|Home | About | Journals | Submit | Contact Us | Français|
Our objective was to develop a formal process that allows two bioinformaticians to provide sophisticated support for 3-6 Affymetrix GeneChip experiments per week. The process provides investigators with support for experiment design, quality assessment, exploratory multivariate analysis, linear modeling, and interpretation of differential expression statistics in the context of biological systems, as well as publication. The process provides administrators with reports on utilization and productivity. Software associated with the process is based on a MySQL relational database and R/Bioconductor tools. Information about our process, data model, software, utilization, and productivity will be presented. In addition, we will highlight 1) use of mixture models to analyze genome-wide expression in heterogeneous cell populations, 2) use of principal component analysis and the surrogate variable analysis of Leek and Storey to capture more general types of latent variation, and 3) integration of time series oligonucleotide array data with cell viability data for the purpose of understanding drug interactions at the level of both transcription and survival. In particular, our model of cell survival supports models of heterogeneous cell populations as well as complicated dependence of cell survival on drug dose and time.