NoiseMaker: simulated screens for statistical assessment.

Area of Science:

• Computational Biology
• Bioinformatics
• Drug Discovery

Background:

• High-throughput screening (HTS) is crucial for drug discovery and research.
• Evaluating hit identification methods for HTS data presents challenges due to limited information on sensitivity and specificity.

Purpose of the Study:

• To introduce NoiseMaker, an open-source software tool for simulating realistically noisy virtual screening data.
• To enable researchers to assess the performance of various hit identification techniques on simulated data.

Main Methods:

• Developed the open-source NoiseMaker software in C# for generating virtual screening data with realistic noise.
• Applied established hit identification methods to simulated data to evaluate their ability to recover true hits and avoid false positives.
• Demonstrated the utility of NoiseMaker in explaining conflicting reports on the B score hit identification method.

Main Results:

• NoiseMaker allows for the quantitative assessment of hit identification method performance.
• Simulations provide insights into the sensitivity and specificity of different techniques.
• The tool aids in understanding and resolving discrepancies in published HTS results.

Conclusions:

• NoiseMaker is a valuable tool for optimizing hit identification strategies in HTS.
• The software enhances the reliability and interpretability of HTS data across various screening types.
• It supports informed decision-making in drug discovery and basic research by clarifying method performance.