Related Experiment Video
Updated: Dec 10, 2025
07:41
Performing Data Mining And Integrative Analysis Of Biomarker in Breast Cancer Using Multiple Publicly Accessible Databases
Published on: May 17, 2019
9.3K
Type2 diabetes mellitus prediction using data mining algorithms based on the long-noncoding RNAs expression: a
Faranak Kazerouni1, Azadeh Bayani2, Farkhondeh Asadi3
1Department of Laboratory Medicine, School of Allied Medical Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran.
BMC Bioinformatics
|August 29, 2020
Summary
Support Vector Machine (SVM) and logistic regression models show high accuracy in predicting Type 2 Diabetes Mellitus (T2DM) using long non-coding RNA (lncRNA) expression data. These machine learning approaches offer promising tools for early T2DM diagnosis.
Area of Science:
- Biomedical Informatics
- Genomics
- Computational Biology
Background:
- Type 2 Diabetes Mellitus (T2DM) affects approximately 90% of diabetic patients.
- Long non-coding RNAs (lncRNAs) play a significant role in T2DM pathogenesis and diagnosis.
- Machine learning (ML) and Data Mining (DM) techniques enhance disease analysis and prognosis.
Purpose of the Study:
- To evaluate and compare the diagnostic performance of four ML models for T2DM.
- To identify the optimal data mining approach for T2DM prediction using lncRNA expression.
- To assess the utility of specific lncRNAs as biomarkers for T2DM detection.
Main Methods:
- Applied four classification models: K-nearest neighbor (KNN), Support Vector Machine (SVM), logistic regression, and Artificial Neural Networks (ANN).
- Utilized six lncRNA variables (LINC00523, LINC00995, HCG27_201, TPT1-AS1, LY86-AS1, DKFZP) and demographic data for analysis.
- Evaluated model performance using Area Under the Curve (AUC), sensitivity, specificity, and Receiver Operating Characteristic (ROC) curves.
Main Results:
- SVM and logistic regression achieved the highest mean AUC of 95% with a standard deviation of 0.05.
- KNN demonstrated the highest mean sensitivity (96%), while SVM exhibited the highest specificity (86%).
- ANN achieved a mean AUC of 93% (SD 0.03), and KNN achieved a mean AUC of 91% (SD 0.09).
Conclusions:
- SVM and logistic regression are the most effective data mining approaches for T2DM prediction based on lncRNA expression.
- KNN and ANN also showed high mean AUC values and small standard deviations, indicating robust performance.
- This study highlights the potential of lncRNAs as biomarkers for early T2DM detection and diagnosis.
