Cell type discrimination. Cell types are discriminated by estimating the probability of each type. Normally and always, the biomarkers for discriminating a pair of cell types (e.g., ESCs versus iPSCs) are used to discriminate cell types. After running the biomarkers from the start point to the end point, the probability of cell types will be summarized and a cell type will be assigned to the sample in basis of the higher probability. (a) A simulated sample with 98% probability of iPSCs and only 2% of ESCs, so the sample is discriminated as iPSCs. (b) A real iPSC sample with 100% probability, and no misclassification was (0% ESCs) found in this sample.