haplotype analysis - help help help help help (May/21/2010 )

hi,

i am using the population based case-control association test for a complex disease (myopia), and i am in the candidate gene approach but not whole-genome.

i have genotyped all the tagged SNPs in the gene ( pairwise tagging by hapmap, r2>0.8, mirror allele frequency >0.2), after i did association test for each SNPs independently,

my boss asked me to do the haplotype analysis too, but i dunt know the meaning of haplotype analysis,

why it is needed? i have done association test for each SNP already.... is this a meaningless test?

i have search for a lot of journals, they said haplotype is a combination of alleles that inherited as a single unit and is being used for geneticist to locate or confer the disease variant???
each tagged SNP represent a small segement of genomic region in a gene????

why somethimes single SNP assocaition test apper negative, but haplotype assocaition test appear positive???

Hi,

SNP Genotyping is an efficient method as you know for complex diseases. But, the real challenge lies in how you use those results.
It is always good to do both single marker analysis and haplotyping as asked by your professor.
In single marker analysis you see the involvement of the (associated allele) genotypes which could be the least possible combination associated with your phenotype.
Where as haplotyping is, for eg. consider, the first locus has alleles A and T with three possible genotypes AA, AT, and TT, the second locus having G and C, again giving three possible genotypes GG, GC, and CC. For a given individual, there are therefore nine possible configurations for the genotypes at these two loci.
For individuals that are homozygous at one or both loci, it is clear what the haplotypes are (in this case single marker is more powerful when you have only one SNP analyzed for a gene); it is only when an individual is heterozygous at both loci that the gametic phase is ambiguous.
Therefore you use some specific programs like PHASE, HAPLOVIEW, HAPLOSTAT etc which could determine the possible haplotypes associated with the marker at that specific loci.
Mostly you can perform this with locus specific, gene specific combinations for a specific phenotype you are looking for.

Foe your case control data, i guess you will have to just use doing an LD plot where you group (they say it as 'block' in LD) the SNPs w.r.t position or gene and then do the haplotype association with any of the softwares mentioned above.

Again see for both the results, that is ur single SNP analysis and haplotyping bcoz for some diseases the SS itself gives more sig results!!

Hope i didn't confuse you
Its all simple if you remove the thread one by one...
Good luck for ur research

Regards,
Celltweety