Skip to main content

Table 6 Performance of the models built on ChEMBL_22 and ChEMBL_23 data

From: Large scale comparison of QSAR and conformal prediction methods and their applications in drug discovery

MethodModel dataPrediction setModel targetsSensitivitySpecificityCCR
QSARChEMBL_23ChEMBL_242820.630.800.71
MCP_inclChEMBL_242820.630.670.65
MCP_exclChEMBL_242820.560.610.58
QSARChEMBL_22ChEMBL_232820.640.840.74
ChEMBL_242820.600.810.71
MCP_inclChEMBL_232820.610.720.66
ChEMBL_242820.610.670.64
MCP_exclChEMBL_232820.560.680.62
ChEMBL_242820.540.610.57
  1. MCP model results are given at 80% confidence level. MCP_incl and MCP_excl indicate the ‘both’ prediction class was included in the result calculation or was ignored, respectively