no bigwig file; cannot stat

[thugib]

I am getting an error in the output when I run CutnRunTools. It is: cp: cannot stat '/home/crtools_output/51_S51/peakcalling/macs2.narrow/51_S51.cpm.norm.bw': No such file or directory However, it has already written 5 files and 4 directories to /home/crtools_output/51_S51/peakcalling/macs2.narrow/

It wrote the control_lambda.bdg, peaks.narrowPeak, peaks.xls, summits.bed and treat_pileup.bdg files as well as blacklist_filtered, blacklist_filtered.fa, fimo.result and random1000 directories. Only the .cpm.norm.bw file is missing.

I am running this on Rocky Linux 9.5 .

My command is: ./run_bulkModulemod.sh /scratch/programs/CutnRunTools/CUT-RUNTools-2.0-master/config/bulk-config.json 51_S51

[neuromancerkeag]

ya so I actually get this error as well, and when I go into the "xxx.gene.bw.log" it tells me there's an initial problem with line 246 in bulk-pipeline.sh for file "$path_deeptools/bamCoverage" but that the final error is that colormaps can't be imported from matplotlib. I looked through everything, seems fine, but I found this error someone else had: https://stackoverflow.com/questions/71516506/importerror-cannot-import-name-listedcolormap-from-partially-initialized-modu and it appears that if you already have a file named with a package/function/etc that the calling errors because there's some conflict.

I don't think is my exact issue, as I run cut&runtools from the activated conda environment and I think the pipeline does not like this because the file "bamCoverage" is the exact same syntax as the function "bamCoverage" from deeptools

Currently running through this at the moment, but I will update if the issue resolves

[neuromancerkeag]

so, none of those things worked, instead, I looked through the closed issues and another user 3 years ago resolved most of the meme and .bw bugs here: https://github.com/fl-yu/CUT-RUNTools-2.0/issues/20

[thugib]

Thank you very much for the reply and the link. I have tried installing the macs2, samtools, etc. according to the versions they list on their installation web page, including setting my repodata_fns to repodata.json and using the classic Conda, solver, but this did not work. My "...glibc version and package needs are too modern to reliably run Python 3.6-based environments anymore."

[neuromancerkeag]

okay, so my first recommendation is to not install manually but instead install the .yaml environment files, they should go through fine. I'm operating on Ubuntu 24.04 and I have not encoutered this problem, but when I did try to install manually, the process would always get stuck on deeptools. You also need to specifically install deeptools version 3.5.0 https://github.com/fl-yu/CUT-RUNTools-2.0/issues/56 and doing this after you run the .yaml files. If you're trying to run on a cluster, windows, or Mac, unfortunately I don't know what you can do except talk to your cluster admin, stop using windows, or try to work from a virtual environment on Mac.