Betamax: towards optimal sampling strategies for high-throughput screens

Area of Science:

• Genetics
• Bioinformatics
• Statistical Genetics

Background:

• Sample size determination is crucial for high-throughput genetic screening.
• A priori sample size estimation can be unreliable due to difficulties in modeling effect sizes and variance.
• Improving statistical power is essential for robust genetic discoveries.

Purpose of the Study:

• To introduce a novel approach for estimating and continuously updating statistical power during genetic screening.
• To optimize sample allocation strategies for maximizing overall study power.
• To enhance the efficiency and reliability of high-throughput genetic screening designs.

Main Methods:

• Development of a method to estimate and dynamically update statistical power.
• Integration of power estimates into a sequential sampling strategy.
• Simulation studies to compare the proposed method against naive equal sample size allocation.

Main Results:

• The proposed method demonstrated significant gains in study recall compared to the naive strategy.
• Continuous power updating allows for adaptive sample size adjustments.
• Optimized sampling allocation achieved higher statistical power with the same total sample size.

Conclusions:

• Dynamic sample size estimation and adaptive allocation improve statistical power in genetic screening.
• This approach offers a more reliable and efficient alternative to traditional fixed sample size designs.
• The findings have implications for optimizing resource allocation in large-scale genetic studies.