CADLIVE Automatic Visualizing System for Large-Scale Biochemical Maps

Abstract

Molecular biological network data accumulate rapidly. Computer-aided visual tools are required tohelp researchers to gain insights into the data. CADLIVE (Computer-Aided Design of LIVing sys-tEms) is a suite of software for such purpose, which features a graphical user interface to operatelarge-scale biological network data [2], but users have to manually describe the layouts of networkcomponents in the CADLIVE graphical representations. It needs a lot of work to make intuitiveand heuristic layouts for complicated networks. General graph drawing techniques are not suitablefor biochemical reaction networks. Even the most analogous method specially designed for drawingmetabolic pathways [1] is inadequate for producing satisfactory layouts of gene regulatory networks.In this paper, we propose a grid layout method for the purpose of automatically visualizing net-worksinCADLIVE.Thealgorithmemphasizestoshowrelationshipsbetweenthenetworkcomponents.The layouts are obtained through minimizing a cost function of layouts. A discrete local minimizationtechnique is used in a modi ed simulated annealing to facilitate the search of acceptable minima.