Related Experiment Video
Updated: May 28, 2026
09:34
A Combinatorial Single-cell Approach to Characterize the Molecular and Immunophenotypic Heterogeneity of Human Stem and Progenitor Populations
Published on: October 25, 2018
A compressed sensing based approach for subtyping of leukemia from gene expression data
Wenlong Tang1, Hongbao Cao, Junbo Duan
1Department of Biomedical Engineering, Tulane University, New Orleans, Louisiana 70118, USA. wtang@tulane.edu
Journal of Bioinformatics and Computational Biology
|October 7, 2011
Summary
Compressed sensing (CS) effectively classifies leukemia subtypes using gene expression data. This novel approach achieves high accuracy, promising improved diagnosis and personalized cancer therapies.
Area of Science:
- Genomics
- Bioinformatics
- Signal Processing
Background:
- High-throughput genomic data analysis requires advanced computational methods.
- Compressed sensing (CS) is a signal processing technique with potential for sparse data representation.
- The application of CS to genome-wide data analysis remains underexplored.
Purpose of the Study:
- To propose and evaluate a novel Compressed Sensing (CS)-based approach for genomic data classification.
- To apply the CS method to classify leukemia subtypes using gene expression data.
- To assess the performance and robustness of the CS method for leukemia subtyping.
Main Methods:
- Utilized Compressed Sensing (CS) theory for signal reconstruction from sparse representations.
- Employed four statistical features for significant gene selection from 7,129 genes.
- Applied the CS-based classification to gene expression data for leukemia subtyping.
Main Results:
- Achieved 97.4% classification accuracy using cross-validation on selected genes.
- Attained 94.3% accuracy on an independent dataset.
- Demonstrated robustness of the CS method to noise, maintaining good performance.
Conclusions:
- The CS-based method effectively detects leukemia subtypes through gene expression analysis.
- This approach offers potential for improved accuracy in leukemia diagnosis.
- The findings suggest CS is a viable tool for personalized cancer therapy development.
