Entering edit mode
4.1 years ago
papriscope
•
0
This is a simple stats question, I'm a newbie and learning and I just want to make sure I'm not missing some step.
If I have summary statistics for various SNPs with betas given, and I wish to 'flip' so to say the reference / effect alleles,
- is it simply flipping the sign of the beta? and then calculating the
OR = exp(beta)
,95%CI = exp(beta+1.96*SE)
orexp(beta-1.96*SE)
like I would with any categorical variable switching correct. - I'm using the z-score/ normal distribution to generate the confidence interval, is that okay? or should I be aware of something?
So for example if the Ref/Effect
A/T beta = -0.23
SE =0.16
T/A beta= 0.23
SE=0.16
OR=1.25 (95% CI 0.91-1.72)
Thank you!