Dear Dr Sawe

I do not know of a way of running a cross validation within lme4 in R. However there are packages for jack knife and bootstrap testing of model fitting. It is even possible to fit a Bayesian model within R.

Without more details it is hard to know if these are what you are looking for.  

 Here is a page that uses R that might be of use

http://machinelearningmastery.com/how-to-estimate-model-accuracy-in-r-using-the-caret-package/

In agreement with David, I also believe that lme4 does not support cross-validation. However, I used to build my training and validation sets using the following user friendly R package :  

https://cran.r-project.org/web/packages/cvTools/cvTools.pdf

It works very fine. If you need support in coding cross-validation, just send me message, and I may support you (without any needs of coauthorship).

I hope to be helpful,

Dear Dr Sawe

I have just looked at the cvTools package. I confess that I had not heard of it before - there are only so many R packages one can earn - but I think it may be just what you need. 

I have one caveat. There is a function for model selection. While this may be useful it might be wise to use this in conjunction with the AIC, BIC or some similar method of model selection. Other than that this package looks excellent

Thank you Dr Rigal dos Santos 

I normally write a forloop manually to do just that by myself. It is not that complicated.