Skip to main content

Table 7 Statistical parameters for the best models for 2–6 variables for the physicochemical property log K, considering the 31 structures as the training set

From: QuBiLS-MAS, open source multi-platform software for atom- and bond-based topological (2D) and chiral (2.5D) algebraic molecular descriptors computations

Size R 2 Q 2loo Q 2boot a (Q 2) F Models Equations
2 0.778 0.734 0.738 −0.208 49.16 log K = 1.596 (±0.885) + 3.809 (±0.582) (19)
       TS[1]_MX_B_AB_nCi_2_SS12_T_KA_a-h − 0.118 (±0.011)  
       KH[1]_MX_F_AB_nCi_2_MP2_T_KA_h  
3 0.863 0.826 0.820 −0.259 57.14 log K = −32.132 (±3.841) − 75.624 (±9.789) (20)
       TS[1]_RA_F_AB_nCi_2_MP2_T_KA_h + 135.484 (±13.179  
       TS[4]_PN_Q_AB_nCi_2_MP0_T_KA_h + 1782.101 (±257.835)  
       KH[2]_PN_B_AB_nCi_2_SS8_T_KA_v-h  
4 0.915 0.887 0.879 −0.324 70.59 log K = −66.472 (±6.939) − 0.223 ± 0.021) (21)
       AC[2]_MX_B_AB_nCi_2_SS7_T_KA_r-h + 0.407 (±0.089)  
       TS[5]_HM_B_AB_nCi_2_SS8_T_KA_v-h + 131.848 (±10.928)  
       TS[4]_PN_Q_AB_nCi_2_MP0_T_KA_h + 3323.451 (±355.509)  
       KH[2]_PN_B_AB_nCi_2_SS8_T_KA_v-h  
5 0.932 0.902 0.890 −0.376 68.53 log K = −70.522 (±6.342) − 0.246 (±0.020) (22)
       AC[2]_MX_B_AB_nCi_2_SS7_T_KA_r-h + 0.422 (±0.081)  
       TS[5]_HM_B_AB_nCi_2_SS8_T_KA_v-h + 144.507 (±9.991)  
       TS[4]_PN_Q_AB_nCi_2_MP0_T_KA_h + 4616.536 (±15.439)  
       GV[2]_MX_Q_AB_nCi_2_MP3_X_KA_h + 3536.215 (±324.863)  
       KH[2]_PN_B_AB_nCi_2_SS8_T_KA_v-h  
6 0.942 (0.960)a 0.914 (0.937)a 0.898 (0.925)a −0.414 (−0.465)a 65.26 (91.74)a log K = −81.005 (±6.216) − 0.233 (±0.020) (23)
       AC[2]_MX_B_AB_nCi_2_SS7_T_KA_r-h − 39,144.250 (±4.757)  
       AC[2]_MN_B_AB_nCi_2_MP2_A_KA_c-h + 0.572 (±17.485)  
       TS[5]_HM_B_AB_nCi_2_SS8_T_KA_v-h + 120.683 (±1.681)  
       TS[4]_PN_Q_AB_nCi_2_MP0_T_KA_h + 0.804 (±0.354)  
       TS[6]_HM_Q_AB_nCi_2_SS0_A_KA_h + 3979.089 (±310.376)  
       KH[2]_PN_B_AB_nCi_2_SS8_T_KA_v-h  
  1. aCompound 31 excluded, taken as outlier, is not taken into account in the training set