ReadMe

The datafile:

Note: we found that the dashes and underscores present in the snp annotations provided by illumina caused errors when running our R code. We there for dummy coded these (dashes/underscores) as Q and Z in the analysis file.

The data file contains the following:

A header row that identifies the following variables:

ID, BMI, SEX, AGE, lnTRI, lnPAI1, PAI14G5G + SNP1.........+SNPN

ID = individual ID
BMI = body mass index
SEX = 0 =male       1 =female
AGE = chronological age measured in years
lnTRI = natural log of the triglyceride measurement
lnPAI1 = natural log of the PAI-1 measurement
PAI14G5G = genotype (if available) at the PAI1 4G5G variant 
SNP1-SNPN = all analyzed SNPs with genotypes coded additively for the effect of the minor allele (0,1,2 copies of the minor allele)

Note for R code:
The included R code is for the median regression analysis. However changing the tau variable to 0.75 will yield the R code used to generate the Upper quartile regression analysis