Using Mass Spectrometry to Identify Ovarian Cancer Biomarkers
Author Information
Author(s): Ye Jingjing, Liu Hao, Kirmiz Crystal, Lebrilla Carlito B, Rocke David M
Primary Institution: University of California, Davis
Hypothesis
Can glycomics mass spectrometry data be effectively analyzed to identify biomarkers for ovarian cancer diagnosis?
Conclusion
The TGDR-AUC algorithm is effective in combining potential biomarkers for ovarian cancer diagnosis and provides good performance.
Supporting Evidence
- The TGDR-AUC algorithm achieved an empirical AUC of 0.95 for the low-dimensional data.
- The method was validated through a simulation study showing asymptotic properties.
- The algorithm identified 63 significant biomarkers from a total of 1228 in high-dimensional data.
Takeaway
Researchers used a special method to look at blood samples and found ways to tell if someone has ovarian cancer by looking at tiny sugar molecules.
Methodology
The study used a constrained threshold gradient descent regularization algorithm to analyze mass spectrometry data and identify biomarkers.
Potential Biases
Potential bias due to the small sample size relative to the number of biomarkers.
Limitations
The results may overestimate the AUC in high-dimensional data with small sample sizes.
Participant Demographics
73 patients, including 49 ovarian cancer patients and 24 healthy patients.
Statistical Information
P-Value
7.827e-07
Confidence Interval
(0.9513, 0.9560)
Statistical Significance
p<0.05
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website