Identification of differentially expressed molecular functions associated with breast cancer using Gibbs sampling.

The aim of the present study was to identify differentially expressed molecular functions (DEMFs) for breast cancer using the Gibbs sampling approach. Molecular functions (MFs) were obtained on the basis of the Bayesian Approach for Geneset Selection package. Subsequently, MFs were converted into Markov chains (MCs) prior to calculating their probabilities, utilizing the MC Monte Carlo algorithm. DEMFs were identified with probabilities ≥0.8 and the gene compositions were studied. Finally, a co-expression network was constructed via the empirical Bayes method and a pathway enrichment analysis of genes in DEMFs was performed. A total of 396 MFs were identified and all transformed to MCs. With the threshold, 2 DEMFs (structural molecule activity and protein heterodimerization activity) were obtained. The DEMFs were comprised of 297 genes, 259 of which were mapped to the co-expression network. These 297 genes were identified to be enriched in 10 pathways, and ribosome was the most significant pathway. The results of the present study revealed 2 DEMFs (structural molecule activity and protein heterodimerization activity) which may be associated with the pathological molecular mechanisms underlying breast cancer, based on Gibbs sampling.