Related Experiment Video
Updated: Apr 24, 2026
08:27
Large-Scale Multi-Omics Genome-Wide Association Studies Mo-GWAS: Guidelines for Sample Preparation and Normalization
Published on: July 27, 2021
4.6K
A case study of normalization, missing data and variable selection methods in lipidomics
1Department of Mathematics and Statistics, University of Turku, Turku, FI - 20014, Finland.
Statistics in Medicine
|September 5, 2014
Summary
Lipidomics analysis can identify disease biomarkers. This study developed statistical methods for preprocessing lipidomics data, improving biomarker discovery for clinical applications.
Area of Science:
- Biomedical Science
- Computational Biology
- Statistical Genetics
Background:
- Lipidomics offers potential for early disease detection through biomarker identification.
- Efficient statistical methodologies are crucial for translating lipidomics findings into clinical practice.
Purpose of the Study:
- To identify an effective statistical methodology for lipidomics data analysis.
- To find interpretable and predictive biomarkers for clinical applications.
Main Methods:
- Data preprocessing including normalization and multiple imputation for handling experimental variability and missing data.
- Comparison of stepwise variable selection and penalized regression models using cross-validation on imputed datasets.
- Utilizing a permutation test for global association testing.
Main Results:
- Data preprocessing methods demonstrated modest improvements in classification precision.
- No single variable selection method consistently outperformed others across different study designs.
- Lipidomics profiles were identified as significant predictors in both case studies.
Conclusions:
- The developed statistical approach aids in identifying robust lipidomics biomarkers.
- Effective data preprocessing is essential for reliable biomarker discovery in lipidomics.
- Lipidomics data holds significant predictive value for disease identification.
