Inferring Time-Lagged Association Rules from Microarray Time-Series Data

GALLO, CRISTIAN ANDRÉS; CARBALLIDO, JESSICA ANDREA; PONZONI, IGNACIO

Las Termas de Chillan

Conferencia; The 1 st International Conference on Bioinformatics SoIBio 2010; 2010

Iberoamerican Bioinformatics Society

Gene regulatory networks have an important role in every process of life, including celldifferentiation, metabolism, the cell cycle and signal transduction. The amount of large scale andgenome wide time-series data is becoming increasingly available, which provides the opportunityto reverse engineer the time-delayed gene regulatory networks that govern the majority of thesemolecular processes. In particular, this paper aims to reconstruct regulatory networks frommicroarray time-series data by using delayed correlations between genes. Thereby, a new modelfree computational toolbox termed GRNCOP2 (Gene Regulatory Network inference byCombinatorial OPtimization v. 2) was developed in order to address the underlying associationbetween genes that can span any unit(s) of time intervals. This toolbox overcomes the limitationsof the original version1 by means of several improvements. The proposed method was applied toa time-series data of 20 yeast genes and the results were compared against several approachesavailable in the literature1, 2, 3. In order to assess and contrast the performance of the methods, the proportion of associations (i.e., rules without repetition) that share any GO biological processterms of a Gene Ontology annotation downloaded from the Saccharomyces cerevisiae GenomeDatabase4 was measured. Figure 1 shows the scores achieved for the algorithms with an accuracyof 0.75 for simultaneous and time-delayed rules. As it can be observed, GRNCOP2 outperformsall the referential methods in terms of the proportion of rules that validate with the GOannotations. This demonstrates that the improvements incorporated in this new version of thealgorithm provide a usable and powerful model-free approach to dissecting high-order dynamictrends of gene-gene interactions, carefully validated with publicly available cell cycling datasetsthrough the well recognized S. cerevisiae Genome Database.