The present invention is hydrogen (H), carbon (C), nitrogen (N), oxygen (O) , sulfur (S), such as being composed of five kinds of elements within the octane number of atoms other than hydrogen in pure organic compound consisting of not more than 25 molecules all-water partition coefficient to predict (Octanol-Water Partition Coefficient) with high accuracy provides a mathematical model. The model is, octanol - for a number of the organic compound satisfying the above conditions, the experimental values of the water distribution coefficient is known, any of a variety of molecules presenter (molecular descriptor) as independent variables, the octanol-water distribution Many of the best multiple linear regression model in which the dependent variable coefficients obtained after using a genetic algorithm (genetic algorithm), receives the value of the molecules presenters included with this model octanol-water partition coefficient output that ANN (artificial neural network) was further improved by constructing the predictive performance of multiple linear regression - a hybrid artificial neural network model (hybrid model), is one example of a QSPR model, if you know the specific values of the molecules presenters included in the model any molecule that somehow, octanol of pure compound consisting of a molecule-water partition coefficient gives predictions. As such, the present invention provides for the reliable octanol organic compounds in a number of the experimental condition of unknown - by giving to provide a way to predict the value of the distribution coefficient can maintain water saving the cost and time to experiment, and related industries lays effects such as the ease of research and development activities. ; 展开▼
