Matching Gene Activity with Physiological Functions

• Published in: Proceedings of the National Academy of Sciences - PNAS 2002-03, Vol.99 (5), p.2603-2608
• Main Authors: Huang, Wei, Sher, Yuh-Pyng, Peck, Konan, Yuan Cheng B. Fung
• Format: Article
• Language: English
• Subjects: Animals; Biological Sciences; biomechanics; Blood Pressure; Blood vessels; Body tissues; Chromosome Mapping; Complementary DNA; Correlation coefficients; Correlations; Elasticity; Genes; Genes - physiology; Genetics; Lungs; Oligonucleotide Array Sequence Analysis; Oxygen; Oxygen - analysis; Physical Sciences; Physiology; Pulmonary artery; Pulmonary Artery - anatomy & histology; Pulmonary Artery - physiology; Pulmonary Gas Exchange; Rats; Renovations; Tunica Intima - anatomy & histology
• ISSN: 0027-8424; 1091-6490
• DOI: 10.1073/pnas.042684399

Matching the activity of the genes with biomechanics and physiology is an effective way to use cDNA microarray technology. Required are data on the change of activities of genes associated with specific physiological functions with respect to a continuous variable such as time. For each pair of data (gene and physiological function) as functions of time, we can compute a coefficient of correlation, R. The correlation is perfect if R is +1 or -1; it is nonexistent if R = 0. By evaluating R for every gene in a microarray, we can arrange the genes in the order of the number R, thus learning which genes are best correlated with the mechanical or physiological function. We illustrate this procedure by studying the blood vessels in the lung in response to pulmonary hypoxic hypertension, including the remodeling of vascular morphometry, the elastic moduli, and the zero-stress state of the vessel wall. For each physiological function, we identify the top genes that correlate the best. We found that different genes correlate best with a given function in large and small arteries, and that the genes in pulmonary veins which respond to arterial functions are different from those in pulmonary arteries. We found one set of genes matching the remodeling of arterial wall thickness, but another set of genes whose integral of activity over time best fit the wall thickness change. Our method can be used to study other thought-provoking problems.