use gamma function to implement wishart

I need to implement Wishart random generator for MCMC derivation of variance. The spark mllib library has a GammaGenerator but not WishartGenerator. When I look at the two distributions carefully, they are almost the same except for a extra 1/2 in the power of the beta parameter. Can I somehow use Gamma to substitute for Wishart? if so, what is the trick in terms of adjusting the shape and scale (or nu and V for Wishart) paramters?