A method for effectively comparing gene effects in multiple conditions in RNAi and expression-profiling research

Abstract

Aim: To develop a new analytical method to address the issues of traditional contrast analysis for comparing gene effects in RNAi and expression-profiling research. Methods & results: I propose a new method consisting of contrast variable, standardized mean of contrast (SMC) and c+-probability analysis for comparing gene effects in multiple conditions. Compared with traditional contrast analysis, this new method has the following major advantages: it directly addresses the primary question of interest, namely the assessment of the strength of comparison; SMC and c+-probability capture data variability and are robust to sample size. The simulation and application studies show that traditional contrast analysis produces misleading results and erroneous conclusions whereas the new method produces reasonable results and sensible conclusions. Conclusions: The new method may have a broad utility in comparing gene effects in multiple conditions including selecting hits in RNAi research and identifying differential expression in microarray experiments.