Fig. 4From: DNA-based molecular classifiers for the profiling of gene expression signaturesProfiling of the miRNA expression levels for the diagnosis of HCC. A In silico training of a minimal linear classifier to discriminate HCC from healthy individuals. Performance of the classifier with data from the training dataset, 94.7% and 95.6% of HCC from healthy individuals were classified correctly, with an AUC of 0.9904. B Performance of the classifier with data from the validation dataset, 96.0% and 94.2% of HCC from healthy individuals were diagnosed correctly, with an AUC of 0.9871. C Minimal linear classifier that included five miRNAs with weights ranging from  − 2.6 to 2.4. D Results of DNA-based molecular classifier for 10 cases each of HCC and healthy individuals use synthetic miRNAsBack to article page