Hi, I would like to use the latest 1000 genomes project to pick tag SNPs for a genetic association study. Is there an efficient and quick way to do this? Are there any tutorials or software that make the task of picking tagSNPs from the 1000 genomes project easier? Thanks for your time!
To get proxy SNPs, r^2, Dprime etc. You can have a look at the following link from BROAD (1000G is included):
http://www.broadinstitute.org/mpg/snap/ldsearchpw.php
all the best Thomas
ArielSonique asked: How do you decide which of the 0.8 LD SNPs to then pick?
I usually don't because it can be a lot of computational work to decide the putative function of each SNP and to then prioritize all those proxy SNPs to the one (maybe two) that you'll genotype. Search for this topic on BioStar as it has been asked before and many great answers were provided. Try this question, for example.
That said, I'd let the genotyping data I have - say from GWAS - tell me which SNPs already have genotyping data. Or, we would use data from the genotyping assay provider tell us which SNP(s) to genotype because they had the assay for sale (as opposed to designing a new assay for another SNP).
Tagger by the Broad Institute.
http://www.broadinstitute.org/mpg/tagger/server.html
Web-based interface, sends you a link via email once it's done.
Is there a tool that generate tagSNPs for common biallelic CNVs?
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
SNAP is out tool of choice - and we do a lot of tag SNP work - because it contains the phase I 1000G data. For absolute LD, use r^=1.0; for strong LD, use r^2=0.8.
Larry, is there any criteria on physical position, genetic position in ASSAY selection on chromosome, i am working on selecting assay in rice and i have more than 80 SNP assays with their physical and genetic position on chromosome and now i want to reduce this to 10 SNPs per chromosome. i request you to share your knowledge or any other articles useful for me in this work. I am sorry if posted this in wrong.
Thanks very much! I will play around with SNAP. As I understand it, SNAP gives you all the SNPs that are say, in r^2=0.8 with another SNP on a given gene. How do you decide which of the 0.8 LD SNPs to then pick?
That is a good question. For some reason 0.8 is a global accepted threshold for r2 when you tag a specific gene/region. I should add that I work within genetic association of common diseases. It might be different for other diseases.
I wish there were something like Snagger to load the 1000genomes data and pick SNPs. I am not very confident about which SNPs to eliminate in any given LD block all correlated at 0.8. How would you guys do it?