Discovery of Vascular Endothelial Growth Factor Receptor 2 Inhibitors Employing Junction Tree Variational Autoencoder with Bayesian Optimization and Gradient Ascent DOI Creative Commons
Gia-Bao Truong, Thanh-An Pham, Van-Thinh To

et al.

ACS Omega, Journal Year: 2024, Volume and Issue: 9(47), P. 47180 - 47193

Published: Nov. 12, 2024

In the development of anticancer medications, vascular endothelial growth factor receptor 2 (VEGFR-2), which belongs to protein tyrosine kinase family, emerges as one most significant targets interest. The ongoing Food and Drug Administration (FDA) approval novel therapeutic medicines toward VEGFR-2 emphasizes urgent need discover sophisticated molecular structures that are capable reliably limiting activity. Recognizing huge potential deep-learning-based model advancements, we focused our study on exploring chemical space find small molecules potentially inhibiting VEGFR-2. To achieve this goal, utilized junction tree variational autoencoder in combination with two optimization approaches latent space: local Bayesian initial data set gradient ascent nine FDA-approved drugs targeting results yielded a 493 uncharted molecules. Quantitative structure–activity relationship (QSAR) models docking were used assess generated for their inhibitory using predicted pIC50 binding affinity. QSAR constructed RDK7 fingerprints CatBoost algorithm achieved remarkable coefficients determination (R2) 0.792 ± 0.075 0.859 respect internal external validation. Molecular was implemented 4ASD complex optimistic retrospective control (the ROC-AUC value 0.710 activity threshold −7.90 kcal/mol). Newly possessing acceptable corresponding both assessments shortlisted checked interactions at site important residues, including Cys919, Asp1046, Glu885.

Language: Английский

Exploring the Antidiabetic Potential of Pyrimidine‐Derived Chalcones: Synthesis, Biological Evaluation, and Molecular Modeling DOI Creative Commons
Miraj Fatima, Samina Aslam,

Aroog Fatima

et al.

ChemistrySelect, Journal Year: 2025, Volume and Issue: 10(2)

Published: Jan. 1, 2025

Abstract In the current research work, we have prepared a series of pyrimidine moiety‐containing molecules due to their promising medicinal profile. First all, two acetyl derivatives ( 15 , 16 ), and then they reacted with different aryl aldehydes form various chalcones in 63–84% yield. The synthesized compounds were characterized by analytical techniques screened for antidiabetic activity. Almost all 17 – 43 ) showed good excellent Among compounds, 30 remarkable activity IC 50 values 5.118 µ m 5.187 respectively, as compared standard reference drug acrabose = 37.38 ). While 18, 19, 21, 22, 23, 27, 31, 33, 38, 42, also Additionally, most biopotent drugs' molecular docking studies supported distinct connections between substituent moieties domains agreed experimental findings. dynamics simulation study active highest binding propensity enzyme revealed robustness complexes from study.

Language: Английский

Citations

1
Discovery of novel inhibitors of dengue viral RNA-dependent RNA polymerase by molecular docking, in vitro assay, DFT, and MD simulations DOI

Chaochun Wei,

Keli Zong,

Wei Li

et al.

International Journal of Biological Macromolecules, Journal Year: 2025, Volume and Issue: 305, P. 141328 - 141328

Published: Feb. 20, 2025

Language: Английский

Citations

0
Discovery of Vascular Endothelial Growth Factor Receptor 2 Inhibitors Employing Junction Tree Variational Autoencoder with Bayesian Optimization and Gradient Ascent DOI Creative Commons
Gia-Bao Truong, Thanh-An Pham, Van-Thinh To

et al.

ACS Omega, Journal Year: 2024, Volume and Issue: 9(47), P. 47180 - 47193

Published: Nov. 12, 2024

In the development of anticancer medications, vascular endothelial growth factor receptor 2 (VEGFR-2), which belongs to protein tyrosine kinase family, emerges as one most significant targets interest. The ongoing Food and Drug Administration (FDA) approval novel therapeutic medicines toward VEGFR-2 emphasizes urgent need discover sophisticated molecular structures that are capable reliably limiting activity. Recognizing huge potential deep-learning-based model advancements, we focused our study on exploring chemical space find small molecules potentially inhibiting VEGFR-2. To achieve this goal, utilized junction tree variational autoencoder in combination with two optimization approaches latent space: local Bayesian initial data set gradient ascent nine FDA-approved drugs targeting results yielded a 493 uncharted molecules. Quantitative structure–activity relationship (QSAR) models docking were used assess generated for their inhibitory using predicted pIC50 binding affinity. QSAR constructed RDK7 fingerprints CatBoost algorithm achieved remarkable coefficients determination (R2) 0.792 ± 0.075 0.859 respect internal external validation. Molecular was implemented 4ASD complex optimistic retrospective control (the ROC-AUC value 0.710 activity threshold −7.90 kcal/mol). Newly possessing acceptable corresponding both assessments shortlisted checked interactions at site important residues, including Cys919, Asp1046, Glu885.

Language: Английский

Citations

0