**210**wrote:

Dear all

I have studied Linkage Disequilibrium on genotyped data (SNP chip) from 6 month ago. I read in many article that they did haplotype Reconstruction by several algorithm, such as Parsimony methods: Clark’s algorithm, Likelihood methods: E-M algorithm, Bayesian methods: PHASE algorithm. as I am new in this kind of analysis, i was confusing that the where do the result of haplotype reconstruction was used in LD analysis?

it use in calculation of r2 and D' or in Phase persistence ? or it calculated just for count haplotype frequency in population?

in Phasing process we calculate a correlation or regression between same SNP pair r2 in each distance for two population Because in different populations or breeds a pair of SNP may have the same value of r2 but different LD phases (Goddard et al., 2006). i have 5 different population of cattle and when i calculated the r between two breeds the amount of r was too low . when i check the amount of r2 as random in two population they had considerable different. and now, i dont know why?

is there any one that explain for me where and how we used the haplotype reconstruction result in LD analysis?

if you are going to calculate ld decay with plink you are not need to haplotype reconstruction.

we do haplotype reconstruction before ld analysis. because at the first stage we need to inference our haplotypes (we need to reconstruct or resolve maternal and paternal haplotypes), after doing this, we can calculate ld (r-square). of course i think when we calculate r2 with plink software, we don't need to haplotype reconstruction, because the plink software firstly reconstruct our haplotypes and secondly calculate r2.

50To borrow a link from Pierre: http://whathaveyoutried.com? Open questions like this (the "help me understand this?" kind) are difficult to answer. Is there a discrete question that you need help on?

5.7k