interpreting Q-Q plot of a GWAS

[S Sam]

Hi everyone,

I hope I can ask this query related to the interpretation of a Q-Q plot.

In the attached plot, obtained from a genome-wide case/control study, the negative log of the p-values are plotted on the x (observed) and y (expected) axis. Few SNPs at the extreme distribution deviates from the expected null distribution. But, since the observed p-values for these SNPs are HIGHER (i.e. less significant) than the expected p-values, how should this be interpreted? as opposed to a situation if the observed p-values were more significant compared to the expected values?

Many thanks

[Christopher Chang]

This may be an artifact of sample size: if you have a sufficiently small number of cases or a small number of controls, it may be impossible for any p-value to be smaller than e.g. 2 * 10^{-6}.

[S Sam]

Thank you very much for your response