Identification of haplotype blocks

Question

[abascalfederico]

Hi all,

I am working with 1000G data for a chromosomal region. I can see that there are different and (more or less) well defined haplotype blocks within that region. I would like to break the region into these "linked" blocks to do a phylogenetic analysis of each block. Can anyone point to a method to identify those haplotype blocks automatically?

Many, many thanks, Federico

Answer 1

[abascalfederico]

I've found the answer. plink --bfile mydata --blocks

[Floris Brenk]

Also take a look at haploview (https://www.broadinstitute.org/haploview/haploview). This uses the same algorithm as plink and is good for seeing the LD structure. btw how are you planning to do the phylogenetic analysis?

[abascalfederico]

Thank you, Floris.

I was planning on using Phyml or RAxML for the phylogenetic analyses, running them for each block separately. However, looking at the alignments I think the blocks are not completely "correct", I mean, there is recombination inside a given block. And recombination mixes phylogenetic histories, so this is a problem.

[jianhaichen]

Did you solve this problem? I met a similar problem. I need to compute phylogeny for each haplotype region of many population samples.