CaronLab / MhcVizPipe

A reporting pipeline for visualization of immunopeptidomics MS data.
MIT License
9 stars 6 forks source link

KeyError when running #4

Closed courcelm closed 4 years ago

courcelm commented 4 years ago

I hoped yesterday it was related to missing jquery but it is not.

Traceback (most recent call last): File "/home/mathieu/MhcVizPipe/python/install/lib/python3.7/site-packages/MhcVizPipe/gui.py", line 1022, in runanalysis = analysis.make_report() File "/home/mathieu/MhcVizPipe/python/install/lib/python3.7/site-packages/MhcVizPipe/ReportTemplates/report.py", line 704, in make_report self.sample_overview_table(className='col-6') File "/home/mathieu/MhcVizPipe/python/install/lib/python3.7/site-packages/MhcVizPipe/ReportTemplates/report.py", line 147, in sample_overview_table within_length = self.peptide_numbers[sample.sample_name]['total'] KeyError: 'Test'

courcelm commented 4 years ago

I pasted sequences and clicked load data.

kevinkovalchik commented 4 years ago

That's new... Not sure, I think the changes to the Popen call might have broken NetMHCpan as I parse the results from the stdout.

kevinkovalchik commented 4 years ago

I'll delete the changes on my side and see if it fixes it. So long as the paths to NetMHCpan etc in the config are absolute it shouldn't be necessary to access the environment path.

courcelm commented 4 years ago

I dit the same on my end. I got a new error.

Traceback (most recent call last): File "/home/mathieu/MhcVizPipe/python/install/lib/python3.7/site-packages/MhcVizPipe/gui.py", line 1020, in run_analysis cl_tools.find_best_files() File "/home/mathieu/MhcVizPipe/python/install/lib/python3.7/site-packages/MhcVizPipe/Tools/cl_tools.py", line 208, in find_best_files with open(grouping/'images'/'gibbs.KLDvsClusters.tab', 'r') as f: FileNotFoundError: [Errno 2] No such file or directory: '/tmp/mhcvizpipe/2020-09-15_11:01:22.637595/gibbs/Test/unsupervised/4groups_12066/images/gibbs.KLDvsClusters.tab'

kevinkovalchik commented 4 years ago

Can you share your peptide list? The change is working on my side.

Also, see if there is an html file in /tmp/mhcvizpipe/2020-09-15_11:01:22.637595/gibbs/Test/unsupervised/4groups_12066. If not then something went wrong in gibbscluster.

courcelm commented 4 years ago

It was a single PEPTIDE . Large list is working fine.

courcelm commented 4 years ago

All looks good now!

kevinkovalchik commented 4 years ago

Great! I should add a check on how many peptides are uploaded. I do this before GibbsCluster runs on anything, but not upfront.