ibd

Interested in a couple approaches to this:

IBDseq is one option. However, I think it only identifies regions of shared identity in pairwise samples. Genotypes should be diploid unphased (X/X). I downloaded the jar file to my lab computer desktop and a working commandline is here:

java -Xmx4000m -jar ibdseq.r1206.jar nthreads=2 gt=lumpy.pass.diploid.vcf out=res chrom=Pv_Sal1_chr06

MCMC_IBDfinder is potentially more powerful. Solves the problem of inferring IBD in multiple samples using mcmc. I have it compiled on Kure (it was easy to do), but it requires special input formats, genetic distances, and no LD in dataset...

ChromosomePainter is perhaps another option, paper here. Seems to have multiple components like fineSTRUCTURE to it. This is similar to what Miotto 2013 used I think. It is now in my PATH on juliano lab computer. Need to run it chromosome-by-chromosome. Can convert a VCF file to the PHASE format required for this using vcf2phase.r.

fs example.cp -idfile data.ids -phasefiles chr06.txt -ploidy 1 -go

This seems to produce some expected output. Just need a way of looking at it now. Maybe fineStructure GUI or the provided R library?