Using Peptide-Level Proteomics Data for Detecting Differentially Expressed Proteins.

Area of Science:

• Proteomics
• Mass Spectrometry
• Bioinformatics

Background:

• High-throughput protein quantification is crucial in biological research.
• Label-free mass spectrometry methods are increasingly used for protein abundance determination.
• Conventional analysis aggregates peptide-level data to protein-level, potentially losing information and introducing inconsistencies.

Purpose of the Study:

• To develop and validate a novel method for detecting differentially expressed proteins.
• To improve the accuracy of differential expression analysis by leveraging peptide-level statistics.
• To provide a user-friendly bioinformatics tool for this new approach.

Main Methods:

• A new method combining peptide-level expression-change statistics for differential protein expression analysis.
• Controlled spike-in experiments were utilized for validation.
• Implementation in the Bioconductor package PECA.

Main Results:

• The proposed method of averaging peptide-level expression changes demonstrated higher accuracy in identifying differentially expressed proteins compared to the conventional protein-level approach.
• Improved performance was particularly noted in scenarios with few replicate samples or small expression differences.
• The PECA package provides an accessible implementation of this technique.

Conclusions:

• Combining peptide-level expression-change statistics offers a more accurate and robust method for differential protein expression analysis.
• This approach enhances the reliability of findings, especially in challenging experimental conditions.
• The PECA package facilitates the adoption of this advanced technique in proteomics research.