Identification of Cancer Related Genes Using a Comprehensive Map of Human Gene Expression

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Public gene expression datasets facilitate large-scale meta-analyses and biological discoveries, including novel cancer gene identification.
• The Affymetrix HG-U133Plus2 platform is widely used for gene expression profiling.

Purpose of the Study:

• To compile and systematically annotate the largest human gene expression dataset from public repositories.
• To identify novel cancer gene candidates by analyzing gene expression changes in solid tissue tumors compared to healthy tissues.

Main Methods:

• Compiled and quality-controlled gene expression data from approximately 40,000 Affymetrix HG-U133Plus2 arrays.
• Systematically curated and categorized sample annotations using ontology assistance.
• Applied unsupervised analysis and a mixed-effects linear model to identify differentially expressed genes in cancer.

Main Results:

• Generated a comprehensive gene expression matrix of ~20,000 genes and ~28,000 samples.
• Identified 1,285 genes with systematic expression changes in cancer.
• The identified gene list is enriched with known cancer genes and includes novel candidates.

Conclusions:

• The compiled dataset represents the largest systematically annotated human gene expression resource publicly available.
• This resource enables detailed investigation of gene expression patterns across diverse biological samples, aiding cancer research.
• The study proposes a list of novel cancer gene candidates for further validation and therapeutic development.