Skip to main content
. 2018 Mar 6;11:73. doi: 10.3389/fnmol.2018.00073

Figure 1.

Figure 1

Variable Importance (Gini) for the top 30 predictor genes. Gini variable importance measures reflect the mean decrease in impurity by splits of a given variable in the classification tree, weighted by the proportion of samples reaching that node. A greater “mean decrease Gini” indicates that the gene plays a greater role in partitioning the data into the defined classes.