search

bgruening / galaxytools

:microscope::books: Galaxy Tool wrappers
MIT License
116 stars 227 forks source link

bump flye to 2.9.3 #1372

Closed thanhleviet closed 7 months ago

thanhleviet commented 9 months ago
bgruening commented 9 months ago

@thanhleviet unfortunately it still fails.

thanhleviet commented 9 months ago

@bgruening How can I see the tool_test_output.html ?

There were problems with 5 test(s) - out of 9 test(s) executed. See tool_test_output.html for detailed breakdown.
flye (Test #1): passed
flye (Test #2): passed
flye (Test #3): passed
flye (Test #4): failed
flye (Test #5): failed
flye (Test #6): failed
flye (Test #7): failed
flye (Test #8): failed
flye (Test #9): passed

All of these tests passed on my local VM, using planemo 0.75.19 as well. tool_test_output.html.zip

bgruening commented 9 months ago

Try running planemo test --biocontainers ... The planemo output is also available here on GitHub at the gh actions summery page

thanhleviet commented 9 months ago

Ran with planemo t --biocontainers, it pulled quay.io/biocontainers/flye:2.9.3--py310h2b6aa90_0, and no fails at all tool_test_output.html.zip So no idea why it failed on here 🧐 I couldn't find tool_test_output.html under that action summary page as well.

bgruening commented 9 months ago

On this page https://github.com/bgruening/galaxytools/actions/runs/7299218189?pr=1372 scroll down to All tool test results. Download those artefacts.

bgruening commented 9 months ago 

Output assembly_info:  different than expected
Expected text '418681' in output ('#seq_name    length  cov.    circ.   repeat  mult.   alt_group   graph_path
contig_1    418704  17  N   N   1   *   *,1,*
')
bgruening commented 9 months ago

grafik

thanhleviet commented 9 months ago

lol, flye results are so ridiculously stochastic

bgruening commented 7 months ago

@thanhleviet I think I know whats going on here. Flye just needs way more memory than the CI provides. So this failes here randomly. I pushed an update to master after testing locally.

Thanks @thanhleviet

thanhleviet commented 7 months ago

Thank you @bgruening , but indeed Flye does not perform consistently, see https://rrwick.github.io/2024/02/15/misassemblies.html

Misassembly 2: Listeria welshimeri The second misassembly happened with Listeria welshimeri ATCC-35897 (reads here), where Flye deleted a 3956 bp chunk of the genome. George first noticed it when running hybracter (which uses Flye), and he saw that the deletion occurred about 70% of the time. I tried to replicate the problem on my computer, and I found it to depend on thread count: using 1 thread or 5+ threads resulted in a correct contig (no deletion) while using 2–4 threads resulted in the deletion.6