smudgeplot extract error: PloidyList: 0A0B is not a valid smudge label

[lilymaya]

Hello,

I have been using the new smudgeplot extract function. It worked great for 4 of my genomes, but for the 5th I got this error message: 'PloidyList: 0A0B is not a valid smudge label'.

It is strange because all the other smudgeplot functions have worked fine for this genome. All genomes are from the same species and I have no reason to suspect that this genome should be particularly different to the others. The coverage is also reasonably high ~66X.

I am wondering if this is a bug in the new function, or if it something inherent in the sample.

Thanks so much for your help!

[KamilSJaron]

Hi, a smudge with A0B0 is a bit funny, not sure how that happened.

You should be able to get through this error if you grep out all A0B0 lines from the smudge annotation (I think it's with .sma suffix) file you give to the extraction command. Basically, the extraction command does not really need any smudgeplot run, it just needs to know what coverage pairs represent each smudge, and get you lists of k-mer pairs with those coverages.

[lilymaya]

That worked! Thanks so much!

Like you said, it's strange that those lines were created. It seems there were only 26 out of ~8700 so I am hoping it is ok to just ignore these.

[KamilSJaron]

If you trust the model (e.g. if the 1n coverage as expected) then it's safe to ignore those.

If you could share the data, I would not mind reproducting the error and try to patch it, but as it stands, it would be hard to recreate it.

[lilymaya]

Thanks, Kamil! Yes, the coverage seems to be what we would expect.

The data is this publicly available dataset: https://www.ncbi.nlm.nih.gov/sra/SRR17230313

[lilymaya]

Just thought I'd add these observations I've had:

• when the genome you use is very high coverage, it will output less extract files (i.e. 1n coverage of ~80 gave extract files for 1A1B, 2A2B, 2A1B, 3A1B, 3A2B, but 1n coverage of ~150 gave only 1A1B, 2A2B, 2A1B)
• for this reason I downsampled some of my genomes because I wanted to be able to access these other output files. When I did this, I found that I got the same extract error as above, with around ~20-30 lines of the .sma file being either 0A0B or 1A0B.