context_type="ID" - Error in py_call_imp

[eltonjrv]

Dear developers,

Thanks for this tool! I've successfully run cosmic_fit for both SBS96 and DINUC context types, but am getting the following error when attempting to run it for the "ID" context type

>cosmic_fit(samples="./run05-IDcontext/", output="./run05-IDcontext/output/", input_type='vcf', context_type="ID", collapse_to_SBS96=FALSE, cosmic_version=3.3, exome=FALSE, genome_build="GRCh38") Starting matrix generation for SNVs and DINUCs...Completed! Elapsed time: 23.23 seconds. Starting matrix generation for INDELs...Completed! Elapsed time: 216.67 seconds. Matrices generated for 13 samples with 0 errors. Total of 173331 SNVs, 3965 DINUCs, and 641666 INDELs were successfully analyzed. Error in py_call_impl(callable, call_args$unnamed, call_args$named) : ValueError: cannot reindex on an axis with duplicate labels Run reticulate::py_last_error() for details

reticulate::py_last_error() is not very informative and I can't spot what I should change to have this command working well for "ID" as it did for "96" and "DINUC".

Any light is much appreciated, Thanks, Elton

[eltonjrv]

Dear all, I wonder if anyone was able to look at this issue opened a while ago. Thanks

[marcos-diazg]

Dear @eltonjrv,

Thanks for your interest in our tools, and sorry for the delay. Could you please share your input data, or the corresponding mutational matrix generated as part of the run? Just in order to be able to reproduce your issue on our end.

Thanks in advance. Best,

Marcos

[eltonjrv]

Thanks for getting in touch, Marcos. I've attached the ID signatures' matrices generated during the run and placed within output/ID/ subdir. Is this what you meant by mutational matrix? I'm not sure I can provide you with the full input vcf files, due to data protection issues. Sorry. Looking forward for a light to solve this. Thanks, Elton output_ID_vcf_all.zip

[eltonjrv]

Hi Marcos, Just touching base on whether the matrices I provided you with last week would work for the trials on your end. Let me know if not, so I can generate dummy-labeled vcf files then. Thanks, Elton

[mdbarnesUCSD]

Hi @eltonjrv,

I am looking into this on behalf of Marcos, thanks for sharing the matrices. I suspect this is an issue with an older version of tools that are installed in your environment. I would recommend updating to the following packages with pip: SigProfilerAssignment 0.1.7 SigProfilerMatrixGenerator 1.2.26 sigProfilerPlotting 1.3.23 numpy 1.26.4

We do not support numpy >=2.0.0 yet, so that's why 1.26.4 is being specified. Please let me know if this resolved your issue. Thanks!

[mdbarnesUCSD]

Please reach out if you are still encountering this issue.

[eltonjrv]

Hi all,

The ID signatures detection has indeed worked out after the tools version update. Thanks a lot.

Best, Elton

Em seg., 15 de jul. de 2024 às 19:18, mdbarnesUCSD @.***> escreveu:

Hi @eltonjrv https://github.com/eltonjrv,

I am looking into this on behalf of Marcos, thanks for sharing the matrices. I suspect this is an issue with an older version of tools that are installed in your environment. I would recommend updating to the following packages with pip: SigProfilerAssignment 0.1.7 SigProfilerMatrixGenerator 1.2.26 sigProfilerPlotting 1.3.23 numpy 1.26.4

We do not support numpy >=2.0.0 yet, so that's why 1.26.4 is being specified. Please let me know if this resolved your issue. Thanks!

— Reply to this email directly, view it on GitHub https://github.com/AlexandrovLab/SigProfilerAssignmentR/issues/4#issuecomment-2229111156, or unsubscribe https://github.com/notifications/unsubscribe-auth/AC4CUEDVWEB3MNEDNVBE2DDZMQG7RAVCNFSM6AAAAABJBHLRGGVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMRZGEYTCMJVGY . You are receiving this because you were mentioned.Message ID: @.***>

-- Elton J. R. Vasconcelos DVM, PhD