SynRoute: A Retrosynthetic Planning Software

Latendresse, Mario; Malerich, Jeremiah P.; Herson, James; Krummenacker, Markus; Szeto, Judy; Vu, Vi-Anh; Collins, Nathan; Madrid, Peter B.

doi:10.1021/acs.jcim.3c00491

Cited by 5 publications

(8 citation statements)

References 34 publications

(66 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the Retro* 45 data set (Table 4), ReSynZ using 13k templates and the buyable molecules from Sigma-Aldrich shows a success rate of 85.71% for this data set, comparable to Retro*'s 86.84%, and outperforms DFPN 46 and 3N-MCTS. 12 We also notice that the solution success rate of ReSynZ increases significantly as the number of templates increases (Tables 3 and S2); other template-based CASP methods, e.g., SynRoute, 44 have a similar trend. However, the ReSynZ method achieves a higher solution success rate based on fewer templates (Table S2).…”

Section: Comparison Of Different Searching Strategiesmentioning

confidence: 75%

“…We performed retrosynthesis analysis using two data sets. For the Synroute 44 data set (Table 3), ReSynZ achieved an overall success rate of 82.4% with our 13k-rule data set, while SynRoute achieved a success rate of 74.8% with its 10k templates, and AiZynth-Finder 14 achieved 76% with its 10k templates. For the Retro* 45 data set (Table 4), ReSynZ using 13k templates and the buyable molecules from Sigma-Aldrich shows a success rate of 85.71% for this data set, comparable to Retro*'s 86.84%, and outperforms DFPN 46 and 3N-MCTS.…”

Section: Comparison Of Different Searching Strategiesmentioning

confidence: 91%

See 1 more Smart Citation

Retrosynthesis Zero: Self-Improving Global Synthesis Planning Using Reinforcement Learning

Guo,

Yu,

et al. 2024

J. Chem. Theory Comput.

View full text Add to dashboard Cite

The field of computer-aided synthesis planning (CASP) has witnessed significant growth in recent years. Still, many CASP programs rely on large data sets to train neural networks, resulting in limitations due to the data quality and prior knowledge from chemists. In response, we propose Retrosynthesis Zero (ReSynZ), a reaction template-based method that combines Monte Carlo Tree Search with reinforcement learning inspired by AlphaGo Zero. Unlike other single-step reaction template-based CASP methods, ReSynZ takes complete synthesis paths for complex molecules, determined by reaction rules, as input for training the neural network. ReSynZ enables neural networks trained with relatively small reaction data sets (tens of thousands of data) to generate multiple synthesis pathways for a target molecule and suggest possible reaction conditions. On multiple data sets of molecular retrosynthesis, ReSynZ demonstrates excellent predictive performance compared to existing algorithms. The advantages, such as self-improving model features, flexible reward settings, the potential to surpass human limitations in chemical synthesis route planning, and others, make ReSynZ a valuable tool in chemical synthesis design.

show abstract

Section: Comparison Of Different Searching Strategiesmentioning

confidence: 75%

Section: Comparison Of Different Searching Strategiesmentioning

confidence: 91%

Retrosynthesis Zero: Self-Improving Global Synthesis Planning Using Reinforcement Learning

Guo,

Yu,

et al. 2024

J. Chem. Theory Comput.

View full text Add to dashboard Cite

show abstract

“…Hong et al proposed an experience-guided Monte Carlo tree search (EG-MCTS), in which knowledge is learned from synthesizing experiences instead of rollout [146]. SynRoute, proposed by Latendresse et al [147], uses a relatively small number of reaction templates as well as a literature-based reaction database to search practical synthetic routes to target compounds. For each reaction template, a machine learning classifier is trained to make predictions.…”

Section: Graph Neural Networkmentioning

confidence: 99%

Application of Modern Intelligent Algorithms in Retrosynthesis Prediction

Liao,

Shi,

Zhu

2024

Preprint

View full text Add to dashboard Cite

In recent years, with the rapid advance of computer science, various modern intelligent algorithms have successively emerged. Transformer, based on multi-head attention mechanism, is one of the most favored AI models among in this century. The introduction of these algorithms leads to dramatic progress in retrosynthesis prediction. Unlike conventional retrosynthesis prediction models, retrosynthesis prediction based on intelligent algorithms can automatically extract chemistry knowledge from chemical reaction datasets to predict retrosynthesis routes. In this review, we provide a comprehensive overview of retrosynthesis prediction based on modern intelligent algorithms, particularly artificial intelligence algorithm. After introducing the related deep learning model, the existing chemical reaction datasets and molecular representations are presented. Subsequently, the current state-of-the art of AI-assisted retrosynthesis prediction models in recent years is discussed, including template-based models, template-free models, and semi-template-based models. Additionally, we conclude by comparing retrosynthesis prediction models across different categorizations. Finally, several challenges and limitations of these current methods are summarized, with a view to promising directions for future research.

show abstract

“…In multistep retrosynthesis, a single step predictor is used within a search algorithm, such as Monte-Carlo tree search (MCTS) 3 , to iteratively break down the target and subsequent intermediates until arriving at commercial starting materials. The field of AI-driven retrosynthesis prediction has received considerable attention lately 4,5 , and several models and tools have been developed to speed up the synthesis-planning process 3,4,[6][7][8][9][10][11][12][13] . AiZynthFinder 3,14 , for instance, is frequently used by chemists in industrial projects 13,15 .…”

Section: Introductionmentioning

confidence: 99%

“…The field of AI-driven retrosynthesis prediction has received considerable attention lately 4,5 , and several models and tools have been developed to speed up the synthesis-planning process 3,4,[6][7][8][9][10][11][12][13] . AiZynthFinder 3,14 , for instance, is frequently used by chemists in industrial projects 13,15 . The single step predictions in AI-driven retrosynthesis are carried out with a machine learning model which predicts reactants for a given product, either implicitly via a reaction template or explicitly via SMILES or molecule graphs.…”

Section: Introductionmentioning

confidence: 99%

Constrained synthesis planning with disconnection-aware transformer and multi-objective search

Westerlund,

Saigiridharan,

Genheden

2024

Preprint

View full text Add to dashboard Cite

Designing synthesis routes with shared intermediates for a set of target compounds is a common task in drug discovery. Multistep retrosynthesis tools such as AiZynthFinder are frequently used by chemists to generate possible routes. Although these tools can find solved routes for a majority of target compounds, they may not generate routes which comply with specific bond constraints. Such bond constraints could be defined in the project to obtain synthesis routes with common intermediates for the set of compounds. Here, we present a novel retrosynthesis approach which aims to generate routes in the feasible region defined by these constraints. The constraints are divided into bonds to break and bonds to freeze. First, we introduce a filter in the search which discards any reaction that violates the frozen bond constraints. Second, we benchmark four possible strategies for breaking selected bonds in the search for synthetic routes. We show that a combination of disconnection-aware transformer and multi-objective search generates routes which satisfy bond constraints for more targets in the PaRoutes dataset compared to the standard search (75.57% vs. 54.80%). Finally, we apply the strategy on a set of drug compounds to exemplify a real-world scenario. Using our novel approach, we can build a short joint synthesis route which satisfies the given bond constraints and covers eight of the ten compounds.

show abstract

SynRoute: A Retrosynthetic Planning Software

Cited by 5 publications

References 34 publications

Retrosynthesis Zero: Self-Improving Global Synthesis Planning Using Reinforcement Learning

Retrosynthesis Zero: Self-Improving Global Synthesis Planning Using Reinforcement Learning

Application of Modern Intelligent Algorithms in Retrosynthesis Prediction

Constrained synthesis planning with disconnection-aware transformer and multi-objective search

Contact Info

Product

Resources

About