AssertionError thrown during stitching

[cjw85]

The process ran until

...
[03:05:22 - Stitch] Processing ctg254.
[03:05:29 - Stitch] Processing ctg255.
[03:05:36 - Stitch] Processing ctg256.
[03:05:47 - Stitch] Processing ctg257.
[03:05:59 - Stitch] Processing ctg258.
Traceback (most recent call last):
File "/usr/local/bioinfo/src/Medaka/medaka-1.4.3/venv/bin/medaka", line 11, in
load_entry_point('medaka==0.4.3', 'console_scripts', 'medaka')()
File "/usr/local/bioinfo/src/Medaka/medaka-1.4.3/venv/lib/python3.6/site-packages/medaka-0.4.3-py3.6-linux-x86_64.egg/medaka/medaka.py", line 170, in main
args.func(args)
File "/usr/local/bioinfo/src/Medaka/medaka-1.4.3/venv/lib/python3.6/site-packages/medaka-0.4.3-py3.6-linux-x86_64.egg/medaka/stitch.py", line 81, in stitch
joined = stitch_from_probs(args.inputs, regions=args.regions, model_yml=args.model_yml)
File "/usr/local/bioinfo/src/Medaka/medaka-1.4.3/venv/lib/python3.6/site-packages/medaka-0.4.3-py3.6-linux-x86_64.egg/medaka/stitch.py", line 62, in stitch_from_probs
end_1_ind, start_2_ind = get_sample_overlap(s1, s2)
File "/usr/local/bioinfo/src/Medaka/medaka-1.4.3/venv/lib/python3.6/site-packages/medaka-0.4.3-py3.6-linux-x86_64.egg/medaka/common.py", line 487, in get_sample_overlap
assert len(pos1_ovl) == len(pos2_ovl)
AssertionError

Originally posted by @chklopp in #16 (comment)

[cjw85]

Hi @chklopp

This is not something we have been able to reproduce, and it's not something for which we have been able to identify a possible cause. If you are willing to share your data we might be able to debug it further, we can do this privately if required.

Alternatively you can invoke the medaka stitch program on single contigs (avoiding the ones which are breaking with e.g.:

medaka stitch medaka/consensus_probs.hdf --regions ctg254 ctg255 ctg256 ctg257 ctg259 ...

The consensus_probs.hdf file will have been created by medaka_consensus

[txje]

I have one of these too and I'm happy to share if it will help debugging

[cjw85]

@txje

You should have received an email with details of how to send data to us.

[cjw85]

@txje

I've got your files and can reproduce the error. There is a corner-case in how contigs are broken into chunks for inference for which medaka stitch does not make allowance. There is a simple fix which will appear in the next release tomorrow.

[txje]

Thanks for looking into this!

[cjw85]

This is resolved in v0.5.0, now on pypi.

[Siadjeu]

Hi,

I have the same problem as mentioned above with the newest version of medaka v0.9.2 installed with conda:

[02:22:00 - Stitch] Processing tig00001220:0-.
[02:22:01 - Stitch] Processing tig00001224:0-.
[02:22:02 - Stitch] Processing tig00001233:0-.
Traceback (most recent call last):
File "/user/wupf2892/miniconda2/envs/medaka/bin/medaka", line 11, in
sys.exit(main())
File "/user/wupf2892/miniconda2/envs/medaka/lib/python3.6/site-packages/medaka/medaka.py", line 428, in main
args.func(args)
File "/user/wupf2892/miniconda2/envs/medaka/lib/python3.6/site-packages/medaka/stitch.py", line 116, in stitch
joined = stitch_from_probs(args.inputs, regions=args.regions)
File "/user/wupf2892/miniconda2/envs/medaka/lib/python3.6/site-packages/medaka/stitch.py", line 87, in stitch_from_probs
s1, s2)
File "/user/wupf2892/miniconda2/envs/medaka/lib/python3.6/site-packages/medaka/common.py", line 275, in overlap_indices
assert len(pos1_ovl) == len(pos2_ovl)
AssertionError
Failed to stitch consensus chunks.

I need help, thanks!