#06227: Protein Inference in "Shotgun" Mass Spectrometry Proteomics
A software solution that improves protein inference from identified peptides in shotgun proteomic experiments
………………………………………………………………………………………………………………
TECH FIELD(S)
Mass Spectrometry Bioinformatics ………………………………………………………………………………………………………………
FEATURES
- Peptide Detectability is calculated as the probability of identifying a peptide in a standard proteomics experiment
- These probabilities are combined in another calculation to determine the most likely protein composition
- Assigned proteins are given a confidence score from 0-1
………………………………………………………………………………………………………………
BENEFITS
Spectrometry hardware now allows the fine separation of complex protein mixtures. And while peptides may be faithfully identified from ion fragmentation data, their parent proteins are harder to characterize. Often only 10-25% of proteins are identified in high throughput experiments—the protein inference problem.
The protein inference problem stems from degenerate peptides—those that comprise multiple proteins. This solution uses peptide detectability—a peptide’s probability of observation—to better infer proteins.
Machine learning algorithms incorporate information from the protein sequence and structure to calculate this probability. A secondary calculation then gives a confidence score for assigned proteins.
Experimental results have shown dramatic increases (50-100%) in correctly identified proteins.
………………………………………………………………………………………………………………
INVENTOR(S)
Randy Arnold (Department of Chemistry), Pedja Radivojac, and Haixu Tang, (Indiana University School of Informatics)
………………………………………………………………………………………………………………
INTELLECTUAL PROPERTY STATUS
U.S. Utility Filing available for licensing
………………………………………………………………………………………………………………
CONTACT INFORMATION
For additional information on this technology, please contact Doug Bodde, Business Development Manager, IURTC:
Indiana University Research & Technology Corporation
351 West 10th Street
Indianapolis, IN 46202
Telephone: 812.855.1608
dbodde@iu.edu
………………………………………………………………………………………………………………
Still have a question? Ask here!