Skip to main content
. 2024 Jan 5;40(1):btae004. doi: 10.1093/bioinformatics/btae004

Table 1.

AUC performance of CircSI-SSL algorithm under different training sample proportions.

Ratio WTAP FXR1 C17ORF85 QKI TAF15 AUF1 AVG
1:9 0.887 0.981 0.964 0.961 0.986 0.971 0.958
2:8 0.914 0.980 0.962 0.964 0.986 0.970 0.963
3:7 0.913 0.984 0.962 0.967 0.987 0.973 0.964
4:6 0.926 0.987 0.963 0.964 0.989 0.972 0.967
5:5 0.938 0.983 0.961 0.968 0.988 0.973 0.968
6:4 0.943 0.985 0.960 0.969 0.991 0.977 0.971
7:3 0.938 0.983 0.959 0.971 0.990 0.975 0.969
8:2 0.978 0.989 0.956 0.970 0.989 0.979 0.977