20200303 SUN Immunology Luminex Data Analysis скачать в хорошем качестве

20200303 SUN Immunology Luminex Data Analysis

This lecture was delivered for the B.Sc. Honours students at the Division of Molecular Biology and Human Genetics at Stellenbosch University. The PDF of slides can be found at this link: https://drive.google.com/file/d/1Y1Tw... Luminex assays are a powerful way to parallelize many ELISA immunoassays in a single experiment. Analyzing the data from Luminex can, however, encounter several challenges. This lecture explains some of the fundamentals for establishing a calibration curve for relation fluorescent intensity to analyte concentration, for handling missingness in values that are below the limit of detection or outside the levels of quantitation, and for normalizing data distributions that contain far-flung outliers. Along the way we contrasted interpolation and extrapolation. We compared the "missing completely at random," "missing at random," and "missing not at random" error models and evaluated some simple and complex strategies for imputing these values. We discussed commonly used methods for normalizing data, such as log normalization and related functions, but we also covered Winsorization, a strategy that particularly curbs the bad effects of "lost sheep" data points on either side of the distribution.