skip to main content
US FlagAn official website of the United States government
dot gov icon
Official websites use .gov
A .gov website belongs to an official government organization in the United States.
https lock icon
Secure .gov websites use HTTPS
A lock ( lock ) or https:// means you've safely connected to the .gov website. Share sensitive information only on official, secure websites.

Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: G2Retro as a two-step graph generative models for retrosynthesis prediction
Retrosynthesis is a procedure where a target molecule is transformed into potential reactants and thus the synthesis routes can be identified. Recently, computational approaches have been developed to accelerate the design of synthesis routes. In this paper,we develop a generative framework G2Retro for one-step retrosynthesis prediction. G2Retro imitates the reversed logic of synthetic reactions. It first predicts the reaction centers in the target molecules (products), identifies the synthons needed to assemble the products, and transforms these synthons into reactants. G2Retro defines a comprehensive set of reaction center types, and learns from the molecular graphs of the products to predict potential reaction centers. To complete synthons into reactants, G2Retro considers all the involved synthon structures and the product structures to identify the optimal completion paths, and accordingly attaches small substructures sequentially to the synthons. Here we show that G2Retro is able to better predict the reactants for given products in the benchmark dataset than the state-of-the-art methods.  more » « less
Award ID(s):
2133650
PAR ID:
10471065
Author(s) / Creator(s):
; ; ; ;
Publisher / Repository:
Nature Portfolio
Date Published:
Journal Name:
Communications Chemistry
Volume:
6
Issue:
1
ISSN:
2399-3669
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ; ; ; ; ; (, Nature Communications)
    Abstract To enhance the reaction kinetics without sacrificing activity in porous materials, one potential solution is to utilize the anisotropic distribution of pores and channels besides enriching active centers at the reactive surfaces. Herein, by designing a unique distribution of oriented pores and single crystalline array structures in the presence of abundant acid sites as demonstrated in the ZSM-5 nanorod arrays grown on monoliths, both enhanced dynamics and improved capacity are exhibited simultaneously in propene capture at low temperature within a short duration. Meanwhile, the ZSM-5 array also helps mitigate the long-chain HCs and coking formation due to the enhanced diffusion of reactants in and reaction products out of the array structures. Further integrating the ZSM-5 array with Co3O4nanoarray enables comprehensive propene removal throughout a wider temperature range. The array structured film design could offer energy-efficient solutions to overcome both sorption and reaction kinetic restrictions in various solid porous materials for various energy and chemical transformation applications. 
    more » « less
  2. ; ; ; ; ; (, NeuralPS)
    Computer-aided synthesis planning (CASP) algorithms have demonstrated expertlevel abilities in planning retrosynthetic routes to molecules of low to moderate complexity. However, current search methods assume the sufficiency of reaching arbitrary building blocks, failing to address the common real-world constraint where using specific molecules is desired. To this end, we present a formulation of synthesis planning with starting material constraints. Under this formulation, we propose Double-Ended Synthesis Planning (DESP), a novel CASP algorithm under a bidirectional graph search scheme that interleaves expansions from the target and from the goal starting materials to ensure constraint satisfiability. The search algorithm is guided by a goal-conditioned cost network learned offline from a partially observed hypergraph of valid chemical reactions. We demonstrate the utility of DESP in improving solve rates and reducing the number of search expansions by biasing synthesis planning towards expert goals on multiple new benchmarks. DESP can make use of existing one-step retrosynthesis models, and we anticipate its performance to scale as these one-step model capabilities improve. 
    more » « less
  3. ; ; ; ; ; (, Angewandte Chemie International Edition)
    Abstract Mechanistic understanding of organic reactions can facilitate reaction development, impurity prediction, and in principle, reaction discovery. While several machine learning models have sought to address the task of predicting reaction products, their extension to predicting reaction mechanisms has been impeded by the lack of a corresponding mechanistic dataset. In this study, we construct such a dataset by imputing intermediates between experimentally reported reactants and products using expert reaction templates and train several machine learning models on the resulting dataset of 5,184,184 elementary steps. We explore the performance and capabilities of these models, focusing on their ability to predict reaction pathways and recapitulate the roles of catalysts and reagents. Additionally, we demonstrate the potential of mechanistic models in predicting impurities, often overlooked by conventional models. We conclude by evaluating the generalizability of mechanistic models to new reaction types, revealing challenges related to dataset diversity, consecutive predictions, and violations of atom conservation. 
    more » « less
  4. ; ; (, Chemical Science)
    Monomeric boroles have been gaining attention as reagents for the synthesis of heterocycles due to their ability to insert atoms into the BC 4 ring in a single step. Although unique boron frameworks can be accessed via this methodology, the products feature aryl substitution on the carbon centers as steric bulk is required to preclude borole dimerization. This work demonstrates that insertion chemistry is possible with Diels–Alder dimeric boroles and that such reactivity is not exclusive to monomeric boroles with bulky groups. With 1-phenyl-2,3,4,5-tetramethylborole dimer, the formal 1,1-insertion of a nitrene and sulfur generate the six-membered aromatic 1,2-azaborine and 1,2-thiaborine, respectively. The isolation of the 1,2-thiaborine enabled the synthesis of an η 6 -chromium complex. Benzophenone and diphenylketene readily insert a CO unit to generate BOC 5 seven-membered rings confirming dimeric boroles can serve as monomeric synthons in 1,2-insertion reactions. An epoxide did not furnish the anticipated eight-membered BOC 6 ring, instead provided a bicyclic system with a BOC 3 ring. The insertion chemistry was demonstrated with two other borole dimers featuring different substitution with diphenylketene as a substrate. This work elevates borole insertion chemistry to a new level to access products that do not require bulky substitution. 
    more » « less
  5. ; ; (, Chemical Society Reviews)
    In this review, recent examples featuring C–H functionalization in the synthesis of complex natural products are discussed. A focus is given to the way in which C–H functionalization can influence the logical process of retrosynthesis, and the review is organized by the type and method of C–H functionalization. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email