Entering edit mode
6.8 years ago
Caragh
▴
40
Hi there,
I have run some genome-wide survival analysis using genipe (which uses a Cox proportional hazards model). I have cleaned for all the normal QC measures (info>0.95, maf>0.05, missingness<0.1, hwe>5e-6, removed related individuals, removed principal component (PC) outliers) and have corrected for the first 8 PCs as well as the site that the samples were collected from. However, I am getting a really large genomic inflation factor of 1.2.
I am wondering - is this due to the nature of the survival model? Or am I missing something that I should be correcting for?
Best wishes,
Caragh
The causes might be several. Maybe you have cryptic relatedness? Or some other unaccounted for bias? And you are right, the model can also be the cause. Can you easily use another model?