A System for Learning Atoms Based on Long Short-Term Memory Recurrent Neural Networks
In recent years, researchers in the fields of bioinformatics and cheminformatics have attempted to utilize machine learning methods for molecule modeling, bioactivity prediction, chemical property prediction, biology analysis, etc. In this paper, we present a system that merges the merits of various techniques such as long short-term memory (LSTM) recurrent neural networks, and is designed for learning atoms and solving the classic problems such as single task classification in the field of drug discovery. We have implemented our approach and conducted extensive experiments based on several widely used datasets such as SIDER and Tox21. The experimental results consistently demonstrate the feasibility and superiority of our proposed approach.
