Skip to main content
. 2024 Dec 2;25(23):12942. doi: 10.3390/ijms252312942

Figure 6.

Figure 6

Comparison of DeepEnhancerPPO with 11 other feature reduction strategies on the independent test dataset for the second enhancer classification task: strength classification, assessed using three performance metrics (ACC, MCC, and AUC). Panels (AC) illustrate the results for different numbers of top sub-features, specifically selecting the top 32, 64, and 128 sub-features, respectively.