Premium
Front Cover: Rotational Variance‐Based Data Augmentation in 3D Graph Convolutional Network (Chem. Asian J. 18/2021)
Author(s) -
Kim Jihoo,
Kim Yeji,
Lee Eok Kyun,
Chae Chong Hak,
Lee Kwangho,
Kim Won June,
Choi Insung S.
Publication year - 2021
Publication title -
chemistry – an asian journal
Language(s) - English
Resource type - Reports
SCImago Journal Rank - 1.18
H-Index - 106
eISSN - 1861-471X
pISSN - 1861-4728
DOI - 10.1002/asia.202100918
Subject(s) - cartesian coordinate system , graph , orientation (vector space) , artificial intelligence , computer science , identification (biology) , cover (algebra) , pattern recognition (psychology) , computer vision , mathematics , theoretical computer science , biology , geometry , engineering , mechanical engineering , botany
Data augmentation for 3D graph convolutional network (3DGCN) has been achieved, for the classification task of human β ‐secretase 1 (BACE‐1)‐ligand binding, by randomly rotating known active (1‐labeled) ligands in the 3D Cartesian coordinate and assigning the rotated ones as label‐0. The data augmentation improves the orientation‐recognizing ability of 3DGCN significantly, suggesting the potential for identification of active ligands from a dataset, with rotation of candidates and evaluation of output sigmoid values. More information can be found in the Communication by Won June Kim, Insung S. Choi et al.