3D Molecular Representations Based on the Wave Transform for Convolutional Neural Networks
| Autor: | Alexander Zhebrak, Denis Kuzminykh, Alex Zhavoronkov, Rim Shayakhmetov, Ivan Baskov, Sergey I. Nikolenko, Daniil Polykovskiy, Artur Kadurin |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
0301 basic medicine
Databases Factual Computer science Computer Science::Neural and Evolutionary Computation Gaussian blur Pharmaceutical Science computer.software_genre 01 natural sciences Convolutional neural network 03 medical and health sciences symbols.namesake Wavelet Voxel Drug Discovery Animals Humans Representation (mathematics) business.industry Spatial structure Fingerprint (computing) Pattern recognition 0104 chemical sciences 010404 medicinal & biomolecular chemistry 030104 developmental biology Transformation (function) symbols Molecular Medicine Artificial intelligence Neural Networks Computer business computer |
| Zdroj: | Molecular pharmaceutics. 15(10) |
| ISSN: | 1543-8392 |
| Popis: | Convolutional neural networks (CNN) have been successfully used to handle three-dimensional data and are a natural match for data with spatial structure such as 3D molecular structures. However, a direct 3D representation of a molecule with atoms localized at voxels is too sparse, which leads to poor performance of the CNNs. In this work, we present a novel approach where atoms are extended to fill other nearby voxels with a transformation based on the wave transform. Experimenting on 4.5 million molecules from the Zinc database, we show that our proposed representation leads to better performance of CNN-based autoencoders than either the voxel-based representation or the previously used Gaussian blur of atoms and then successfully apply the new representation to classification tasks such as MACCS fingerprint prediction. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|