Outlier Identification using Nonmetric Multidimensional Scaling of Yeast Cell Cycle Phase using Gene Expression Data

Current researches focused on gene function classification and discovery are with the use of wet laboratory. This research focused on the identification of outlier yeast genes, Saccharomyces cerevisiae involved in a eukaryotic cell cycle using time series normalized gene expression data. A method for identifying outlier genes using Nonmetric Multidimensional Scaling (nMDS) with confidence intervals of 95% and confidence ellipse of 95% is used for the computing method for identifying the goodness of fit per group. This method shows a good identification of outlier genes based on the identified genes per cell cycle phases, using criteria identified for visualization associated with confidence interval. Visualization of the data set captures the group structure of genes based from the cell cycle. It shows the characteristics of the events of the genes and identified outliers are included at the adjacent groups. Based on this study, 25 outlier genes were identified, 6.51% of the gene set population.