Incorporating gene functions as priors in model-based clustering of microarray gene expression data

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Gene expression clustering is vital for gene function discovery.
• Existing methods often overlook known gene functions during clustering.
• Leveraging functional annotations can improve clustering accuracy.

Purpose of the Study:

• To develop a novel model-based clustering approach that integrates prior gene functional information.
• To enhance the accuracy of gene function prediction by utilizing known annotations.

Main Methods:

• Proposed a stratified mixture model incorporating known gene functions as prior probabilities.
• Developed an expectation-maximization (EM) algorithm for model fitting.
• Compared the stratified model with standard model-based clustering.

Main Results:

• The stratified model demonstrated advantages over standard model-based clustering.
• Simulation studies confirmed the effectiveness of the proposed method.
• Application to gene function prediction showed improved performance.

Conclusions:

• Incorporating known gene functions into model-based clustering significantly improves gene function prediction.
• The stratified mixture model offers a powerful approach for analyzing gene expression data.
• This method advances the field of bioinformatics for functional genomics.