Predicting Cell Cycle Genes from E-MAP Profiles by Integrating Multiple Types of Data

Interactions between genes and proteins can be revealed by multiple experimental platforms. The derived interaction networks can be utilized to discover novel genes involved in specific biological process. E-MAP is an experimental platform to measure genetic interactions in a genome-wide scale, which successfully recovered known pathways and also revealed novel protein complexes in S. cerevisiae. However, E-MAP data can be quite noisy, and it is of great challenge to make reliable biological inference based on it. Here we propose a novel approach which aims to discover genes involved in the cell cycle process in S. cerevisiae by combining E-MAP data with other sources of data, such as gene expression, protein phosphorylation, and transcription factor (TF)–DNA binding. From an E-MAP screen with 35 query genes, we predict three unknown genes (YPL158C, YPR174C and YJR054W) as potential cell cycle genes. Our strategy can be applied to other biology processes as well.