THE USE OF CONFIDENCE REGIONS TO DETERMINE OBJECTIVE CRITERIA FOR ANALYSIS OF SINGLE PARAMETER HISTOGRAMS OF MEAN FLUORESCENCE INTENSITY by

Several methods have been proposed for comparing fluorescence intensity curves obtained from flow cytometry analysis. The one comparison method that is most often used is the Kolmogorov-Smirnov (K-S) test, which is known to find statistically significant differences between curves even when clinical or scientific differences are doubtful. In the field of flow cytometry, the combination of the gating process and either Overton Subtraction or K-S test yields a histogram subtraction technique for analysis of flow cytometric frequency data. In order to help biologists assess the results of a K-S or other test for the difference between mean fluorescence intensity curves, we have proposed a method of producing bootstrap-based confidence bands around the difference in histograms to show the range of intensity channels for which a difference between histograms may be identified. As an illustration of the method, we obtain and display confidence bands for histogram differences between pumping cell subsets. The confidence bands are constructed using both pointwise and simultaneous confidence bands. We compare the results of these two approaches. Furthermore, for descriptive purposes, the confidence bands produced were found to provide a useful descriptive display to aid in the interpretation of results. Flow cytometry is a major research tool for immunology and, by John. W. Wilson, PhD THE USE OF CONFIDENCE REGIONS TO DETERMINE OBJECTIVE CRITERIA FOR ANALYSIS OF SINGLE PARAMETER HISTOGRAMS OF MEAN FLUORESCENCE INTENSITY Dongha Kim, MS University of Pittsburgh, 2015