hi i try to Learn Genetic Interactions from Saccharomyces cerevisiae, using Dynamic Bayesian Networks from BNT toolbox in MatLab.
Collecting data from here :http://genome-www.stanford.edu/yeast_stress/.
"In these experiments, the expression patterns in the Yeast S. cervisiae was explored during the response to various environmental transitions. DNA microarrays were used to measure changes in transcript levels as cells responded to temperature shocks, hydrogen peroxide, the drug menadione, the sulfhudryl-oxidizing agendiamide, the disulfide-reducing agent dithiothretiol, hyper- and hypo-osmotic shock, amino-acid starvation, nitrogen source depletion, and progression into stationary phase."
It is a very good dataset for this reason. The problem is that algorithms (REVEAL, DBmcmc, GlobalMIT) for Bayesian structure learning can handle up to 15 genes for inference. SOo i have to choose up to 15 genes ammong 5000. Any suggestion of how to proceed further besides clustering????
Thanks!
if I understood correctly you can extract strongly differential expressed genes (FDR < 0.01) among 5000 genes as input for GRN inference algorithms. but catnet in R is more flexible for inferring Dynamic Bayesian Networks.
good luck