CORAL: Building up the model for bioconcentration factor and defining it’s applicability domain

CORAL (CORrelation And Logic) software can be used to build up the quantitative structure – property/activity relationships (QSPR/QSAR) with optimal descriptors calculated with the simplified molecular input line entry system (SMILES). We used CORAL to evaluate the applicability domain of the QSAR models, taking a model of bioconcentration factor (log BCF) as example. This model's based on a large training set of more than 1000 chemicals. To improve the model is predictivity and reliability on new compounds, we introduced a new function, which uses the Delta(obs) = log BCF(expr) − log BCF(calc) of the predictions on the chemicals in the training set. With this approach, outliers are eliminated from the phase of training. This proved useful and increased the model's predictivity.

Root mean square error (RMSE) has been examined as an endpoint in the SMILES-based QSPR analysis of the bioconcentration factor of 1035 substances and was used to calculate the applicability domain and outliers.Figure optionsDownload as PowerPoint slideResearch highlights
► Root mean square error (RMSE) has been examined as an endpoint in the SMILES-based QSPR analysis of the bioconcentration factor of 1035 substances;
► QSPR model of the RMSE obtained for sub-training and calibration sets was used for prediction of this endpoint for test set. The ability of the RMSE to calculate the applicability domain and outliers has been demonstrated;
► Previous QSPR model for bioconcentration factor has been improved.