Fig 1.
Procedure of the Sort Difference Backward Elimination (SDBE) algorithm.
Fig 2.
Gene selection procedure in the GSEA–SDBE method.
Fig 3.
Volcano plot of differentially expressed genes.
The red and blue dots represent upregulated and downregulated genes, respectively.
Fig 4.
Genes sorted by importance in descending order.
Fig 5.
Enrichment plots for the four gene sets (pathways) that were strongly related to breast cancer.
Table 1.
Gene sets (pathways) that were strongly related to breast cancer.
Fig 6.
Polylines of classification metrics, MCC, and ROC_AUC_score in 19 iterations.
(a) MCC as the object of difference analysis. (b) ROC_AUC_score as the object of difference analysis.
Fig 7.
Polylines of classification metrics at the 19th iteration of the Sort Difference Backward Elimination (SDBE) algorithm.
(a) MCC as the object of difference analysis. (b) ROC_AUC_score as the object of difference analysis. Various metric lists from stage 1 of the algorithm were illustrated by red polylines (RF_improtance).
Table 2.
MCC as the object of difference analysis: 10-fold cross-validation classification metrics of the top three genes.
Table 3.
ROC_AUC_score as the object of difference analysis: 10-fold cross-validation classification metrics of the top three genes.
Fig 8.
Kaplan–Meier survival graphs for expression of VEGFD, TSLP, and PKMYT1.
Red and blue curves denote high-risk and low-risk groups, respectively.
Table 4.
Results of survival analysis for high-risk and low-risk groups according to three genes.
Table 5.
Information on the datasets used for performance comparison.
Table 6.
Classification metrics (%) of four optimization algorithms for five cancer datasets.