We then applied QuantiSNP to each dataset to detect chromosomal aberrations. As the data is generated from samples of a normal individual, any detected aberrations will be false positive events. For various settings of the algorithm (different L and Bayes Factor thresholds), we then counted the number of false positive events. In this manner we are able to automatically define a Bayes Factor threshold and prior setting, , which maximizes power for a given Type I error rate. (Further details in Figure 2.)
