Kinetically-driven reactivity of sulfinylamines enables direct conversion of carboxylic acids to sulfinamides

Chemical Science, Journal Year: 2023, Volume and Issue: 14(46), P. 13384 - 13391

Published: Jan. 1, 2023

Sulfinamides are some of the most centrally important four-valent sulfur compounds that serve as critical entry points to an array emergent medicinal functional groups, molecular tools for bioconjugation, and synthetic intermediates including sulfoximines, sulfonimidamides, sulfonimidoyl halides, well a wide range other S(iv) S(vi) functionalities. Yet, accessible chemical space sulfinamides remains limited, approaches largely confined two-electron nucleophilic substitution reactions. We report herein direct radical-mediated decarboxylative sulfinamidation first time enables access from broad structurally diverse carboxylic acids. Our studies show formation prevails despite inherent thermodynamic preference radical addition nitrogen atom, while machine learning-derived model facilitates prediction reaction efficiency based on computationally generated descriptors underlying reactivity.

Language: Английский

---

Design rules applied to silver nanoparticles synthesis: A practical example of machine learning application.

Computational and Structural Biotechnology Journal, Journal Year: 2024, Volume and Issue: 25, P. 20 - 33

Published: Feb. 17, 2024

The synthesis of silver nanoparticles with controlled physicochemical properties is essential for governing their intended functionalities and safety profiles. However, process involves multiple parameters that could influence the resulting properties. This challenge be addressed development predictive models forecast endpoints based on key parameters. In this study, we manually extracted synthesis-related data from literature leveraged various machine learning algorithms. Data extraction included such as reactant concentrations, experimental conditions, well antibacterial efficiencies toxicological profiles synthesized were also extracted. a second step, completeness, employed regression algorithms to establish relationships between desired build models. core size efficiency trained validated using cross-validation approach. Finally, features' impact was evaluated via Shapley values provide insights into contribution features predictions. Factors duration, scale choice capping agents emerged most significant predictors. study demonstrated potential aid in rational design paves way safe-by-design principles by providing optimization achieve provides valuable dataset compiled sources time effort researchers. Access datasets notably aids computational advances field nanotechnology.

Language: Английский

---

A genetic optimization strategy with generality in asymmetric organocatalysis as a primary target

Chemical Science, Journal Year: 2024, Volume and Issue: 15(10), P. 3640 - 3660

Published: Jan. 1, 2024

A genetic optimization strategy to discover asymmetric organocatalysts with high activity and enantioselectivity across a broad substrate scope.

Language: Английский

---

Do Chemformers Dream of Organic Matter? Evaluating a Transformer Model for Multistep Retrosynthesis

Journal of Chemical Information and Modeling, Journal Year: 2024, Volume and Issue: 64(8), P. 3021 - 3033

Published: April 11, 2024

Synthesis planning of new pharmaceutical compounds is a well-known bottleneck in modern drug design. Template-free methods, such as transformers, have recently been proposed an alternative to template-based methods for single-step retrosynthetic predictions. Here, we trained and evaluated transformer model, called the Chemformer, retrosynthesis predictions within discovery. The proprietary data set used training comprised ∼18 M reactions from literature, patents, electronic lab notebooks. Chemformer was purpose both multistep retrosynthesis. We found that performance especially good on reaction classes common discovery, with most showing top-10 round-trip accuracy above 0.97. Moreover, reached higher compared model. By analyzing experiments, observed synthetic routes, leading commercial starting materials 95% target compounds, increase more than 20% model compound set. In addition this, discovered suggested novel disconnections corresponding templates, which are not included These findings were further supported by publicly available ChEMBL conclusions drawn this work allow design synthesis tool where template-free models harmony optimize recommendations.

Language: Английский

---

Top 20 influential AI-based technologies in chemistry

Artificial Intelligence Chemistry, Journal Year: 2024, Volume and Issue: 2(2), P. 100075 - 100075

Published: July 27, 2024

The beginning and ripening of digital chemistry is analyzed focusing on the role artificial intelligence (AI) in an expected leap chemical sciences to bring this area next evolutionary level. analytic description selects highlights top 20 AI-based technologies 7 broader themes that are reshaping field. It underscores integration tools such as machine learning, big data, twins, Internet Things (IoT), robotic platforms, smart control processes, virtual reality blockchain, among many others, enhancing research methods, educational approaches, industrial practices chemistry. significance study lies its focused overview how these innovations foster a more efficient, sustainable, innovative future sciences. This article not only illustrates transformative impact but also draws new pathways chemistry, offering broad appeal researchers, educators, industry professionals embrace advancements for addressing contemporary challenges

Language: Английский

---

ROBERT: Bridging the Gap Between Machine Learning and Chemistry

Wiley Interdisciplinary Reviews Computational Molecular Science, Journal Year: 2024, Volume and Issue: 14(5)

Published: Sept. 1, 2024

ABSTRACT Beyond addressing technological demands, the integration of machine learning (ML) into human societies has also promoted sustainability through adoption digitalized protocols. Despite these advantages and abundance available toolkits, a substantial implementation gap is preventing widespread incorporation ML protocols computational experimental chemistry communities. In this work, we introduce ROBERT, software carefully crafted to make more accessible chemists all programming skill levels, while achieving results comparable those field experts. We conducted benchmarking using six recent studies in containing from 18 4149 entries. Furthermore, demonstrated program's ability initiate workflows directly SMILES strings, which simplifies generation predictors for common problems. To assess ROBERT's practicality real‐life scenarios, employed it discover new luminescent Pd complexes with modest dataset 23 points, frequently encountered scenario studies.

Language: Английский

---

Toward Next-Generation Heterogeneous Catalysts: Empowering Surface Reactivity Prediction with Machine Learning

Engineering, Journal Year: 2024, Volume and Issue: 39, P. 25 - 44

Published: Jan. 5, 2024

Heterogeneous catalysis remains at the core of various bulk chemical manufacturing and energy conversion processes, its revolution necessitates hunt for new materials with ideal catalytic activities economic feasibility. Computational high-throughput screening presents a viable solution to this challenge, as machine learning (ML) has demonstrated great potential in accelerating such processes by providing satisfactory estimations surface reactivity relatively low-cost information. This review focuses on recent progress applying ML adsorption prediction, which predominantly quantifies solid catalyst. models that leverage inputs from different categories exhibit levels complexity are classified discussed. At end review, an outlook current challenges future opportunities ML-assisted catalyst is supplied. We believe summarizes major achievements discovery through can inspire researchers further devise novel strategies accelerate design and, ultimately, reshape industry landscape.

Language: Английский

---

Machine learning models and performance dependency on 2D chemical descriptor space for retention time prediction of pharmaceuticals

Journal of Chromatography A, Journal Year: 2024, Volume and Issue: 1730, P. 465109 - 465109

Published: June 18, 2024

Language: Английский

---

Machine Learning for Reaction Performance Prediction in Allylic Substitution Enhanced by Automatic Extraction of a Substrate-Aware Descriptor

Journal of Chemical Information and Modeling, Journal Year: 2025, Volume and Issue: 65(1), P. 312 - 325

Published: Jan. 2, 2025

Despite remarkable advancements in the organic synthesis field facilitated by use of machine learning (ML) techniques, prediction reaction outcomes, including yield estimation, catalyst optimization, and mechanism identification, continues to pose a significant challenge. This challenge arises primarily from lack appropriate descriptors capable retaining crucial molecular information for accurate while also ensuring computational efficiency. study presents successful application ML predicting performance Ir-catalyzed allylic substitution reactions. We introduce SubA, an innovative substrate-aware descriptor that is inspired fact specific atoms or motifs reactants drive outcomes. By employing graph matching algorithms backbone identification incorporating atomic properties derived density functional theory calculations, SubA extracts essential at both level level. Compared four mainstream descriptors, achieves reduced dimensionality enhanced accuracy with over 2% mean absolute error reduction random scaffold splitting evaluations. It demonstrates better generalization when confronted previously unreported substrate combinations extended experiments. Furthermore, interpretable analysis shows predictor focuses on key features, offering insights into mechanisms.

Language: Английский

---

Connecting the Complexity of Stereoselective Synthesis to the Evolution of Predictive Tools

Chemical Science, Journal Year: 2025, Volume and Issue: unknown

Published: Jan. 1, 2025

This review provides an overview of predictive tools in asymmetric synthesis. The evolution methods from simple qualitative pictures to complicated quantitative approaches is connected with the increased complexity stereoselective

Language: Английский

---