Passing Messages between Biological Networks to Refine Predicted Interactions

Abstract

Regulatory network reconstruction is a fundamental problem in computational biology. There are significant limitations to such reconstruction using individual datasets, and increasingly people attempt to construct networks using multiple, independent datasets obtained from complementary sources, but methods for this integration are lacking. We developed PANDA (Passing Attributes between Networks for Data Assimilation), a message-passing model using multiple sources of information to predict regulatory relationships, and used it to integrate protein-protein interaction, gene expression, and sequence motif data to reconstruct genome-wide, condition-specific regulatory networks in yeast as a model. The resulting networks were not only more accurate than those produced using individual data sets and other existing methods, but they also captured information regarding specific biological mechanisms and pathways that were missed using other methodologies. PANDA is scalable to higher eukaryotes, applicable to specific tissue or cell type data and conceptually generalizable to include a variety of regulatory, interaction, expression, and other genome-scale data. An implementation of the PANDA algorithm is available at www.sourceforge.net/projects/panda-net.

BGRMI: A method for inferring gene regulatory networks from time-course gene expression data and its application in breast cancer research

“…Many recent GRN reconstruction methods, e.g. RNEA76, PANDA77, PTHGRN78, APG79, CMGRN80, BVS12 also have this feature. However, these algorithms have their advantages and disadvantages.…”

From systems biology to P4 medicine: applications in respiratory medicine

“…2) XUE et al [91] used other network-centric procedures to reveal an unexpected loss of inflammatory signature in COPD patients, as well as an activation-independent core signature for human and murine macrophages. 3) GLASS et al [92] used the network inference analysis PANDA (Passing Attributes between Networks for Data Assimilation) [93], designed for improved integration of individual with public datasets, and discovered network rewiring of lymphocyte activation signalling circuits in a known gene variant implicated in COPD by genome-wide association studies. 4) FANER et al [94] unravelled differences in the molecular pathogenesis of emphysema and bronchiolitis by performing correlation network analysis of lung transcriptomics on COPD patients.…”

Development of statistical tools for integrating time course ‘omics’ data

“…For example the algorithm 'passing messages between biological networks to refine predicted interactions' (PANDA) [75] outperformed methods based on semi-supervised learning [76], model reconstruction methods [77] and methods using data-driven mutual information in gene expression to predict regulatory networks [78,79].…”

“…The integration of experimentally obtained protein-protein interactions (PPI) [9,97,98] with gene expression data, has been shown to increase specificity of networks built [75].…”

“…While integrating PPI information has shown to increase specificity of networks built [75], however these PPI maps are considered incomplete [99] and more importantly PPIs have been found to alter under specific conditions [100]. It is therefore important to create and analyse conditional specific molecular interaction networks to gain new insights and discover condition specific correlated molecular expression.…”