Protheragen-ING - Molecular Property Prediction

Molecular property prediction , a powerful way to evaluate, select, and generate molecules, is significant for drug discovery and substance discovery. Accurate prediction shortens the time and saves cost to accelerate the overall process of drug discovery. Thus, the key is to find suitable and effective computational methods. Recent studies show that Multilevel Graph Convolutional neural Network, which considers much rule-based information, has a good performance in analyzing the complex relations and structures among a large collection of molecular compounds.