Plant & Animal Genomes XIII Conference

January 15-19, 2005
Town & Country Convention Center
San Diego, CA

The PowerAtlas: A Power And Sample Size Atlas For Microarray Research Based Upon 458 Public Datasets

Grier P Page¹ , David B Allison¹ , Jelai Wang¹ , Jode W Edwards²

¹  Department of Biostatistics University of Alabama at Birmingham 1665 University Blvd RPHB 327 Birminghma, Al 35294

²  USDA-ARS Department of Agronomy Iowa State University Ames, IA 50011

Microarrays are an exciting new tool allowing biologists to simultaneously measure the mRNA abundance of thousands of genes. One of the most important issues facing investigators planning microarray experiments is choosing the optimal sample size, or number of replicate chips. While statistical methods exist to estimate optimal sample sizes from previously conducted experiments, biological researchers generally don’t have the software and expertise available to estimate optimal sample sizes and very often do not have data from previous experiments. To address the inability of investigators to choose optimal sample sizes, we have constructed a Microrarray Power Atlas (www.poweratlas.org) containing estimated sample size requirements for 458 experiments from Gene Expression Omnibus. Our objectives in constructing a Power Atlas are to i.) provide investigators easily accessible information to use in planning microarray experiments, ii.) use the Atlas to educate investigators on basic statistical principles in biological research, iii.) develop a biological database directed at dissemination of sophisticate statistical computations and graphical representations of them. In the near future data from the Stanford Microarray Database as well as NASCArray will be added. In addition investigators will be able to use their own data for estimating power for future studies.