PAULYLSER Solute Descriptors
GNNGraph-Convolutional Neural Network
Graph-convolutional neural net for the prediction of LSER solute descriptors
For the solute descriptors A, B, E, S, L individual neural nets were trained with 2 hidden layers including 32 and 64 neurons.
For training of the models the curated dataset of Ulrich et al. was used, the number of chemicals varies for the different solute descriptors.
Number of training set chemicals varies for each descriptor.
Data augmentation strategy was applied to enlarge the number of input features for training of the neural net.