Visualization of module alignment discovery

Abstract

Network alignment is an efficient approach for understanding similarities and dissimilarities of protein interaction networks (PINs) from different species. However, it's difficult to visualize matching status for large-scale PINs even if accurate alignments are found. As we know, PINs can be structured by clustering based on detecting modularity. In this paper, we combine clustering with network alignment to display alignment results, and pairs of modules that most proteins in them are matched, named aligned modules, are discovered intuitively by some visualization tools we developed. Experiments on Homo sapiens and Saccharomyces cerevisiae PIN datasets demonstrate that the visualization is helpful for analyzing large scale network alignments, especially when module alignments are discovered in a visual way.