A novel clustering algorithm based on Bayesian model was introduced into the analysis of large-scale gene expression profiles.

A parametric probabilistic model is sought for the SBM spectral densities, and a Bayesian framework is used to statistically update it to local records.