Janssen Research & Development, Janssen Pharmaceutica N.V., Turnhoutseweg 30, Beerse B-2340, Belgium.
Arcadia Inc., 28 k2, Bolshoy Sampsonievskiy pr., St. Petersburg 194044, Russia.
J Chem Inf Model. 2022 Jul 25;62(14):3307-3315. doi: 10.1021/acs.jcim.2c00344. Epub 2022 Jul 6.
This work introduces , a new algorithm for reaction atom-to-atom mapping (AAM) based on a transformer neural network adopted for the direct processing of molecular graphs as sets of atoms and bonds, as opposed to SMILES/SELFIES sequence-based approaches, in combination with the Bidirectional Encoder Representations from Transformers (BERT) network. The graph transformer serves to extract molecular features that are tied to atoms and bonds. The BERT network is used for chemical transformation learning. In a benchmarking study with IBM RxnMapper, which is the best AAM algorithm according to our previous study, we demonstrate that our AAM algorithm is superior to it on our "Golden" benchmarking data set.
本文介绍了一种新的基于转换器神经网络的反应原子到原子映射（AAM）算法，该算法用于直接处理分子图作为原子和键的集合，而不是基于 SMILES/SELFIES 序列的方法，并结合了来自转换器的双向编码器表示（BERT）网络。图转换器用于提取与原子和键相关的分子特征。BERT 网络用于化学转化学习。在与 IBM RxnMapper 的基准研究中，根据我们之前的研究，它是最好的 AAM 算法，我们证明了我们的 AAM 算法在我们的“Golden”基准数据集上优于它。
