Comments for UNC Computational Genetics http://compgen.unc.edu/wp A systems genetics research team at UNC-Chapel Hill Mon, 23 Nov 2009 22:56:24 +0000 http://wordpress.org/?v=2.7.1 hourly 1 Comment on NPUTE by adarob http://compgen.unc.edu/wp/?page_id=57&cpage=1#comment-8 adarob Thu, 19 Jun 2008 00:29:54 +0000 http://compgen.unc.edu/?page_id=57#comment-8 Prakash, You simply need to remove all columns before the markers and make sure that the file is formatted as a comma-separated (csv) file. Once the values have been imputed, you can add the ID columns back. -Adam Prakash,

You simply need to remove all columns before the markers and make sure that the file is formatted as a comma-separated (csv) file. Once the values have been imputed, you can add the ID columns back.

-Adam

]]> Comment on NPUTE by vsp_123 http://compgen.unc.edu/wp/?page_id=57&cpage=1#comment-5 vsp_123 Sat, 22 Mar 2008 14:16:12 +0000 http://compgen.unc.edu/?page_id=57#comment-5 Hello, How can a haplotype format be obtained from a regular merlin format genotype data (below), so I can use NPUTE to impute missing data? Ind ID, Family ID, Father ID, Mother ID, Sex, Disease status, marker1, marker 2, ... Thanks, Prakash Hello,

How can a haplotype format be obtained from a regular merlin format genotype data (below), so I can use NPUTE to impute missing data?

Ind ID, Family ID, Father ID, Mother ID, Sex, Disease status, marker1, marker 2, …

Thanks,
Prakash

]]> Comment on SNP reading in FastMap by Daniel Gatti http://compgen.unc.edu/wp/?p=92&cpage=1#comment-3 Daniel Gatti Mon, 27 Aug 2007 13:48:05 +0000 http://compgen.unc.edu/?p=92#comment-3 Is this the SNP Browser tool? I didn' t know that it would make frequency histograms of SDPs. I understand the 2^(n-1) part, but why do you subtract 1 to get 32767? Can you post/send me the "new clean version of the Perlegen data"? I'll try to get your results. I need to get this code right because if the SDPs are wrong, then our association mapping is wrong, too. Is this the SNP Browser tool? I didn’ t know that it would make frequency histograms of SDPs. I understand the 2^(n-1) part, but why do you subtract 1 to get 32767?

Can you post/send me the “new clean version of the Perlegen data”? I’ll try to get your results. I need to get this code right because if the SDPs are wrong, then our association mapping is wrong, too.

]]> Comment on SNP reading in FastMap by Leonard McMillan http://compgen.unc.edu/wp/?p=92&cpage=1#comment-2 Leonard McMillan Sat, 25 Aug 2007 16:33:24 +0000 http://compgen.unc.edu/?p=92#comment-2 <p>Yes, choosing any strain at random (or the first) and using it as the reference "0" in the case of ties should work with regard to getting the correct number of SDPs.</p> <p>Your SDP count for Perlegen is still too high. The maximum number for 16 strains is 2^(16-1)-1 = 32767, and we expect to have considerably fewer. I have written a SNP viewer that already does this sort of analysis. </p> <p>Using the new clean version of the Perlegen data, it gets 20972 for the entire dataset. It also tells you how many SDPs there are per chromosome (there are 5496 on Chromosome 1). The tool also tells you how these numbers are affected by the removal of a strain. For instance, removing the wild-derived strains (CAST, WSB, PWD, and MOLF) results in 2023 SDPs out of 2047 possible.</p> Yes, choosing any strain at random (or the first) and using it as the reference “0″ in the case of ties should work with regard to getting the correct number of SDPs.

Your SDP count for Perlegen is still too high. The maximum number for 16 strains is 2^(16-1)-1 = 32767, and we expect to have considerably fewer. I have written a SNP viewer that already does this sort of analysis.

Using the new clean version of the Perlegen data, it gets 20972 for the entire dataset. It also tells you how many SDPs there are per chromosome (there are 5496 on Chromosome 1). The tool also tells you how these numbers are affected by the removal of a strain. For instance, removing the wild-derived strains (CAST, WSB, PWD, and MOLF) results in 2023 SDPs out of 2047 possible.

]]> Comment on Inferring InDels as well as Missing Genotypes by Leonard McMillan http://compgen.unc.edu/wp/?p=1&cpage=1#comment-1 Leonard McMillan Thu, 02 Aug 2007 18:05:24 +0000 http://compgen.unc.edu/?p=1#comment-1 <p>From a computer science perspective this seems doable. One could imagine incorporating into an HMM the possibility of three unobserved states, one for the minority allele, one for the majority, and one for a deletion.<br /> <br>Another question is if any association mapping tools could take advantage of this information.<br /> <br>Ironically, no more than 10 years ago the primary marker type was short repeat-count variations (microsatellites). These can be described entirely in terms of sequence deletions. Today's marker of choice is SNPs, and most SNP based association tools ignore the possibility of deletions.<br /> </p> From a computer science perspective this seems doable. One could imagine incorporating into an HMM the possibility of three unobserved states, one for the minority allele, one for the majority, and one for a deletion.

Another question is if any association mapping tools could take advantage of this information.

Ironically, no more than 10 years ago the primary marker type was short repeat-count variations (microsatellites). These can be described entirely in terms of sequence deletions. Today’s marker of choice is SNPs, and most SNP based association tools ignore the possibility of deletions.

]]>