Brent Pedersen

Hi, yes, duphold should work very well on long read data. a lot of those do seem extreme. How deep is the coverage for this sample? I would look at...

Hi, perhaps you can plot a few of these with [samplot](https://github.com/ryanlayer/samplot)? If you show the result here, I might be able to tell more what's going on.

Yes, something is wrong. Would be nice to see these in IGV or samplot. If you can share one chromosome (bam,vcf) then I can have a look.

thanks, can you also send the reference fasta via that service?

got it. will go through this in next few days.

Hi Edoardo, this is expected. The problem is that simply parsing and then writing the variants becomes a bottleneck. It uses htslib, so it doesn't actually "parse" the sample parts...

yeah, you'll probably have to build yourself with cargo, or wait for me to get this into bioconda. What does: ``` grep -m4 model /proc/cpuinfo ``` show?

hi, you can run duphold without a snp vcf and just find changes in depth in your SVs. your SNP vcf has: ``` ##INFO= ``` which does not follow spec....

shoot. I mean FORMAT, not INFO. what does your VCF have for AD in FORMAT?

oh, I see your AD is: ``` ##FORMAT= ``` should be `Number=R, type=Integer` and describe the depths for each allele.