dc.contributor.author |
Kuzminykh D. |
|
dc.contributor.author |
Polykovskiy D. |
|
dc.contributor.author |
Kadurin A. |
|
dc.contributor.author |
Zhebrak A. |
|
dc.contributor.author |
Baskov I. |
|
dc.contributor.author |
Nikolenko S. |
|
dc.contributor.author |
Shayakhmetov R. |
|
dc.contributor.author |
Zhavoronkov A. |
|
dc.date.accessioned |
2019-01-22T20:46:15Z |
|
dc.date.available |
2019-01-22T20:46:15Z |
|
dc.date.issued |
2018 |
|
dc.identifier.issn |
1543-8384 |
|
dc.identifier.uri |
https://dspace.kpfu.ru/xmlui/handle/net/148683 |
|
dc.description.abstract |
© 2018 American Chemical Society. 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. |
|
dc.relation.ispartofseries |
Molecular Pharmaceutics |
|
dc.subject |
3D convolutional neural networks |
|
dc.subject |
autoencoders |
|
dc.subject |
wave transform |
|
dc.subject |
wavelets |
|
dc.title |
3D Molecular Representations Based on the Wave Transform for Convolutional Neural Networks |
|
dc.type |
Article |
|
dc.relation.ispartofseries-issue |
10 |
|
dc.relation.ispartofseries-volume |
15 |
|
dc.collection |
Публикации сотрудников КФУ |
|
dc.relation.startpage |
4378 |
|
dc.source.id |
SCOPUS15438384-2018-15-10-SID85052122681 |
|