Differential expression and network inferences through functional data modeling.

Publication Type:

Journal Article


Biometrics, Volume 65, Issue 3, p.793-804 (2009)


2009, Center-Authored Paper, Computer Simulation, Data Interpretation, Statistical, Gene Expression Profiling, Humans, Male, Models, Biological, Models, Statistical, Neoplasm Proteins, Oligonucleotide Array Sequence Analysis, Prostatic Neoplasms, Public Health Sciences Division, Signal Transduction, Tumor Markers, Biological


Time course microarray data consist of mRNA expression from a common set of genes collected at different time points. Such data are thought to reflect underlying biological processes developing over time. In this article, we propose a model that allows us to examine differential expression and gene network relationships using time course microarray data. We model each gene-expression profile as a random functional transformation of the scale, amplitude, and phase of a common curve. Inferences about the gene-specific amplitude parameters allow us to examine differential gene expression. Inferences about measures of functional similarity based on estimated time-transformation functions allow us to examine gene networks while accounting for features of the gene-expression profiles. We discuss applications to simulated data as well as to microarray data on prostate cancer progression.