PatternLab

PatternLab provides analysis of quantitative proteomics data to detect differential protein expression and perform feature selection and normalization for proteomic studies.

Key Features:

• Spectral Count Normalization: Implements techniques to normalize spectral counting data for quantitative estimation of protein abundance.
• Differential Protein Expression Analysis: Implements ACFold and nSVM methods to identify differentially expressed proteins.
• ACFold Method: Combines expression fold changes, the AC test, and false-discovery rate calculations and is tailored for experiments with fewer than three replicates or varied protocols.
• nSVM Methodology: The nSVM (natural support vector machine) is rooted in evolutionary computing and statistical learning theory and is suited to designs with multiple readings per state and selection of minimal protein sets for classification.
• Unified Feature Selection and Normalization Strategies: Integrates multiple feature selection and normalization strategies tailored to different experimental designs.
• Graphing Tools: Includes graphing tools to visualize high-throughput experimental proteomic data.

Scientific Applications:

• Differential Proteomics: Distinguishing between biological states by identifying protein expression differences.
• Low-Replicate or Variable-Protocol Experiments: Analysis of experiments with fewer than three replicates or differing protocols using ACFold.
• Biomarker Discovery and Diagnostic Development: Selection of minimal protein sets for classification to support biomarker discovery and development of diagnostic kits.
• Classification and Early Detection: Use of nSVM for classifier selection in experimental designs with multiple readings per state.

Methodology:

Implements spectral count normalization; ACFold combining expression fold changes, the AC test, and false-discovery rate calculations; nSVM (natural support vector machine) based on evolutionary computing and statistical learning theory; and integration of feature selection and normalization strategies with graphing for visualization.