Machine Learning for Analyzing Peptide Fragmentation Patterns
Author Information
Author(s): Zhou Cong, Bowler Lucas D, Feng Jianfeng
Primary Institution: University of Sussex
Hypothesis
What factors influence peptide fragmentation during CID?
Conclusion
The intensity patterns of fragmentation spectra are informative and can be used to analyze the influence of various characteristics of fragmented peptides on their fragmentation pathway.
Supporting Evidence
- The model can accurately predict intensity patterns for given MS/MS spectra.
- The study identified significant features influencing peptide fragmentation.
- The predictions include mean values and variances to tolerate noise in experimental data.
Takeaway
This study uses a computer model to understand how peptides break apart in mass spectrometry, which can help identify proteins better.
Methodology
A Bayesian neural network was used to analyze ion intensity information from MS/MS spectra and predict intensity patterns based on selected features.
Limitations
The study primarily focused on mobile and partial-mobile peptides, with fewer non-mobile data, which may lead to missing important fragmentation rules.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website