Description:
Regression model to predict inhibitors of dihydrofolate reductase (DHFR). The model was built with the Multiple Linear Regression technique by using a total of 9 QuBiLS-MAS descriptors.
Training and testing datasets:
A total of 237 training compounds and 124 testing compounds were extracted from the Sutherland, et al., 10.1021/jm0497141
Internal performance:
For a 10-fold cross-validation repeated 100 times: Squared R = 0.7648, MAE = 0.4506, and RMSE = 0.6142.
External performance:
Squared R = 0.7641, MAE = 0.5288, and RMSE = 0.6621.
Regression equation:
pLC50 =
0.0355 * AC[2]_K_Q_BB_nCi_2_MP5_X_LGP[3]_psa_MAS + 1375.445 * ES_N3_Q_BB_nCi_2_MP10_A_LGP[7-8]_a_MAS + 0.7885 * ES_S_Q_AB_nCi_2_MP13_X_KA_dc3_MAS + 0.2569 * IB_N2_B_BB_nCi_2_SS1_M_NSRW_v-est_MAS + -0.2248 * TS[1]_K_Q_BB_nCi_2_MP5_P_LGP[4]_li_MAS + -251.9427 * TS[3]_I50_F_BB_nCi_2_MP8_H_T_LGP[6]_alk_MAS + 10.4064 * TS[4]_P2_Q_BB_nCi_2_SS9_D_LGP[8]_alk_MAS + 7.2157 * TS[4]_V_F_BB_nCi_2_SS5_P_LGP[3-8]_r_MAS + 0.1621 * TS[6]_S_F_BB_nCi_2_NS4_X_LGP[3-5]_m_MAS + 2.1613