ABSTRACT
Recently, as data usage gets higher importance in almost every field of study, the relationship between two variables is getting a special value. Several ways to detect some specific relationships such as conditional dependence between two variables have been used in the past 10 years. In this study, we investigated a Markov Chain Monte Carlo (MCMC) method, named by Reversible Jump MCMC (RJMCMC), its alternatives and copula methods performance by detail for biological data. Biological data, usually, has lots of variables with only limited samples that cannot be modeled in classical ways. The variables can be proteins or genes and the conditional dependence between variables is the target. We presented the accuracies of the alternative methods to compare them in terms of some measures. Finally, we observed that RJMCMC and vine copula approaches are more accurate than most other discussed methods.
