Fig. 3From: Hypoxia classifier for transcriptome datasetsSelected decision trees. The labels in each node indicate the node’s class (N, normoxic, and H, hypoxic), the probability of samples in the node to be classified as hypoxic, and the percentage of samples of the training set in each node. Each node is colored according to the probability of samples in said node to be classified as hypoxic. A Overall best performing tree (tree #125). B Best performing tree among those not using MIR210HG (tree #241) C Best performing tree on mouse data (tree #42)Back to article page