What's in a Workflow: Understanding Time Course DNA Microarray Data using Multivariate Curve Resolution

<p>Time course DNA microarray experiments have become more prevalent in systems biology studies as they allow the investigation of transcriptome-wide changes over time. Multivariate curve resolution (MCR), a chemometrics technique, has recently been effectively applied to microarray time course data [1]. Successful application of MCR to time course microarray data requires some modifications to normal practices in both chemometrics and bioinformatics. These modifications concern issues of measurement quality, data transformation, and error estimation. As a consequence of analyzing a variety of DNA microarray time course studies, we have developed a set of standard workflow procedures that appear to be effective in extracting the desired information and avoiding some of the pitfalls associated with this type of data. In this work we describe some of the aspects of these procedures with examples from relevant data sets.