cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz /cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv 7839_adt_0,7839_adt_1 7839_adt -p 4 --max-mismatch-feature 2 --feature antibody --max-mismatch-cell 0 --umi-length 12
And the stderr file says 
Traceback (most recent call last): File "<stdin>", line 46, in <module> File "/usr/lib/python3.9/
subprocess.py", line 373, in check_call raise CalledProcessError(retcode, cmd) subprocess.CalledProcessError: Command '['generate_count_matrix_ADTs', '/cromwell_root/regev-lab/resources/cellranger/3M-february-2018.txt.gz', '/cromwell_root/fc-113b2ab4-b775-4321-96f6-5f28729d0963/input_new/adt_tags.csv', '7839_adt_0,7839_adt_1', '7839_adt', '-p', '4', '--max-mismatch-feature', '2', '--feature', 'antibody', '--max-mismatch-cell', '0', '--umi-length', '12']' returned non-zero exit status 255.
Do you have any advice on how best to proceed? Thank you again for your help.
Best, 
Jackson

You received this message because you are subscribed to a topic in the Google Groups "Cumulus Support" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cumulus-support/8gt4AP0RIOE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cumulus-suppo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cumulus-support/CAOhT0GAVtLcQh19H5NAdP_XxGjme9TdBdSUjomXuXfnCLE8EYw%40mail.gmail.com.

Yiming Yang

unread,
Oct 24, 2022, 8:40:12 PM10/24/22
to Jackson Smith, Bo Li, Yiming Yang, Mia Petljak, Cumulus Support
Hi Jackson,

The failed sample "7839_adt" has 2 channels (HC2CMDMXY and HM3G2DSX3), and it is the first channel's low quality that caused the "Detected less than 10 feature barcodes in the first 10000 reads" error.

To get rid of it, you can either change the order of the two rows of "7839_adt" sample to enforce processing the latter channel first, or delete the row for the first channel. I've checked both ways locally and they both succeeded.

Moreover, you don't have to rerun everything again. In the attachment is an input sample sheet I created with only "7839_adt" sample's entries included and putting HM3G2DSX3 ahead of HC2CMDMXY. I also use the FASTQ folders generated by your last job's mkfastq for the Flowcell column. Therefore, when rerunning your Cellranger job, you only need to:

1. Use this updated input sample sheet;
2. In workflow page, set "run_mkfastq" to "false".

Otherwise, if you only want to rerun with HM3G2DSX3 channel, just delete the last row in the attached sample sheet.


Sincerely,
Yiming



--
γνθι σεαυτόν.
7839_adt_sample_sheet.csv

Jackson Smith

unread,
Oct 25, 2022, 4:27:18 PM10/25/22
to Yiming Yang, Bo Li, Yiming Yang, Mia Petljak, Cumulus Support
Dear Yiming and Bo,

Thank you so much for your help with this. I greatly appreciate you taking the time to find the fast solution for us, and testing it as well. We will run this now and be out of your hair! Thank you for your patience with all of these sample sheet issues. 

All the best,
Jackson
Reply all
Reply to author
Forward
0 new messages