|
Abstract:
|
One of the challenges of RNA-seq analysis is differential expression analysis, i.e., the analysis to identify genes whose expression levels differ across the levels of one or more categorical factors of interest. In addition to the factor of primary scientific interest, RNA-seq datasets often contain several covariates. As in any experiment or observational study, covariates may hold information about the heterogeneity of the experimental or observational units used in the investigation. Other covariates may track variation that is created during the complex process of measuring RNA transcript abundance levels using RNA-seq technology. Either ignoring relevant covariates or accounting for the effects of irrelevant covariates can be detrimental in RNA-seq analysis. In this talk, we propose a backward selection strategy for RNA-seq differential expression analysis based on pseudo-variables. The proposed method’s performance is investigated and compared with the performance of other existing methods via a simulation study.
|
