Skip to main content
Fig. 4 | BMC Genomics

Fig. 4

From: Molecular signature comprising 11 platelet-genes enables accurate blood-based diagnosis of NSCLC

Fig. 4

Performances of three independent classifiers on early-stage vs. healthy samples, MI samples and on RT-qPCR data. (a) AUC (Area under the curve) plot representing the performances of three independent classifiers i.e. Gradient Boosting Machines (GB), Random Forest (RF), and Linear Discriminant Analysis (LDA) in distinguishing tumor and healthy samples using Δ Ct values of 11 genes from 10 NSCLC patients and 7 healthy controls. (b) AUC plot depicting the improvement in the classification accuracy by augmenting the data-points with artificial samples, using the EigenSample technique. (c) Classification performance based on the proposed 11 gene-panel the on TEP profiles of non-metastatic NSCLC patients and healthy controls from [15]. (d) Classifier performances on experimental data of 10 NSLC and 7 healthy samples. e Receiver Operating Characteristics (ROC) plot depicting the performances of three independent classifiers in distinguishing healthy and myocardial infarction episode samples using normalized intensity from platelets microarray dataset [21]

Back to article page