Skip to main content
Fig. 1 | BMC Bioinformatics

Fig. 1

From: Using empirical biological knowledge to infer regulatory networks from multi-omics data

Fig. 1

IntOMICS framework. IntOMICS framework takes as input (i) gene expression matrix GE with \(m\) samples and \(n_1\) genes, (ii) the associated copy number variation matrix \(CNV\) (\(m\) x \(n_2\)), (iii) the associated DNA methylation matrix of beta-values \(METH\) (\(m\) x \(n_3\)) sampled from the same individuals, and (iv) the biological prior knowledge matrix \(B\) (\(n_1\) x \(n_1\)) with information on known interactions among molecular features. An automatically tuned MCMC algorithm [30] estimates parameters and empirical biological knowledge. Conventional MCMC algorithm with additional Markov blanket resampling step is used to infer resulting regulatory network structure consisting of three types of nodes: GE nodes (highlighted in green) refer to gene expression levels, CNV nodes (highlighted in blue) refer to copy number variations, and METH nodes (highlighted in red) refer to DNA methylation. Edge weight \(wi\) represents the empirical frequency of given edge over samples of network structures

Back to article page