Extended Data Fig. 2: False positive detection by the four BulkECexplorer classification methods.

The false positive rate for each of the 1st run actual negative markers (n = 109) is shown for each of the classification methods and across all eligible datasets in the BulkECexplorer (transcript level >0 TPM, transcript level >1 TPM, GMM, zTPM). Gene names are shown adjacent to the Y axis alongside their median TPM.