Hi, I am trying to see the association between a group of over 91 thousand SNPs and a phenotypic trait but when I run the code, all p-values are 1, which seems a bit odd (I have 91728 SNPs and 1000 individuals). I think probably I'm missing something? Maybe the input data is not right or I missed a step? Any suggestion would be welcomed.
Here is what I've done:
Prepare my input files: I have transformed my SNP data (PLINK files) into a Diploid HapMap format using TASSEL and then to a csv file just as the snpdata.csv provided in the example function. My phenotypic data is also in csv format as in the phenodaya.csv format, with the columns ID, BMI and Age (values are all positive integers)
snp <- read.snpData("snps.csv", sep = ",",na.string = "NN")
pdat <- read.csv("phen.csv", header = TRUE, sep = ",")
Run the same command lines as in the example function:
#clustering of the SNPs clust <- qtcatClust(snp) # create a genotype object for the HIT analysis geno <- qtcatGeno(snp, clust) #create a phenotype object for the HIT analysis pheno <- qtcatPheno(names = pdat[, 1], pheno = pdat[, 2], covariates = model.matrix(~ pdat[, 3])) # run the core analysis, the HIT fitted <- qtcatHit(pheno, geno) qtcatQtc(fitted)
Any idea of what might be happening? My files look exactly like the ones in the example, but I could email them to you if that was helpful. Thanks in advance!!!