These are chat archives for dereneaton/ipyrad

7th
Jun 2018
Nathan Layman
@northbynate
Jun 07 2018 18:16
Any ideas on why I can't get the full vcf file 'V' as an output? I need the invariant sotes as well.
Nathan Layman
@northbynate
Jun 07 2018 18:24
Getting a
Local variable 'maxsnplen' referenced before assignment error
Eaton Lab
@eaton-lab
Jun 07 2018 19:24
@northbynate I'll take a look at that. Though I've actually been thinking about getting rid of the full vcf option since nobody seems to use it, and storing the information to build it is actually what makes step 6 take so long to run for really large assemblies. We can make things much faster without it. Out of curiosity, how do you plan to use it?
Nathan Layman
@northbynate
Jun 07 2018 19:33
We're writing scripts to calculate per site theta for each contig and want to know how many sites there are in each.
Eaton Lab
@eaton-lab
Jun 07 2018 19:37
If you don't need the read depth information then it might be easier to parse the .loci file, which has the length and highlights variable site columns. The full VCF is usually really really huge, since it stores so much extra information, so unless you're interested in that info it's not ideal. And most (all?) other programs seem to only output VCF for variable sites. But, since we do currently support it I think we should be able to fix it and make it work again for now if you want it.
Nathan Layman
@northbynate
Jun 07 2018 19:43
That's something we've been looking at. The problem is with cleaning. It's nice to
Know read depth info so we can try filtering based on different read depths or to resolve ambiguities.
We're working on a tetraploid so we're having to do everything by hand.
Eaton Lab
@eaton-lab
Jun 07 2018 19:46
I see.
Nathan Layman
@northbynate
Jun 07 2018 19:51
Is there a connection between the locus number in the .loci file and the chrom column in the VCF file?
That sound probably be enough. We don't care about read depth on the invariant sites.
Eaton Lab
@eaton-lab
Jun 07 2018 20:09
For denovo assembly I believe yes. For reference I'll have to check.
Nathan Layman
@northbynate
Jun 07 2018 20:35
We went of a reference.
Nathan Layman
@northbynate
Jun 07 2018 20:44
Went for. Sorry for all the autocorrect bs. Thanks for your help!