A weighted sample size for microarray datasets that considers the variability of variance and multiplicity.

Area of Science:

• Genomics
• Biostatistics

Background:

• Microarray experiments are crucial for identifying differentially expressed genes across clinical phenotypes.
• Standard sample size calculations are inadequate for microarrays due to the high number of genes analyzed.
• The multiplicity problem in high-throughput screening requires specialized statistical approaches.

Purpose of the Study:

• To develop a novel sample size calculation method for microarray studies.
• To address the multiplicity issue inherent in analyzing tens of thousands of genes.
• To ensure reliable identification of significant genes in genomic data.

Main Methods:

• Proposed a sample size calculation method incorporating variance across the entire gene set.
• Utilized the Bonferroni correction to manage the multiplicity problem.
• Employed k-means cluster analysis to group genes by variance and calculated weighted-averaged sample sizes.

Main Results:

• The proposed method demonstrated that sample size is dependent on the number of genes per microarray chip.
• The weighted sample size calculation effectively controlled the Type I error rate.
• The method provides a more accurate sample size determination for gene expression studies.

Conclusions:

• The developed sample size calculation method is suitable for microarray experiments.
• This approach enhances the reliability of detecting differentially expressed genes.
• Accurate sample size determination is critical for robust genomic research findings.