Fig. 4
From: GECKO is a genetic algorithm to classify and explore high throughput sequencing data
GECKO voting mode for small sample sizes. a GECKO’s voting mode will run 10 separate genetic algorithms with added Gaussian noise. The best solutions of these runs will be fed into a final genetic algorithm to produce a final solution. b GECKO output showing the t-SNE separation of patients with complete response to chemotherapy from those that did not using five k-mers from the winning individual. Triangles correspond to the test dataset that was excluded from GECKO training can thus be used to estimate overfitting
