A novel approach to detect differentially expressed genes from count-based digital databases by normalizing with...

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Gene expression analysis using sequence tag counts is crucial for identifying cancer-related genes.
• Computational methods for differential gene expression, often using Binomial or beta-Binomial distributions, are vital in cancer biology.
• Selecting an appropriate statistical model is a persistent challenge in these analyses.

Purpose of the Study:

• To develop a novel Bayesian algorithm-based method for differential gene expression analysis.
• To introduce the Electronic Differential Gene Expression Screener (EDGES) for identifying candidate genes in cancer.
• To establish a robust statistical model for count-based gene expression data.

Main Methods:

• Developed EDGES, a Bayesian algorithm utilizing geometric averaging of 12 common housekeeping genes for statistical model determination.
• Applied EDGES to publically available Serial Analysis of Gene Expression (SAGE) and Expressed Sequence Tag (EST) data.
• Validated EDGES using gene expression microarray analysis and quantitative reverse transcription real-time PCR.

Main Results:

• EDGES successfully identified differentially expressed genes in lung, breast, and colorectal cancers.
• The method demonstrated effectiveness in pinpointing genes relevant to the cancerous phenotype.
• Experimental validation confirmed the reliability of the identified gene sets.

Conclusions:

• The developed EDGES method offers a novel approach to differential gene expression screening in cancer research.
• Normalization of calibrators provides new insights into count-based digital subtraction for cancer studies.
• This work highlights the importance of robust statistical modeling in analyzing high-throughput gene expression data for cancer biology.