I have developed a mathematical technique which reveals "global" information about the distribution of causal variants for aquantitative trait. It is similar to the Visscher technique for extracting heritability, in that it does not identify specific loci, but rather properties of their distribution. I have done some preliminary work with this technique on height data, and it seems tobe giving very nice results -- we have indications about the density of alleles as a function of MAF, their average effect size and the total number.

I do not need full access to the SNP level data to do the analysis. At minimum I need a small amount of time from someone who is familiar with computations on such data. I can explain what I need to have computed in a relative brief conversation, and it can run on an ordinary PC or even laptop. Once the resulting set of numbers is reported back to me I can produce the results which characterize gene-level distributions. No information regarding individual genotypes need pass to me.