GPNN: A Neural Network Method for Detecting Gene Interactions in Disease Studies
Author Information
Author(s): Alison A Motsinger, Stephen L Lee, George Mellick, Marylyn D Ritchie
Primary Institution: Vanderbilt University Medical School
Hypothesis
The study aims to evaluate the power of GPNN for identifying high-order gene-gene interactions.
Conclusion
GPNN may be a useful method for detecting gene-gene and gene-environment interactions, particularly in Parkinson's disease.
Supporting Evidence
- GPNN has high power to detect gene-gene interactions in simulated data models.
- The model correctly predicts PD status 59.66% of the time.
- GPNN detected a significant interaction between the DLST gene and sex.
Takeaway
This study shows that a special computer program can help find how different genes work together to affect diseases like Parkinson's.
Methodology
The study used a genetic programming optimized neural network (GPNN) to analyze simulated data and real datasets from Parkinson's disease cases and controls.
Limitations
The sample sizes may be too small for detecting high-order interactions, and the study's findings need replication in larger datasets.
Participant Demographics
305 Parkinson's disease patients and 321 controls, with a majority being male.
Statistical Information
P-Value
p < 0.001
Statistical Significance
p < 0.001
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website