Fig. 3From: Using published pathway figures in enrichment analysis and machine learningGene importance from random forest breast cancer models built on pathway genes. Represented as circles, the gene content is sized and ordered by importance (large-to-small, bottom-to-top) for the top twenty-one pathways ordered by prediction accuracy (left-to-right). The “Top Gene” from Table 3 is represented by the largest, bottom-most circle and is labeled for each pathway. The vast majority of circles (gray) denote genes found only in PFOCR pathways among all significant pathways with “important” genes. The genes that are also present within Reactome and KEGG results are highlighted in red and blue, respectively. There were no significant pathways with gene importance scores above 0.1 from WikiPathwaysBack to article page