search

B-UMMI / chewBBACA

BSR-Based Allele Calling Algorithm
GNU General Public License v3.0
134 stars 27 forks source link

chewbbaca_AlleleCallEvaluator bug #204

Open mcanses opened 3 months ago

mcanses commented 3 months ago

Hello, I am using the chewbbaca_AlleleCallEvaluator on galaxy europe and locally on the machine. When I use this tool on the allele call output, the "Core-genome Neighbor-Joining Tree" plot is missing in the result and two identical files are available for the allecall web report, while only one would be required. Could you please resolve the issue?

Thank you

rfm-targa commented 3 months ago

Hello @mcanses,

Thank you for your interest. The core-genome tree might not be displayed if the cgMLST includes 0 loci or if an internal problem leads to errors. The stdout might include relevant information about possible issues. Can you share the stdout and the command you used to run the module? Do you have multiple HTML files for the report files? Can you open the report and visualize all the components except the core-genome tree? It would also help us identify the issue if you could share the report files (e.g., a Zip with all the files shared through email), but we understand if it's impossible.

Kind regards,

Rafael

mcanses commented 3 months ago

Hello, Thank you for your answer. I am using chewbbaca (Version 3.3.3) via galaxy https://usegalaxy.eu/ The galaxy support told me that this is not a problem of galaxy, but there is a bug in the tool itself and not in galaxy. I can open the report and visualize all the components except the core-genome tree which is not there. Find in attachment a zip folder containing the outputs of the chewbbaca_AlleleCallEvaluator and the command and stdout Thank you

Best regards

Marina

rfm-targa commented 2 months ago

Hello @mcanses,

Sorry for the delay. I ran some tests to verify if the issue was related to the file names of the input files or the loci files, but that did not reveal any issue. I think the best approach is to run the same analysis with chewBBACA's latest version, v3.3.10. Have you tried that, or can you install the latest version locally to test it? The latest version includes several bug fixes that might target the issue you are encountering. If not, we would need some input data we could use to reproduce the issue.

Kind regards,

Rafael

mcanses commented 2 months ago

Hello, I did not try the version v3.3.10, and i can not install it locally, since I am using the tool in Galaxy, where the version is 3.3.30. I can send you the data used as input for the allele call evaluator and the cgMLST scheme. What else do you need? can I send it by email to you? THanks Best regards

Marina

rfm-targa commented 2 months ago

Hello @mcanses,

The input data for the AlleleCallEvaluator module and the schema should be enough to identify the issue. You can send the data to imm-bioinfo@medicina.ulisboa.pt (it might exceed the size limit even if it's compressed; a link to download the data from WeTransfer or a similar service is fine). Thank you for helping us track this issue.

Best regards,

Rafael

indexofire commented 1 month ago

I have a closed issue when running on shell, the locus alignment file return nothing so that AlleleCallEvaluator returned wrong "Core-genome Neighbor-Joining Tree". By checking the codes I found out the reason was mafft runing uncorrectly in my conda env. After make sure mafft can poweron, everything goes well. Hope it helped.