Modeling Framework for Biomarker Discovery in Metabonomics
Author Information
Author(s): Ghosh Samiran, Grant David F, Dey Dipak K, Hill Dennis W
Primary Institution: Department of Mathematical Sciences, Indiana University Purdue University, Indianapolis, Indiana, USA
Hypothesis
The mass spectral profile of patients will differ significantly from that of healthy individuals.
Conclusion
The proposed modeling process is a novel approach for analyzing metabonomic data and requires clinical validation of potential biomarkers.
Supporting Evidence
- The study used a novel semi-parametric approach to distinguish urinary metabolic profiles.
- Data sets were analyzed in replicate to address instrument variability.
- The proposed method could be extended to other domains as long as the physical mechanism remains the same.
Takeaway
Researchers created a new way to analyze urine samples from trauma patients to find possible biomarkers that could help in diagnosis.
Methodology
A semiparametric approach was developed using Dirichlet mixture of beta distribution to analyze mass spectrometry data from urine samples.
Limitations
The study primarily focused on modeling aspects and requires clinical validation of the identified biomarkers.
Participant Demographics
6 normal healthy individuals and 6 patients with different degrees of acute trauma.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website