This is a test version of Biostars. For the public version, visit https://www.biostars.org.
genomic inflation factor calculation

hallo, everyone.

I'm new to GWAS data analysis. When I do QQ plot for my association analysis result produced by snptest software. The observed p-value is deviated from expected p-value. But the inflation factor lambda is quite small, it's only 1.03. How to explain this phenomenon. Followings are my codes. 

chisq = qchisq(gwas$PVAL,1,lower.tail=FALSE);
lambda <- median(chisq) / qchisq(0.5,1)
gwas inflation factor qq plot
• 6,914 views • link
written by lilingjoyo • lilingjoyo

Which one the correct one?

lambda <- median(chisq) / qchisq(0.5,1) 
or
lambda <- median(chisq) / qchisq(0.456,1)
• 0 views • link
written by always_learning always_learning

@always_learning

I think you have gotten confused:

  qchisq(0.5,1)

is (approx) 0.456. You want to use: 

chisq <- qchisq(gwas$PVAL, 1, lower.tail = F)
lambda <- median(chisq) / qchisq(0.5,1)

OP's question cannot be answered without looking at their QQ plot, but I don't see why lambda of 1.03 would be an unusual value to obtain.

• 0 views • link
written by anailis • anailis

0 answers

No answers yet.

Log in to answer this question.

More posts like this