Fig. 2
From: A novel multi-tissue RNA diagnostic of healthy ageing relates to cognitive health status
ROC curves showing predictive performance of the healthy ageing classifier based on LOOCV (kNN = 5) for muscle, brain, and skin. Using only the 150 probe-sets identified in the first stage of the project, this ‘healthy ageing classifier’ was able to correctly classify young and old samples across independent data sets with an accuracy of ~96 %, 91 %, 85 %, and 78 %. We present two examples of independent muscle data [48, 50] and one example each for human brain [50] and skin data [11] with areas under the curve of 0.99, 0.94, 0.78, and 0.85, respectively, reflecting excellent separation of the age groups and hence accurate multi-tissue performance