You are here

GxE model for binary variable

3 posts / 0 new
Last post
diana's picture
Offline
Joined: 04/17/2019 - 08:58
GxE model for binary variable

Hello! I am trying to fit a GxE model. The phenotype is a binary variable, and the moderator is measured on a 6-point scale .

I want to use the code posted in there, but this code is suit for continuous variable and I don't know how to change it to be suit for binary variable. Could you tell me what lines I should change or add?

Because I'm pressed for time, I am sorry to directly ask for solution instead of learning by myself.

Many many thanks!

AdminRobK's picture
Offline
Joined: 01/24/2014 - 12:15
very tricky...

What, specifically, are the phenotype and moderator? If you'd rather not say, then are either of the following statements plausible concerning the moderator?:
1. Its covariance with the phenotype is equally due to A, C, and E.
2. All of the variance in the moderator is entirely due to either A, C, or E.

Another question: would you feel more comfortable treating the ordinal moderator as (1) an approximately interval-scale variable, or (2) a categorical variable with unordered categories?

I will say that I know of no OpenMx example script for a binary phenotype and a multicategory ordinal moderator. Indeed, I do not believe I have ever read a paper fitting such a model. In fact, OpenMx may simply not be capable of a completely satisfactory solution to that data-analysis problem....

...Are you familiar with Markov-Chain Monte Carlo Bayesian methods?

tbates's picture
Offline
Joined: 07/31/2009 - 14:25
umx::umxGeE

Have a go with umx::umxGeE

Or perhaps umx::umxGxE_window