BNP is a framework that takes in an experimental data file for two groups of samples and builds a gene interaction network for the input genes using supplied information and external biological knowledge. A gene expression data file can be uploaded by indicating the number of control and case samples. The gene expression file should only consist of the genes for which the interaction network will be built.

Current implementation of the BNP provides two improvements to the original publication: (i) fold change values are not discretized and used in continuous form for structure learning reducing information loss (ii) external evidence matrix is enhanced to include over 15 million unique gene interaction pairs for over 30 evidence types.