CGHpower: exploring sample size calculations for chromosomal copy number experiments

Area of Science:

• Genomics
• Bioinformatics
• Statistical Genetics

Background:

• Sample size determination is critical for microarray experiment planning.
• Statistical power increases with more arrays, but at a higher cost.
• No methods previously existed for sample size calculation in array comparative genomic hybridization (aCGH).

Purpose of the Study:

• To explore power calculations for aCGH experiments comparing two groups.
• To introduce a statistical framework for estimating average power based on sample size using pilot data.

Main Methods:

• The CGHpower method utilizes pilot data to estimate biological diversity between groups.
• It provides a statistical framework for power calculations as a function of sample size.
• The method can be applied during the planning of new studies or to assess past experiments.

Main Results:

• CGHpower estimates statistical power for aCGH experiments.
• The method requires pilot data to estimate biological diversity and power.
• It can be used for planning future studies or evaluating past ones.

Conclusions:

• The proposed method has limitations as its assumptions may not always hold true.
• Violated assumptions can lead to unreliable sample size estimates.
• CGHpower is currently the only available method for aCGH sample size calculation, offering diagnostic plots for assumption assessment.