Related Experiment Video
Updated: Jun 12, 2026
03:08
Using Human Differentially Expressed Gene Lists to Perform Downstream Pathway Enrichment Analysis and Target Prioritization
Published on: October 3, 2025
Predictive response-relevant clustering of expression data provides insights into disease processes
Lisa E M Hopcroft1, Martin W McBride, Keith J Harris
1Inference Group, Department of Computing Science, University of Glasgow, and Gartnavel General Hospital, 1053 Great Western Road, Glasgow G12 0YN, UK.
Nucleic Acids Research
|June 24, 2010
Summary
This study introduces a novel microarray data analysis method, coupling clustering and classification to identify response-relevant genes. The approach reveals a gene cluster protective against hypertension, offering biological insights into salt-sensitive conditions.
Area of Science:
- Bioinformatics
- Genomics
- Systems Biology
Background:
- Microarray data analysis requires robust methods for identifying biologically relevant genes.
- Distinguishing informative genes for specific responses is crucial for understanding complex diseases.
Purpose of the Study:
- To present a novel microarray data analysis method combining model-based clustering and binary classification.
- To identify `response-relevant' genes and create a `meta-covariate' representation for predictive modeling.
- To apply and validate the method on leukemia and hypertension datasets, exploring biological insights.
Main Methods:
- Coupling model-based clustering and binary classification to form gene clusters.
- Utilizing a `meta-covariate' representation of gene clusters in a probit regression model.
- Applying the method to leukemia and rat renal gene expression datasets.
Main Results:
- Identification of a 13-gene cluster, including transcription factors (Arntl, Bhlhe41, Npas2), protective against hypertension.
- Functional analysis implicated transcriptional activation and circadian rhythm signaling pathways.
- Demonstrated applicability to high-dimensional datasets beyond expression data.
Conclusions:
- The novel method effectively identifies response-relevant genes and provides biological insights.
- The identified gene cluster offers potential protective mechanisms against hypertension.
- The meta-covariate approach is versatile and applicable to various high-dimensional biological data.
