论文部分内容阅读
Background: Spatiotemporal variation of gene expression can happen extensively.Thanks to current applications of high throughput technologies, e.g.microarray, SAGE and next generation sequencing, gene differential expressions could be motored simultaneously in large scale.Methods: Here, we developed a method to identify housekeeping genes, tissue-specific/ selective genes and tissue-repressed genes using three novel statistical parameters: Specificity Measure (SPM), Contribution Measure (CTM) and Dispersion Measure (DPM).Accordingly, a number of pattern genes were thus identified from several microarray datasets.Upon these pattern genes, a systematic and genome-wide exploration of these pattern genes were demonstrated in various aspects of chromosomal locus, sub-cellular distribution, physiological function, and association with tissue functions.Results: It is observed that many housekeeping genes are non-evenly expressed across tissues in normal physiological state or under experimental conditions.They tend to lie in intracellular organelles, especially in nucleus and ribosome, and their products function in aspects of nucleic acid binding, nucleic acid metabolic process, transcription and translation.To the contrast, products of tissue-specific genes are normally extracellular, e.g.on plasma membrane or in plasma, in the form of secreted proteins or receptors.They often play roles in receptor binding, homeostasis, organ development and signal transduction.Moreover, tissue selective genes may serve as a bridge between tissues.They even provide clues of tissue development.Conclusions: In summary, our study provides an easier, more sensitive and robust way in quantitative detection of gene expression patterns and further better understanding of tissue function and development .