Closed mrvollger closed 3 years ago
My troubles with chain files were due to my own (hopefully relatable) misunderstandings which I will explain in case they help someone else.
From the UCSC site a chain file header line is defined with these names
chain score tName tSize tStrand tStart tEnd qName qSize qStrand qStart qEnd id
I assumed that if the target (t
) was hg18 for example and that the query (q
) was hg17 that this would be called anhg17-to-hg18.chain
chain because that is the common way to describe an alignment… but it seems the naming convention is opposite for liftOver
and alignment files so this situation would actually be named hg18-to-hg17.chain
. So just some confusion about UCSC naming from me.
Still looking into the indexing error I had.
I think I fixed the pyfaidx error in https://github.com/harrispopgen/mutyper/pull/28, but its a little unclear to me since it only crops up at random and I cannot get a case the reproduces it every time.
Let's leave open until the fix is merged
Hi Will,
I am probably doing something very silly, but I am having trouble getting the
ancestor
command to work as I would expect.Looking at the documentation it looks like the chain file should be from the outgroup to the reference (https://harrispopgen.github.io/mutyper/cmd.html#Positional%20Arguments) which in my case gives a chain file with a header that looks like (ref has chromosome names, outgroup has contig names):
and not for example the other direction:
(my source for saying the directionality of these chain files http://genome.ucsc.edu/goldenPath/help/chain.html)
However when I run
mutyper ancestor
with the first chain all the SNP positions in my VCF files get turned into Ns in the ancestral fasta output leading to an empty vcf fromvariants
. whereas if I run with the second file I get results, but it seems opposite what the documentation suggests if I am reading it right (and we thought the results looked weirdish).Here is my calls using
mutyper
and I have deposited an example of all the data I am working with here: https://eichlerlab.gs.washington.edu/help/mvollger/share/mutyper/small_example/
Dir structure to help navigation:
Also a perhaps related issue that I often get errors from
ancestor
when it tries to rebuild an fai even thought it has already been made in a pervious setup step. This becomes an issues when multipleancestor
commands run at the same time I think.Thanks in advance, Mitchell