New 1.8.14 release?

[abretaud]

Are you using the latest release? That's the point :laughing:

Describe the bug Is there a plan for officially release a new (1.8.14?) version of funannotate? Currently the galaxy tool/workflow is broken due to different bugs that seem to be fixed on the master branch. As we are preparing for a big training session in may, I'd love to make sure everything's working perfectly before that

[nextgenusfs]

Yes I can tag one. I've not had much time to devote to this. I was hoping to test the v1.8.14 (current master) with augustus v3.5 myself but haven't had a chance to do it -- if I could get definitive confirmation that it works well with v3.5 than I can tag that in the conda build which should fix a lot of peoples augustus related install issues (hopefully).

[abretaud]

Hi! I don't know if it helps, but I've just run the tests from the [galaxy tools](https://github.com/galaxyproject/tools-iuc/tree/main/tools/funannotate) using the current code in master branch.

I did it like that:

• created a conda env with funannotate 1.8.13 and pip (it installs augustus 3.5.0)
• installed latest funannotate with pip install git+https://github.com/nextgenusfs/funannotate.git

The tests for funannotate predict pass with latest code (though they no longer with 1.8.13)

With funannotate compare, I had these errors (maybe due to a minimal test dataset?):

[Apr 06 04:55 PM]: Running 1.8.14
[Apr 06 04:55 PM]: Now parsing 4 genomes
[Apr 06 04:55 PM]: working on Genus species
[Apr 06 04:55 PM]: working on Other species
[Apr 06 04:55 PM]: working on Other beast
[Apr 06 04:55 PM]: working on Yet another
[Apr 06 04:55 PM]: No secondary metabolite annotations found
[Apr 06 04:55 PM]: Summarizing PFAM domain results
[Apr 06 04:55 PM]: Summarizing InterProScan results
[Apr 06 04:55 PM]: Loading InterPro descriptions
[Apr 06 04:55 PM]: Summarizing MEROPS protease results
[Apr 06 04:55 PM]: Summarizing CAZyme results
[Apr 06 04:55 PM]: found 0 CAZy familes
[Apr 06 04:55 PM]: No COG annotations found
[Apr 06 04:55 PM]: No SignalP annotations found
[Apr 06 04:55 PM]: Summarizing fungal transcription factors
[Apr 06 04:55 PM]: No transcription factor IPR domains found
[Apr 06 04:55 PM]: Running GO enrichment for each genome
Traceback (most recent call last):
  File "/conda/envs/__funannotate@1.8.13/lib/python3.8/site-packages/funannotate/aux_scripts/enrichment_parallel.py", line 76, in <module>
    lib.runMultiProgress(GO_safe_run, file_list, procs, progress=False)
  File "/conda/envs/__funannotate@1.8.13/lib/python3.8/site-packages/funannotate/library.py", line 1254, in runMultiProgress
    p = multiprocessing.Pool(cpus)
  File "/conda/envs/__funannotate@1.8.13/lib/python3.8/multiprocessing/context.py", line 119, in Pool
    return Pool(processes, initializer, initargs, maxtasksperchild,
  File "/conda/envs/__funannotate@1.8.13/lib/python3.8/multiprocessing/pool.py", line 205, in __init__
    raise ValueError("Number of processes must be at least 1")
ValueError: Number of processes must be at least 1
[Apr 06 04:55 PM]: Running orthologous clustering tool, ProteinOrtho.  This may take awhile...
[Apr 06 04:55 PM]: Calculating dN/dS ratios for each ortholog group, 19 orthologous groups
[Apr 06 04:56 PM]: Compiling all annotations for each genome
[Apr 06 04:56 PM]: Inferring phylogeny using raxml
[Apr 06 04:56 PM]: Found 1 single copy BUSCO orthologs, will use all to infer phylogeny
Traceback (most recent call last):
  File "/conda/envs/__funannotate@1.8.13/bin/funannotate", line 8, in <module>
    sys.exit(main())
  File "/conda/envs/__funannotate@1.8.13/lib/python3.8/site-packages/funannotate/funannotate.py", line 716, in main
    mod.main(arguments)
  File "/conda/envs/__funannotate@1.8.13/lib/python3.8/site-packages/funannotate/compare.py", line 1268, in main
    t = row[3].split(', ')  # convert Eggnog to list
AttributeError: 'float' object has no attribute 'split'

And also found https://github.com/nextgenusfs/funannotate/issues/896

Let me know if I can help by running other tests

[nextgenusfs]

Okay I just pushed fix for #896. Lets try to get rest of these addressed and tag a version.

I can't seem to replicate the errors in compare at least with the integrated tests:

(py3-funannotate) jon@Jons-MacBook-Pro:~$ funannotate test -t compare --cpus 2
#########################################################
Running `funannotate compare` unit testing
CMD: funannotate compare -i Genome_one.gbk Genome_two.gbk Genome_three.gbk -o compare --cpus 2 --ml_model LG+G4 --outgroup botrytis_cinerea.dikarya
#########################################################
-------------------------------------------------------
[Apr 10 07:05 PM]: OS: MacOSX 10.16, 8 cores, ~ 17 GB RAM. Python: 3.7.12
[Apr 10 07:05 PM]: Running 1.8.14
[Apr 10 07:05 PM]: Now parsing 3 genomes
[Apr 10 07:05 PM]: working on Genome one
[Apr 10 07:05 PM]: working on Genome two
[Apr 10 07:05 PM]: working on Genome three
[Apr 10 07:05 PM]: No secondary metabolite annotations found
[Apr 10 07:05 PM]: Summarizing PFAM domain results
[Apr 10 07:06 PM]: Summarizing InterProScan results
[Apr 10 07:06 PM]: Loading InterPro descriptions
[Apr 10 07:06 PM]: Summarizing MEROPS protease results
[Apr 10 07:06 PM]: found 4 MEROPS familes
[Apr 10 07:06 PM]: Summarizing CAZyme results
[Apr 10 07:06 PM]: found 5 CAZy familes
[Apr 10 07:06 PM]: Summarizing COG results
[Apr 10 07:06 PM]: Summarizing secreted protein results
[Apr 10 07:06 PM]: Summarizing fungal transcription factors
[Apr 10 07:06 PM]: Running GO enrichment for each genome
  WARNING: skipping Genome_one.txt as no GO terms
[Apr 10 07:06 PM]: Running orthologous clustering tool, ProteinOrtho.  This may take awhile...
[Apr 10 07:06 PM]: Compiling all annotations for each genome
[Apr 10 07:06 PM]: Inferring phylogeny using iqtree
[Apr 10 07:06 PM]: Found 1 single copy BUSCO orthologs, will use all to infer phylogeny
[Apr 10 07:06 PM]: Compressing results to output file: compare.tar.gz
[Apr 10 07:06 PM]: Funannotate compare completed successfully!
#########################################################
SUCCESS: `funannotate compare` test complete.
#########################################################

[nextgenusfs]

Okay, I downloaded your test files -- they don't have enough annotation which is why it is failing. Looks like perhaps funannotate annotate was not run on these particular test cases? You can use the test cases from the internal tests if that is useful, they are located here: https://osf.io/7s9xh/download?version=1

[nextgenusfs]

The tests for funannotate predict pass with latest code (though they no longer with 1.8.13)

Yes this would be expected I think because Augustus v3.5 is not compatible.

[nextgenusfs]

@abretaud new release is tagged (v1.8.15), let me know if any other problems/issues that need to be addressed. Hope you can get it working on galaxy again.

[abretaud]

Super cool, thanks :tada: I'll let you know how it goes in galaxy

[abretaud]

(it goes well in galaxy, https://github.com/galaxyproject/tools-iuc/pull/4617)