NormalyzerDE

NormalyzerDE performs normalization evaluation and differential expression analysis for LC-MS quantitative proteomics and other omics datasets, implementing a retention time (RT)-segmented normalization to mitigate local and sample-specific biases caused by electrospray current fluctuations during LC-MS gradients.

Key Features:

• Retention Time-Segmented Normalization Approach: Implements a retention time (RT)-segmented normalization method to mitigate local and sample-specific biases caused by electrospray current fluctuations during LC-MS gradients.
• Multiple Normalization Methods: Supports a range of global normalization techniques and enables RT segmentation to be applied in combination with these methods.
• Normalization Quality Assessment: Evaluates the quality of applied normalization to assess the extent to which technical biases are minimized.
• Differential Expression Analysis: Performs differential expression analysis using the empirical Bayes Limma approach.
• Performance Evaluation: RT-segmented normalization approaches were evaluated on two spike-in datasets, detecting 8–36% more peptides and increasing recall by 2–35% compared with conventional methods.
• Enhanced Detection Capabilities: Combining RT-normalization with Limma enabled identification of 108% more spike-in peptides (2597 vs. 1249) in one comparison.

Scientific Applications:

• Quantitative proteomics: Normalization and differential expression analysis of LC-MS-based proteomics datasets to improve peptide detection and quantification.
• Other LC-MS omics: Application to other omics datasets generated by LC-MS where electrospray current fluctuations and retention time–dependent biases affect quantitation.
• Differential expression studies: Identification of differentially expressed features to support analyses of molecular mechanisms and disease-related pathways.

Methodology:

Methods explicitly include retention time (RT)-segmented normalization compatible with global normalization techniques, normalization quality assessment, and differential expression analysis using the empirical Bayes Limma approach, with performance evaluated on two spike-in datasets.