mergeBed ERROR: Requested column 10, but database file - only has fields 1 - 9.

[JiahuaQu]

Description of the bug

I ran the chipseq pipeline in HPC and used bowtie2 as the aligner. After MACS2, the mergeBed step had error: mergeBed ERROR: Requested column 10, but database file - only has fields 1 - 9.

Hope get your help. Thank you. my.log test.sh.o9822496.txt

Command used and terminal output

No response

Relevant files

No response

System information

nextflow # version: 22.10.6 HPC SGE: Sun Grid Engine Singularity CentOS nf-core-chipseq-2.0.0

[JoseEspinosa]

Are you updating the narrow_peak parameter using a custom.config with the -c flag. It is advisable not to do so and use the-params-file flag instead as explained here.

[JiahuaQu]

Are you updating the narrow_peak parameter using a custom.config with the -c flag. It is advisable not to do so and use the-params-file flag instead as explained here.

Thank you for the good suggestion. I ran the program with the -params-file flag again, but failed with new errors. Hope get your further help. Thank you. fail_again.txt

[JoseEspinosa]

See this thread in the nf-core Slack for a possible fix. In short, the problem could be when using Singularity, that you probably need to add a custom bind path for Singularity to mount the directory.

[JiahuaQu]

Because the nf-core slack space does not allow new registry, I can't access to it.

Error: No account found. It looks like there isn’t an account on nf-core tied to this email address.

I tried several different emails which can access to other slack space than nf-core sclack.

Do I need your invitation to slack? Hope to proceed with the chipseq pipeline successfully. Thank you.

[JoseEspinosa]

You can join the nf-core slack using this link as explained in the nf-core website (https://nf-co.re/join) :smile:

[JiahuaQu]

You can join the nf-core slack using this link as explained in the nf-core website (https://nf-co.re/join) 😄

I tried that link before and again this time, and couldn't log into slack. (1) I tried "contibue with Google", but got the error: The email address must match one of the domains listed below. Please try another email. (2) There are only two types of emails allowed: @qbic.uni-tuebingen.de, or @seqera.io. (3) On that page, it says, "Don’t have an email address from one of those domains? Contact the workspace administrator at nf-core for an invitation."

Therefore, could you please invite me into the slack? Thank you very much.

[JoseEspinosa]

Did you use this exact link: https://nf-co.re/join/slack ? In principle it should work whatever your email domain 🤔

[JiahuaQu]

Yes, I used that exact link. My email can access to other spaces which invited me by others before, but not this nf-core space.

[JoseEspinosa]

Sorry, it seems that the invitation link has expired we will regenerate it again but in the meantime maybe try with this link

[ewels]

Link updated / working again. Thanks for letting us know 👍🏻

[JiahuaQu]

Thank you very much, Jose and Phil. I got into slack. I will try the binding suggestion in slack and give you any feedback afterwards.

[JoseEspinosa]

OK, thanks a lot for reporting 😄

[JiahuaQu]

A good news: the program ran with bowtie2 succeeded. Then, I switched to try BWA, but this method still failed. I used exactly the same codes and config files except for changing bowtie2 and Bowtie2Index to bwa and BWAIndex. Details in the below three log files.

1-error_sent_to_email.txt 2-my.log.txt 3-nohup_test.sh.o371946.txt

[JoseEspinosa]

I can reproduce the error by providing a directory with no index files, so could you do the following checks to try to see which could be the reason causing the issue. If you could paste here the output of the 3 it would be awesome.

1. Could you try to see if the index is correctly staged in the working dir by doing:

ls /wynton/group/rajkovic/jiahua/test/09-test_NextFlow-nf-core-chipseq/test/BWA/work/a2/47aba59b81052ee234b8e73e115091/

2. Could you try to see whether the command used to retrieve the index is not working correctly working in your system:

cd /wynton/group/rajkovic/jiahua/test/09-test_NextFlow-nf-core-chipseq/test/BWA/work/a2/47aba59b81052ee234b8e73e115091
find -L ./ -name "*.amb" | sed 's/.amb//'

3. Finally, could you check that you are really providing the necessary files in the folder you set to contain the bwa index:

ls /wynton/group/rajkovic/jiahua/software/NextFlow/genome/NCBI/GRCm38/references/Mus_musculus/NCBI/GRCm38/Sequence/BWAIndex

[JiahuaQu]

Thank you for the suggestion. The below are my output:

1. Could you try to see if the index is correctly staged in the working dir by doing:

   ls /wynton/group/rajkovic/jiahua/test/09-test_NextFlow-nf-core-chipseq/test/BWA/work/a2/47aba59b81052ee234b8e73e115091/

   

2. Could you try to see whether the command used to retrieve the index is not working correctly working in your system:

   cd /wynton/group/rajkovic/jiahua/test/09-test_NextFlow-nf-core-chipseq/test/BWA/work/a2/47aba59b81052ee234b8e73e115091
   find -L ./ -name "*.amb" | sed 's/.amb//'

   

3. Finally, could you check that you are really providing the necessary files in the folder you set to contain the bwa index:

   ls /wynton/group/rajkovic/jiahua/software/NextFlow/genome/NCBI/GRCm38/references/Mus_musculus/NCBI/GRCm38/Sequence/BWAIndex

   

I found there are additional two versions of BWAIndex inside the first level of BWAIndex folder, version0.5.x and version0.6.0, so I checked both of them too.

[JoseEspinosa]

OK, I think actually the problem is these/version* folders inside --bwa-index path. We will try to fix it. But in the meanwhile you could provide the indexes files in a different folder and removing the version005.x and the version0.6.0 of it. Please, let us know if this fixes the issue.

[JiahuaQu]

I did as you suggested and ran the pipeline successfully finally. Thank you very much.

[JoseEspinosa]

OK, nice to hear that. I will close this issue and the other one then. Please, do not hesitate in asking us if you find any other issue.