STATISTICAL-ANALYSIS OF HIGHLY SKEWED IMMUNE-RESPONSE DATA

This paper considers methods of statistical analysis for highly skewed immune response data. Observations from population studies of immunol ogical variables are rarely normally distributed between individuals; typically the distribution shows extreme levels of skewness. In some s ituations, skewness remains considerable even after transforming the d ata. Using resampling techniques, applied to several actual datasets o f ELISA assay data, we consider the robustness of normal parametric me thods, e.g. t tests and linear regression. Despite the skewness of the transformed data, we demonstrate that such methods are quite robust d epending on the number of observations, type of analysis and severity of skewness. We also illustrate how bootstrap resampling can be used t o provide a valid alternative method of analysis that can be used eith er for checking normal parametric analysis or as a direct method of an alysis. We illustrate this combined approach by analysing real data to test for association between human serum antibodies to malaria merozo ite surface proteins, MSP1 and MSP2, and resistance to clinical malari a, and confirm the protective effect of antibodies to MSP1 and demonst rated a similar protective effect for some antibodies to MSP2.