IMAGE METRICS IN THE STATISTICAL ANALYSIS OF DNA MICROARRAY DATA

摘要

Expression profiling using DNA microarrays is an important new method foranalyzing cellular physiology. In"spotted" microarrays, fluorescently labeled cDNA from experimental andcontrol cells is hybridized to arrayed target DNA and thearrays imaged at two or more wavelengths. Statistical is performed onmicroarray images and show that non-additive background,high intensity fluctuations across spots, and fabrication artifacts interferewith the accurate determination of intensity information.The probability density distributions generated by pixel-by-pixel analysis ofimages can be used to measure the precision with whichspot intensities are determined. Simple weighting schemes based on theseprobability distributions are effective in improvingsignificantly the quality of microarray data as it accumulates in amultiexperiment database. Error estimates from image-based metricsshould be one component in an explicity probabilistic scheme for the analysisof DNA microarray data.