Enantioselectivity prediction of pallada-electrocatalysed C–H activation using transition state knowledge in machine learning

Nature Synthesis, Journal Year: 2023, Volume and Issue: 2(4), P. 321 - 330

Published: Jan. 30, 2023

Language: Английский

---

Genetic Optimization of Homogeneous Catalysts

Chemistry - Methods, Journal Year: 2022, Volume and Issue: 2(6)

Published: March 4, 2022

Abstract We present the NaviCatGA package, a versatile genetic algorithm capable of optimizing molecular catalyst structures using well‐suited fitness functions to achieve set targeted properties. The flexibility and generality this tool are validated demonstrated with two examples: i) Ligand optimization exploration for Ni‐catalyzed aryl‐ether cleavage manipulating SMILES function derived from volcano plots, ii) multi‐objective (i. e., activity/selectivity) bipyridine N,N ‐dioxide Lewis basic organocatalysts asymmetric propargylation benzaldehyde 3D fragments. show that evolutionary optimization, enabled by NaviCatGA, is an efficient way accelerating discovery through bypassing combinatorial scaling issues incorporating compelling chemical constraints.

Language: Английский

---

When machine learning meets molecular synthesis

Trends in Chemistry, Journal Year: 2022, Volume and Issue: 4(10), P. 863 - 885

Published: Aug. 31, 2022

Language: Английский

---

Data-driven design of new chiral carboxylic acid for construction of indoles with C-central and C–N axial chirality via cobalt catalysis

Nature Communications, Journal Year: 2023, Volume and Issue: 14(1)

Published: May 31, 2023

Challenging enantio- and diastereoselective cobalt-catalyzed C-H alkylation has been realized by an innovative data-driven knowledge transfer strategy. Harnessing the statistics of a related transformation as source, designed machine learning (ML) model took advantage delta enabled accurate extrapolative enantioselectivity predictions. Powered model, virtual screening broad scope 360 chiral carboxylic acids led to discovery new catalyst featuring intriguing furyl moiety. Further experiments verified that predicted acid can achieve excellent stereochemical control for target alkylation, which supported expedient synthesis large library substituted indoles with C-central C-N axial chirality. The reported approach provides powerful data engine accelerate molecular catalysis harnessing hidden value available structure-performance statistics.

Language: Английский

---

Paving the road towards automated homogeneous catalyst design

ChemPlusChem, Journal Year: 2024, Volume and Issue: 89(7)

Published: Jan. 26, 2024

In the past decade, computational tools have become integral to catalyst design. They continue offer significant support experimental organic synthesis and catalysis researchers aiming for optimal reaction outcomes. More recently, data-driven approaches utilizing machine learning garnered considerable attention their expansive capabilities. This Perspective provides an overview of diverse initiatives in realm design introduces our automated tailored high-throughput silico exploration chemical space. While valuable insights are gained through methods analysis space, degree automation modularity key. We argue that integration data-driven, modular workflows is key enhancing homogeneous on unprecedented scale, contributing advancement research.

Language: Английский

---

AI for organic and polymer synthesis

Science China Chemistry, Journal Year: 2024, Volume and Issue: 67(8), P. 2461 - 2496

Published: June 26, 2024

Language: Английский

---

Data Science Guided Multiobjective Optimization of a Stereoconvergent Nickel-Catalyzed Reduction of Enol Tosylates to Access Trisubstituted Alkenes

ACS Catalysis, Journal Year: 2024, Volume and Issue: 14(7), P. 4699 - 4708

Published: March 13, 2024

Herein we report a method for stereoconvergent synthesis of trisubstituted alkenes in two steps from simple ketone starting materials. The key step is nickel-catalyzed reduction the corresponding enol tosylates that predominantly relies on monophosphine ligand to direct formation either E- or Z-trisubstituted alkene products. Reaction optimization was accomplished using data science workflow including training set design, statistical modeling, and multiobjective Bayesian optimization. campaign significantly improved access both products up ∼90:10 diastereoselectivity >90% yield. After identifying superior ligands only 25 reactions were required each objective (E- Z-isomer formation) converge reaction parameters search space ∼30,000 potential conditions EDBO+ platform. Additionally, hierarchical machine learning model developed predict stereoselectivity untested achieve validation mean absolute error (MAE) 7.1% selectivity (0.21 kcal/mol). Ultimately, present synergistic leveraging integration optimization, thereby expanding stereodefined alkenes.

Language: Английский

---

Mechanistic Inference from Statistical Models at Different Data-Size Regimes

ACS Catalysis, Journal Year: 2022, Volume and Issue: 12(13), P. 7886 - 7906

Published: June 17, 2022

The chemical sciences are witnessing an influx of statistics into the catalysis literature. These developments propelled by modern technological advancements that leading to fast and reliable data production, mining, management. In organic chemistry, models encoded with information-rich parameters have facilitated formulation mechanistic hypotheses across different data-size regimes. Herein, we aim demonstrate through selected examples integration statistical principles homogeneous can streamline not only reaction optimization protocols but also investigation procedures. Namely, highlight how aspects molecular modeling, set design, visualization, nuanced restructuring contribute improving reactivity selectivity, while furthering our understanding mechanisms. By mapping out these techniques at sizes, hope encourage broad application data-driven approaches for studies regardless accessible amount data.

Language: Английский

---

Bridging Chemical Knowledge and Machine Learning for Performance Prediction of Organic Synthesis

Chemistry - A European Journal, Journal Year: 2022, Volume and Issue: 29(6)

Published: Oct. 7, 2022

Recent years have witnessed a boom of machine learning (ML) applications in chemistry, which reveals the potential data-driven prediction synthesis performance. Digitalization and ML modelling are key strategies to fully exploit unique within synergistic interplay between experimental data robust performance selectivity. A series exciting studies demonstrated importance chemical knowledge implementation ML, improves model's capability for making predictions that challenging often go beyond abilities human beings. This Minireview summarizes cutting-edge embedding techniques model designs synthetic prediction, elaborating how can be incorporated into until June 2022. By merging organic tactics informatics, we hope this Review provide guide map intrigue chemists revisit digitalization computerization chemistry principles.

Language: Английский

---

High-throughput screening of CO2 cycloaddition MOF catalyst with an explainable machine learning model

Green Energy & Environment, Journal Year: 2024, Volume and Issue: unknown

Published: Feb. 1, 2024

The high porosity and tunable chemical functionality of metal-organic frameworks (MOFs) make it a promising catalyst design platform. High-throughput screening catalytic performance is feasible since the large MOF structure database available. In this study, we report machine learning model for high-throughput catalysts CO2 cycloaddition reaction. descriptors training were judiciously chosen according to reaction mechanism, which leads accuracy up 97% 75% quantile set as classification criterion. feature contribution was further evaluated with SHAP PDP analysis provide certain physical understanding. 12,415 hypothetical structures 100 reported MOFs under °C 1 bar within one day using model, 239 potentially efficient discovered. Among them, MOF-76(Y) achieved top experimentally among MOFs, in good agreement prediction.

Language: Английский

---