Skip to main content
. 2016 Mar 14;4:e1804. doi: 10.7717/peerj.1804

Figure 1. Pipeline for the development and evaluation of the ColoFinder.

Figure 1

Gene expression data are chosen, normalized, and z-score transformed. A K-fold cross-validation is used for the development of the random survival forest (RSF) prognosis model (K = 5). The 9-genes signature was selected to develop the RSF prognosis model. A fully-developed model based on the optimal parameters identified in cross-validation was then evaluated in an independent dataset.