Skip to main content

Advertisement

Fig. 3 | BMC Medicine

Fig. 3

From: Overcoming intratumoural heterogeneity for reproducible molecular risk stratification: a case study in advanced kidney cancer

Fig. 3

Kaplan-Meier curves for overall survival risk stratification by N-Cadherin, EPCAM, Age and mTOR (NEAT). a Validation cohort. The high risk (n = 4, dashed line) and low risk (n = 18, dotted line) groups identified by NEAT have markedly different prognoses (log-rank p = 7.62 × 10−7) with respective 2-year survival rates of 0% and 78% (precision = 100%, recall = 50%, specificity = 100%, accuracy = 82%, Matthews correlation coefficient = 0.62). Data analysed were independent of wrapper feature selection and of fitting model coefficients. b Development cohort. The identified features and model coefficients were learned on the data shown, which therefore does not provide an independent test. High risk (n = 14, dashed line) and low risk (n = 8, dotted line) groups are clearly separated (log-rank p = 0.00553), with respective 2-year survival rates of 43% and 100%. (precision = 57%, recall = 100%, specificity = 57%, accuracy = 73%, Matthews correlation coefficient = 0.57)

Back to article page