Building QSAR models using shannon entropy and a genetic algorithm
Jrg K. Wegner, Andreas Zell · 2003
For the analysis of the descriptor selection method we worked with the Huuskonen solubility data set with logS results. We used a partitioning of the data set analogue to the work of Liu and So. Additionally we used the Wang dataset to predict the octanol/water partition coefficient logP. The descriptors were calculated with MOE (especially surface descriptors), JOELib(1) (especially atom property descriptors and derivatives) and MolConnZ (topological descriptors). d 1 Abstract d 3