Skip to main content

Table 4 Average areas under ROC curves

From: Statistical-based database fingerprint: chemical space dependent representation of compound databases

Dataset

MACCS keys (166-bit)

ECFP4

1-NN

DFP

SB-DFP

1-NN

DFP

SB-DFP

BRD2

0.938 (0.035)

0.875 (0.019)

0.911 (0.031)

0.974 (0.023)

0.865 (0.037)

0.970 (0.030)

BRD3

0.940 (0.041)

0.873 (0.015)

0.905 (0.029)

0.962 (0.037)

0.861 (0.056)

0.964 (0.032)

BRD4

0.880 (0.038)

0.821 (0.036)

0.871 (0.040)

0.927 (0.037)

0.740 (0.082)

0.941 (0.026)

CREBBP

0.953 (0.025)

0.924 (0.008)

0.963 (0.009)

0.956 (0.027)

0.913 (0.016)

0.972 (0.020)

DNMT1

0.652 (0.045)

0.652 (0.049)

0.855 (0.037)

0.711 (0.058)

0.484 (0.060)

0.834 (0.042)

EHMT2

0.969 (0.033)

0.897 (0.027)

0.965 (0.023)

0.951 (0.050)

0.860 (0.042)

0.947 (0.036)

EP300

0.874 (0.041)

0.810 (0.052)

0.896 (0.026)

0.843 (0.066)

0.592 (0.076)

0.873 (0.052)

HDAC10

0.932 (0.022)

0.916 (0.043)

0.946 (0.025)

0.934 (0.032)

0.821 (0.063)

0.975 (0.016)

HDAC11

0.939 (0.024)

0.899 (0.073)

0.940 (0.034)

0.948 (0.035)

0.786 (0.065)

0.979 (0.018)

HDAC1

0.797 (0.036)

0.755 (0.085)

0.886 (0.041)

0.884 (0.035)

0.688 (0.073)

0.945 (0.030)

HDAC2

0.847 (0.035)

0.808 (0.081)

0.895 (0.042)

0.905 (0.032)

0.750 (0.048)

0.954 (0.024)

HDAC3

0.875 (0.032)

0.862 (0.059)

0.888 (0.032)

0.892 (0.035)

0.725 (0.062)

0.950 (0.025)

HDAC4

0.841 (0.039)

0.781 (0.067)

0.888 (0.021)

0.890 (0.039)

0.672 (0.060)

0.939 (0.034)

HDAC5

0.866 (0.066)

0.838 (0.036)

0.920 (0.016)

0.917 (0.030)

0.840 (0.049)

0.926 (0.035)

HDAC6

0.828 (0.028)

0.825 (0.042)

0.895 (0.011)

0.868 (0.026)

0.743 (0.072)

0.928 (0.021)

HDAC7

0.907 (0.037)

0.925 (0.037)

0.948 (0.012)

0.913 (0.027)

0.864 (0.020)

0.934 (0.024)

HDAC8

0.878 (0.024)

0.883 (0.054)

0.937 (0.011)

0.896 (0.028)

0.762 (0.043)

0.953 (0.019)

HDAC9

0.901 (0.028)

0.933 (0.031)

0.943 (0.012)

0.942 (0.019)

0.885 (0.026)

0.960 (0.018)

KAT2B

0.926 (0.039)

0.893 (0.022)

0.947 (0.033)

0.935 (0.027)

0.928 (0.022)

0.965 (0.027)

KDM1A

0.745 (0.048)

0.701 (0.051)

0.860 (0.055)

0.885 (0.038)

0.721 (0.058)

0.941 (0.034)

KDM4C

0.677 (0.067)

0.608 (0.069)

0.837 (0.044)

0.653 (0.052)

0.527 (0.045)

0.823 (0.048)

L3MBTL1

0.997 (0.001)

0.999 (0.000)

1.000 (0.000)

1.000 (0.000)

1.000 (0.000)

1.000 (0.000)

L3MBTL3

0.990 (0.003)

0.991 (0.002)

0.991 (0.003)

0.989 (0.005)

0.985 (0.004)

0.990 (0.005)

MAP3K7

0.860 (0.042)

0.791 (0.028)

0.861 (0.027)

0.858 (0.042)

0.738 (0.079)

0.911 (0.035)

MGEA5

0.985 (0.005)

0.985 (0.006)

0.979 (0.009)

0.979 (0.009)

0.996 (0.002)

0.992 (0.007)

NCOA1

0.491 (0.074)

0.572 (0.073)

0.682 (0.056)

0.618 (0.047)

0.519 (0.060)

0.722 (0.044)

NCOA3

0.530 (0.057)

0.577 (0.071)

0.680 (0.064)

0.590 (0.045)

0.503 (0.059)

0.709 (0.043)

PRMT1

0.867 (0.058)

0.673 (0.072)

0.843 (0.078)

0.881 (0.081)

0.365 (0.081)

0.934 (0.037)

Average

0.853 (0.132)

0.824 (0.129)

0.898 (0.082)

0.882 (0.113)

0.755 (0.171)

0.926 (0.077)

  1. The best performing methods for each dataset are shown in bold. If there were no significative difference between two or more methods, all of them are marked. Standard deviations are shown in parentheses