I conducted a GWAS study, and found a strongly associated SNP.
Then, I downloaded hapmap3 genotypes as well as expression data generated from hapmap3 genomes.
I looked at gene expression of nearby genes based on genotype of the associated SNP, and I found that genotype of this SNP is associated with expression of a gene near to the marker.
Now, this SNP is on the lower end of minor allele frequency. As such, although there were some homozygotes for the minor allele in our genotyping study, it is not entirely surprising that there were not any in the hapmap3 expression data.
Thus, when I conducted the association test between genotype and gene expression, only AA and AC (no CC) genotypes were available for testing. Our alpha threshold was set at alpha = 0.05 because this was the only variant we tested, and the association was stronger (p = 0.003). However, I am not certain if it is appropriate to report an eQTL as significant if only 2 of the 3 possible genotypes are present.
So, my questions are:
1) Is it valid to report this? 2) If not, what technical or other considerations does one need to be aware of that make it unreliable?
Not what you are asking for but useful enough to share: GTEx has precalculated eQTL SNPs per gene, in multiple tissues, in a pretty big database. That's likely a powerful and reliable source.
Have you resolved the question? Due to my small samples, there are often only (AA and AC) or (AA and CC) in genotypes. And, i don't know how to do with it... could you give me any suggestion?