Related Experiment Videos
Feature selection using logistic regression in case-control DNA methylation data of Parkinson's disease: A
Aishwarya Kakade1, Baby Kumari1, Pankaj Singh Dholaniya1
1Department of Biotechnology, School of Life Sciences, University of Hyderabad, Hyderabad, Telangana 500 046, India.
Journal of Theoretical Biology
|August 19, 2018
Summary
Epigenetic modifications, specifically DNA methylation patterns, are implicated in Parkinson's disease (PD) progression. This study identified key gene methylation patterns using Logistic Regression, improving diagnostic accuracy for Parkinson's disease.
Area of Science:
- Neuroscience
- Genetics
- Epigenetics
Background:
- Parkinson's disease (PD) is a neurodegenerative disorder resulting from dopaminergic neuron loss.
- Its pathogenesis involves complex genetic, epigenetic, and environmental factors.
- Epigenetic modifications, particularly DNA methylation, are increasingly recognized for their role in PD.
Purpose of the Study:
- To analyze DNA methylation patterns in Parkinson's disease.
- To identify potential epigenetic biomarkers for PD using machine learning.
- To evaluate feature reduction methods for classifying PD based on methylation data.
Main Methods:
- Analysis of DNA methylation data from 66 samples (43 controls, 23 PD patients).
- Utilized Logistic Regression (LR) for feature reduction and classifier development.
- Compared LR performance with Random Forest (RF) and Principal Component Analysis (PCA).
Main Results:
- Logistic Regression achieved improved accuracy in classifying PD compared to using all features.
- LR and RF demonstrated superior performance over PCA in feature reduction.
- LR uniquely identified significant PD-associated genes, including COMT, DCTN1, and PRNP.
Conclusions:
- DNA methylation patterns can serve as valuable biomarkers for Parkinson's disease.
- Logistic Regression is an effective method for feature selection in PD epigenetics research.
- Identified genes like COMT, DCTN1, and PRNP warrant further investigation for their role in PD pathogenesis.