The PowerAtlas: a power and sample size atlas for microarray experimental design and research

Area of Science:

• Genomics
• Bioinformatics
• Statistical Genetics

Background:

• Microarrays enable simultaneous measurement of thousands of gene mRNA abundances.
• Estimating required sample sizes for robust statistical power in microarray experiments is a significant challenge.
• Existing statistical methods for sample size calculation are often insufficient for multiple hypothesis testing in microarrays, especially without pilot data.

Purpose of the Study:

• To develop a resource for estimating statistical power and sample sizes for microarray studies.
• To provide a tool that assists investigators in planning efficient microarray experiments.

Main Methods:

• Development of the Microrarray PowerAtlas, a novel computational resource.
• Inclusion of sample size and power estimates from 632 experiments in the Gene Expression Omnibus (GEO).
• Functionality for investigators to upload their own pilot data for customized power and sample size estimations.

Main Results:

• The Microrarray PowerAtlas provides power and sample size estimates based on a large repository of existing microarray experiments.
• The resource allows for study planning by utilizing data from similar previous studies.
• Investigators can derive personalized estimates by uploading their own pilot data.

Conclusions:

• The Microrarray PowerAtlas is a valuable tool for optimizing microarray study design.
• It aids researchers in efficiently determining the necessary sample sizes for achieving desired statistical accuracy.
• The resource facilitates more informed and effective planning of gene expression studies.