2026-09-20T11:17:13Zhttps://dspace.mit.edu/server/oai/request

oai:dspace.mit.edu:1721.1/1472972023-01-20T03:45:46Zcom_1721.1_7582com_1721.1_7581col_1721.1_131023

Coley, Connor W. Tu, Zhengkai Massachusetts Institute of Technology. Center for Computational Science and Engineering 2023-01-19T18:43:46Z 2023-01-19T18:43:46Z 2022-09 2022-09-12T19:54:34.804Z https://hdl.handle.net/1721.1/147297 Synthesis planning and reaction outcome prediction are two fundamental problems in computer-aided organic chemistry for which a variety of data-driven approaches have emerged. Natural language approaches that model each problem as a SMILESto-SMILES translation lead to a simple end-to-end formulation, reduce the need for data preprocessing, and enable the use of well-optimized machine translation model architectures. However, SMILES representations are not efficient for capturing information about molecular structures, as evidenced by the success of SMILES augmentation to boost empirical performance. Here, we describe a novel Graph2SMILES model that combines the power of Transformer models for text generation with the permutation invariance of molecular graph encoders that mitigates the need for input data augmentation. In our encoder, a directed message passing neural network (DMPNN) captures local chemical environments, and the global attention encoder allows for long-range and intermolecular interactions, enhanced by graph-aware positional embedding. As an end-to-end architecture, Graph2SMILES can be used as a dropin replacement for the Transformer in any task involving molecule(s)-to-molecule(s) transformations, which we empirically demonstrate leads to improved performance on existing benchmarks for both retrosynthesis and reaction outcome prediction. S.M. Massachusetts Institute of Technology In Copyright - Educational Use Permitted Copyright MIT http://rightsstatements.org/page/InC-EDU/1.0/ Scalable Model for Reaction Outcome Prediction and One-step Retrosynthesis with a Graph-to-Sequence Architecture Thesis application/pdf Master Master of Science in Computational Science and Engineering Publication unknown unknown <Publication xmlns="https://www.openaire.eu/cerif-profile/1.1/" id="479bc1ac-d59f-4225-b155-3ba2a4506767"> <Type xmlns="https://www.openaire.eu/cerif-profile/vocab/COAR_Publication_Types">http://purl.org/coar/resource_type/c_1843</Type> <Title>Scalable Model for Reaction Outcome Prediction and One-step Retrosynthesis with a Graph-to-Sequence Architecture</Title> <PublishedIn> <Publication> </Publication> </PublishedIn> <PublicationDate>2022-09</PublicationDate> <Authors> <Author> <DisplayName>Tu, Zhengkai</DisplayName> <Affiliation> <OrgUnit> </OrgUnit> </Affiliation> </Author> </Authors> <Editors> </Editors> <Publishers> <Publisher> <DisplayName>Massachusetts Institute of Technology</DisplayName> <OrgUnit /> </Publisher> </Publishers> <License>http://rightsstatements.org/page/InC-EDU/1.0/</License> <Abstract>Synthesis planning and reaction outcome prediction are two fundamental problems in computer-aided organic chemistry for which a variety of data-driven approaches have emerged. Natural language approaches that model each problem as a SMILESto-SMILES translation lead to a simple end-to-end formulation, reduce the need for data preprocessing, and enable the use of well-optimized machine translation model architectures. However, SMILES representations are not efficient for capturing information about molecular structures, as evidenced by the success of SMILES augmentation to boost empirical performance. Here, we describe a novel Graph2SMILES model that combines the power of Transformer models for text generation with the permutation invariance of molecular graph encoders that mitigates the need for input data augmentation. In our encoder, a directed message passing neural network (DMPNN) captures local chemical environments, and the global attention encoder allows for long-range and intermolecular interactions, enhanced by graph-aware positional embedding. As an end-to-end architecture, Graph2SMILES can be used as a dropin replacement for the Transformer in any task involving molecule(s)-to-molecule(s) transformations, which we empirically demonstrate leads to improved performance on existing benchmarks for both retrosynthesis and reaction outcome prediction.</Abstract> <Access xmlns="http://purl.org/coar/access_right" > </Access> </Publication>