Composite MicroRNA Target Predictions and Comparisons of Several Prediction Algorithms

• Activity Number: 523
• Type: Invited
• Date/Time: Thursday, August 10, 2006 : 10:30 AM to 12:20 PM
• Sponsor: Biometrics Section
• Author(s): Jin Zhou*+ and Vincent Melfi and Joe Verducci and Shili Lin
• Companies: The Ohio State University and Michigan State University and The Ohio State University and The Ohio State University
• Keywords: microRNA ; target (gene) predictions ; composite statistics ; gene ontology ; Kendall's distance ; Spearman's footrule distance

Abstract

MicroRNAs (or miRs) are noncoding RNAs whose role is to repress translation by regulating gene expression through binding to mRNA targets. There are computational algorithms for miR target predictions, but their results vary. Thus, it would be useful to consolidate these results to have a greater degree of certainty before engaging in costly experiments. We studied three popular algorithms, miRanda, TargetScan, and PicTar, systematically through the use of three measures of similarity and a statistical test on the gene ontology categories. Two composite statistics were also devised to combine and rank the composite target list. We applied these methods to all human miRs. Our results indicate that TargetScan and PicTar tend to have a greater degree of similarity. We also demonstrate that our composite statistics can be useful tools for short listing genes for downstream experiments.