Without more details, it's hard to give you suggestions. Maybe this paper that compares different tests applied to spectral counts will help: Zang et al (2006) Detecting differential and correlated protein expression in label-free shotgun proteomics,” Journal of Proteome Research.

There are also now other methods to estimate protein abundance from spectral counts.

Otherwise, the G-test is a likelihood ratio test, it's not available in R but probably exists in some R packages and there's also code to be found on the net (e.g. the module Statistics::Distributions::GTest for perl)