QM Assisted ML for 19F NMR Chemical Shift Prediction DOI Creative Commons
Patrick Penner, Anna Vulpetti

Опубликована: Авг. 7, 2023

Ligand-observed 19F NMR detection is an efficient method for screening libraries of fluorinated molecules in fragment-based drug design campaigns. Screening large mixtures makes a high-throughput method. Typically, these are generated from pools well-characterized fragments. By predicting chemical shift, could be arbitrary facilitating example focused screens. In previous publication, we introduced to predict shift using rooted fluorine fingerprints and machine learning (ML) methods. Having observed that the quality prediction depends on similarity training set, here propose assist with quantum mechanics (QM) based methods cases where compounds not well covered by set. Beyond similarity, performance ML associated individual features compounds. A combination both used as procedure split input data sets into those predicted required QM processing. We show proprietary fragment library, known LEF (Local Environment Fluorine), public Enamine set shifts synergize outperform either individually. Models built data, model building workflow tools, can found at https://github.com/PatrickPenner/lefshift https://github.com/PatrickPenner/lefqm.

Язык: Английский

Prediction of 19F NMR chemical shift by machine learning DOI Creative Commons
Yao Li,

Wen-Shuo Huang,

Li Zhang

и другие.

Artificial Intelligence Chemistry, Год журнала: 2024, Номер 2(1), С. 100043 - 100043

Опубликована: Янв. 4, 2024

Fluorine-19 (19F) is a nucleus of great importance in the field Nuclear Magnetic Resonance (NMR) spectroscopy due to its high receptivity and wide chemical shift dispersion. 19F NMR plays crucial roles both organic synthesis biomedicine. Herein, machine learning-based comprehensive prediction model was established based on experimental dataset from book by Dolbier open database nmrshiftdb2. Fluorine radical SMILES (Fr-SMILES) that reflected fluorine equivalence, designed as representation molecule. Model trained with graph convolution network (GCN) algorithm gave low mean absolute error (MAE) 3.636 ppm testing set. This exhibits broad applicability can effectively predict shifts for range molecules. We believe current work will provide powerful tool not only predicting but also aiding analysis identification these diverse compounds. An online platform constructed model, which be found at https://fluobase.cstspace.cn/fnmr.

Язык: Английский

Процитировано

8
Borate-Catalysed Direct Amidation Reactions of Coordinating Substrates DOI Creative Commons
Richard J. Procter, Carla Alamillo‐Ferrer,

Usman Shabbir

и другие.

Chemical Science, Год журнала: 2025, Номер unknown

Опубликована: Янв. 1, 2025

A new borate ester amidation catalyst was developed, that shows higher reactivity with challenging carboxylic acids and amines. Reactions could be performed on multigram scale the generated in situ from commercial reagents.

Язык: Английский

Процитировано

1
Direct Synthesis of Fluorinated Carbon Materials via a Solid‐State Mechanochemical Reaction Between Graphite and PTFE DOI
B. Z. Jang, Qiannan Zhao, Jaehoon Baek

и другие.

Advanced Functional Materials, Год журнала: 2023, Номер 33(47)

Опубликована: Июль 27, 2023

Abstract Fluorinated carbon materials (FCMs) have received significant attention, because of their exceptional stability, which is associated with the strong C‐F bonding, strongest among single bonds. However, fluorination requires extremely toxic and moisture‐sensitive reagents, makes it inapplicable for practical uses. Here, a straightforward relatively safe method are reported scalable synthesis FCMs, by mechanochemical depolymerization polytetrafluoroethylene (PTFE) fragmentation graphite. The resultant FCMs evaluated as anode lithium‐ion batteries (LIBs). An optimized FCM delivered capacities high 951.6 329.3 mAh g −1 at 0.05 10 A , respectively. It also demonstrated capacity retention 76.6% even after 1000 cycles 2.0 .

Язык: Английский

Процитировано

15
An unusual way of augmenting one-electron basis sets: New aug-pecS-n (n = 1, 2) basis sets for H, C, N, and O atoms for NMR shielding constant calculations that require extra diffuse functions DOI
Yury Yu. Rusakov, Irina L. Rusakova

The Journal of Chemical Physics, Год журнала: 2025, Номер 162(16)

Опубликована: Апрель 28, 2025

In this paper, the augmentation of NMR chemical shift-oriented pecS-n (n = 1, 2) basis sets for H, C, N, and O atoms with diffuse functions was proposed. New aug-pecS-n were shown to significantly improve accuracy calculated solvent corrections 1H, 13C, 15N, 17O shielding constants these anion systems. The carried out property-energy consistent method using isotropic dipole polarizability as target property, which is cardinally different from usual approach based on energy minimization.

Язык: Английский

Процитировано

0
QM assisted ML for 19F NMR chemical shift prediction DOI
Patrick Penner, Anna Vulpetti

Journal of Computer-Aided Molecular Design, Год журнала: 2023, Номер 38(1)

Опубликована: Дек. 12, 2023

Язык: Английский

Процитировано

4
On the relativistic effects on 19F nuclear magnetic resonance chemical shifts in the presence of iodine atoms DOI
Irina L. Rusakova, Stepan A. Ukhanev, Yury Yu. Rusakov

и другие.

Journal of Fluorine Chemistry, Год журнала: 2023, Номер 271, С. 110188 - 110188

Опубликована: Сен. 17, 2023

Язык: Английский

Процитировано

2
QM Assisted ML for 19F NMR Chemical Shift Prediction DOI Creative Commons
Patrick Penner, Anna Vulpetti

Опубликована: Окт. 9, 2023

Ligand-observed 19F NMR detection is an efficient method for screening libraries of fluorinated molecules in fragment-based drug design campaigns. Screening large mixtures makes a high-throughput method. Typically, these are generated from pools well-characterized fragments. By predicting chemical shift, could be arbitrary facilitating example focused screens. In previous publication, we introduced to predict shift using rooted fluorine fingerprints and machine learning (ML) methods. Having observed that the quality prediction depends on similarity training set, here propose assist with quantum mechanics (QM) based methods cases where compounds not well covered by set. Beyond similarity, performance ML associated individual features compounds. A combination both used as procedure split input data sets into those predicted required QM processing. We show proprietary fragment library, known LEF (Local Environment Fluorine), public Enamine set shifts synergize outperform either individually. Models built data, model building workflow tools, can found at https://github.com/PatrickPenner/lefshift https://github.com/PatrickPenner/lefqm.

Язык: Английский

Процитировано

2
QM Assisted ML for 19F NMR Chemical Shift Prediction DOI Creative Commons
Patrick Penner, Anna Vulpetti

Опубликована: Авг. 7, 2023

Ligand-observed 19F NMR detection is an efficient method for screening libraries of fluorinated molecules in fragment-based drug design campaigns. Screening large mixtures makes a high-throughput method. Typically, these are generated from pools well-characterized fragments. By predicting chemical shift, could be arbitrary facilitating example focused screens. In previous publication, we introduced to predict shift using rooted fluorine fingerprints and machine learning (ML) methods. Having observed that the quality prediction depends on similarity training set, here propose assist with quantum mechanics (QM) based methods cases where compounds not well covered by set. Beyond similarity, performance ML associated individual features compounds. A combination both used as procedure split input data sets into those predicted required QM processing. We show proprietary fragment library, known LEF (Local Environment Fluorine), public Enamine set shifts synergize outperform either individually. Models built data, model building workflow tools, can found at https://github.com/PatrickPenner/lefshift https://github.com/PatrickPenner/lefqm.

Язык: Английский

Процитировано

1