A merged molecular representation learning for molecular properties prediction with a web-based service

Deep learning has brought a dramatic development in molecular property prediction that is crucial in the field of drug discovery using various representations such as fingerprints, SMILES, and graphs. In particular, SMILES is used in various deep learning models via character-based approaches. However, SMILES has a limitation in that it is hard to reflect chemical properties. In this paper, we propose a new self-supervised method to learn SMILES and chemical contexts of molecules simultaneously in pre-training the Transformer. The key of our model is learning structures with adjacency matrix embedding and learning logics that can infer descriptors via Quantitative Estimation of Drug-likeness prediction in pre-training. As a result, our method improves the generalization of the data and achieves the best average performance by benchmarking downstream tasks. Moreover, we develop a web-based fine-tuning service to utilize our model on various tasks.

Mixed isatin with 3-(2-(aryl)hydrazono)acetylacetone Mn(II), Co(II) and Ni(II) complexes: antibacterial evaluation and molecular properties prediction

The metal complexes {Ni (II), Co (II) and Mn (II)} of 3-(2-(aryl)hydrazono)acetylacetone with isatin were synthesized and screened for their in vitro antibacterial activity against four pathogenic microorganisms {two Gram‐positive and two Gram negative}. The results of antibacterial activities revealed that all the metal complexes 1-9 exhibited moderate activities. Also, Lipinski's rule of five (RO5) of the mixed ligand metal complexes were calculated by SwissADME website. KEY WORDS: Isatin, 3-(2-(Aryl)hydrazono)acetylacetone, Metal complexes, Antibacterial activities, Lipinski rules Bull. Chem. Soc. Ethiop. 2020, 34(3), 533-541. DOI: https://dx.doi.org/10.4314/bcse.v34i3.9

Theoretical Molecular Properties Prediction, Docking and Antimicrobial Studies on Anisidine-Isatin Schiff Bases

The continuous intake of a specific antibiotic for diseases is continuously reducing the immunity of the human in course of time, which includes Covid-19 treatment. Recent research on Schiff bases shows the promising biological activities and good antibacterial results. In this study, three Schiff bases with lactam ring using isatin and three different anisidines in presence of acetic acid were synthesized and characterized. Drug likeness was examined using Molsoft and docking against the target proteins such as 5J6R, 3L9L, 5HVY, covid main protease and 3ZBO proteins for drug suitability. The experimental antibacterial activity against Gram-positive strains like Staphylococcus aureus, Bacillus subtilis and Staphylococcus epidermidis. Among the synthesized compounds, three Schiff bases ortho and meta substituted compounds exhibited good results when compared to para compound, where the methoxyl group position effect was observed.

Novel Organo Phosphonate-1,2,3 Triazole Derivatives: Molecular Properties Prediction

In the present scenario drug discovery and development processes are expensive and time consuming. To resolve this, we utilised the Lipinski’s rule (Ro5) methodology, which appears to be useful in defining drugability. In the present investigation, we reported the synthesis and evolution of antibacterial activity of title compounds and according to Rule of 5 series, twenty novel ((R)-dimethyl (hydroxy(4-((1-(2-nitrophenyl)-1H-1,2,3-triazol-4-yl)methoxy)phenyl)methyl)phosphonate-1,2,3-triazole derivatives were subjected to molecular properties prediction, drug likeness by Molinspiration (Molinspiration, 2020) and Molsoft (Molsoft, 2020) software.

A Merged Molecular Representation Learning for Molecular Properties Prediction with a Web-based Service

Deep learning has brought a dramatic development in molecular property prediction which is crucial in the field of drug discovery. Various methods such as fingerprint, SMILES, graphs have been proposed for representing molecules. Recently, unlabeled molecule data is used to improve performance for various pre-training methods. The main challenge of molecular properties predictions is designing a data representation and model that can show good performance for various datasets. However, performance deviation due to scarcity of dataset exists in constructing the model. We propose a new self-supervised method to learn the characteristics and structures of molecules by integrating existing methods. The key of our model is learning structures with matrix embedding and learning logics that can infer descriptors via QED prediction. As a result, our method improves the generalization of the data and achieves the best average performance by benchmarking downstream tasks. Moreover, we develop a web-based fine-tuning service to utilize our model on various tasks.