Microarray studies aim to discover which genes are differentially expressed under different experimental conditions in biological samples. Before any microarray study begins, the investigator should plan carefully and answer important questions like the following: (a) How many sample tissues should be included in the experiment? (b) How many times should the experiment be replicated? (c) What statistical power does the experiment have to uncover a specified level of genetic differential expression? Experimental designs for microarray studies vary widely. Hence, the answers to these kinds of questions are important for any type of microarray experimental study. In this proposal, we discuss conceptual issues and present computational methods for statistical power and sample size in microarray studies. The proposed research program will encompass choices of experimental design and replication for a study. The proposed analytical approach avoids the use of the observed mean square error and, hence, makes no use of t or F statistics at the level of the individual gene. The presentation in the proposal makes reference to cDNA arrays for illustrations and discussion but the suggested methodology is equally applicable to expression data from oligonucleotide arrays.
The specific aims are listed below.
Aim 1. Derive formulas for computing power and sample size for different types of hypotheses being tested in microarray studies.
Aim 2. Quantitatively assess the effect of experimental design and replication in terms of statistical power for microarray studies.
Aim 3. Develop software for computing power and sample size for different experimental designs.
Aim 4. Demonstrate application of the derived methodology in several practical studies.
| 1000 Genomes Project Consortium; Abecasis, Gonçalo R; Altshuler, David et al. (2010) A map of human genome variation from population-scale sequencing. Nature 467:1061-73 |
| Tsai, Pi-Wen; Lee, Mei-Ling Ting (2005) Split-plot microarray experiments: issues of design, power and sample size. Appl Bioinformatics 4:187-94 |
| Lee, Mei-Ling Ting; Whitmore, G A; Bjorkbacka, Harry et al. (2005) Nonparametric methods for microarray data based on exchangeability and borrowed power. J Biopharm Stat 15:783-97 |
