Abraham’s data set

For modeling the BBB penetration, authors used Abraham s data set of 57 compounds as the training set. The test set consisted of 13 compounds, 7 of which were taken from Abraham s data set and 6 from the data set of Lombardo and workers. In addition to the lipoaffinity descriptor, the other descriptors used by them include molecular weight and TPSA. Two models were developed one based on stepwise MLR and the other one based on ANN. To test the performance of different descriptors, they first carried out a simple LR of the 55 training set compounds (two outliers were removed) using TPSA as the only descriptor (Eq. 41). The equation was comparable to Clark s model (Eq. 33). 

Abraham s data set of 57 compounds was selected as training set for log BB prediction. The test set contained the 13 compounds used by Clark and Liu et al. A three-component model was built from the atom type descriptors, and it estimated the data set of 57 compounds with an r2 = 0.897, q2 = 0.504, and RMSEE = 0.259. The relatively lower q2 resulted from the small size of the data set. Totally, 94 different atom types were identified for the 57 compounds, and half of these atom types occurred only once or twice through the whole data set. When the compounds containing these atom types were left out in cross-validation, the contribution of these atom types could not be predicted accurately, since they did not appear in the training set. After... 

Lombardo et al. [29] Basak et al. [5] Kaliszan and Markuszewski [23] Young s data set + Abraham s data set New compounds Young s data set + Abraham s data set... 

Kaznessis et al. [24] Young s data set + Abraham s data set + new compounds... 

Abraham et al. [2] published a number of papers in which they analyzed Young s data set using MLR and gave a general solvation equation in which various solvent-solute interactions were described by solute descriptors and equation coefficients (Eq. 16)... 

Zhao and coworkers [53] also constructed a linear model using the Abraham descriptors. The MLR model possesses good correlation and predictability for external data sets. In this equation, E is an excess molar refraction (cm3/mol/ 10.0) and S the dipolarity/polarizability, A and B are the hydrogen bond acidity and basicity, respectively, and V is the McGowan characteristic volume (cm3/ mol/100). The large coefficients of A and B indicate too polar molecules having poor absorption. 

Abraham, M.H., Whiting, G.S., Doherty, R.M. and Shuely, W.J. (1990b). Hydrogen Bonding. Part 13. A New Method for the Characterization of GLC Stationary Phases - The Laffort Data Set. J.Chem.Soc.Perkin Trans.2,1451-1460. 

Our group elected to use Abraham s LFER model as our base equation becanse it is representative of the dermal QSPR approaches presently available (Abraham and Martins, 2004). Preliminary analyses applying 16 different LFER equations reviewed by Geinoz et al. (2004) to the entire PSFT (288 treatment combinations) and IPPSF data set (32 treatment combinations) demonstrated a superior fit of our data set to the Abraham equation compared to most other models reviewed. It must be stressed that the purpose of this research was not to identify the optimal LFER for predicting dermal permeation or to validate that this model is predictive of dermal absorption. Rather, we selected this model because it best described the available data and is widely accepted by the scientific community. 

In an excellent paper, Zhao et al. [29] assembled a carefully reviewed literature set of human absorption data on 241 drugs. They showed that a linear regression model built with 5 Abraham descriptors could fit percent human absorption data reasonably well (r2 = 0.83, RMSE = 14%). The descriptors are excess molar refraction (E), polarizability (S), hydrogen bond acidity (A), hydrogen bond basicity (B), and McGowan volume (V), all related to lipophilicity, hydrophilicity, and size. In a follow-on paper, data on rat absorption for 151 drugs was collected from the literature and modeled using the Abraham descriptors [30]. A model with only descriptors A and B had r2 = 0.66, RMSE = 15%. 

---