The database dbEST correctly predicts gene expression in colon cancer patients

Area of Science:

• Genomics
• Bioinformatics
• Molecular Biology

Background:

• Gene expression data is crucial for understanding biological processes and disease.
• Databases like dbEST aggregate vast amounts of gene expression information from diverse experiments.
• Comparing data mining predictions with experimental validation is essential for assessing data reliability.

Purpose of the Study:

• To evaluate the predictive accuracy of gene expression data from the NIH's dbEST database.
• To determine if dbEST can facilitate comparisons across different tissue types and experimental conditions.
• To provide a stronger foundation for findings derived from data mining of gene expression profiles.

Main Methods:

• Quantitative RT-PCR was used to measure the expression of nine specific genes and two calibration genes.
• Gene expression was analyzed in paired normal and cancerous colon tissues from eight patients.
• Predictions from the dbEST database were compared against experimental results.

Main Results:

• The expression of nine genes in normal colon tissue was accurately predicted by dbEST data, spanning two orders of magnitude.
• GUS and beta-actin served as reliable internal calibration standards due to low inter-patient variation.
• While less precise in cancer tissues, dbEST predictions remained valuable for clinical correlation.

Conclusions:

• dbEST demonstrates significant predictive power for gene expression in normal colon tissues.
• The database offers a valuable tool for comparative studies across various tissues and conditions.
• dbEST findings hold promise for validating data mining results and aiding cancer research.