Related Experiment Video
Updated: Sep 9, 2025
10:37
Deep Proteome Profiling by Isobaric Labeling, Extensive Liquid Chromatography, Mass Spectrometry, and Software-assisted Quantification
Published on: November 15, 2017
12.1K
IQUP identifies quantitatively unreliable spectra with machine learning for isobaric labeling-based proteomics.
Yi-Yun Chou1, Jing-Chen Yang1, Tun-Chu Tsai1
1Department of Bioinformatics and Biomedical Engineering, Asia University, Taichung, 413, Taiwan.
Scientific Reports
|August 29, 2025
Summary
We developed IQUP, a machine learning method to identify unreliable peptide-spectrum matches (PSMs) in mass spectrometry proteomics. This improves quantitation accuracy by filtering out low-quality data, enhancing proteomic analysis reliability.
Area of Science:
- Proteomics
- Mass Spectrometry
- Computational Biology
Background:
- Isobaric labeling technology is widely used for quantitative proteomics.
- Current methods using target-decoy search may include peptide-spectrum matches (PSMs) with high quantitation errors.
- This can compromise the overall accuracy of proteomic quantitation.
Purpose of the Study:
- To introduce IQUP, a machine learning-based method for identifying quantitatively unreliable PSMs (QUPs).
- To improve the accuracy of proteomic quantitation in isobaric labeling experiments.
Main Methods:
- PSMs were characterized using 16 spectral and distance-based features.
- A machine learning approach was employed to train models for QUP identification.
- Performance was evaluated on three independent datasets using accuracy, AUC, and MCC metrics.
Main Results:
- The best models achieved high performance with accuracies of 0.883-0.966, AUCs of 0.924-0.963, and MCCs of 0.596-0.691.
- Significant differences in relative error distributions were observed between QUPs and quantitatively reliable PSMs (QRPs).
- Utilizing only predicted QRPs for quantitation significantly reduced peptides with large relative errors (15.3-83.3%) and increased peptides with small relative errors (3.1-25.5%).
Conclusions:
- IQUP demonstrates robust performance and generalizability across diverse datasets.
- The method effectively identifies quantitatively unreliable PSMs.
- IQUP has significant potential to enhance proteomic quantitation accuracy at both PSM and peptide levels.
