Compound Active Site Prediction

The compound active site is predicted by virtual screening of a compound against hundreds of proteins representing different active site types. Compound active site prediction provides comprehensive effects, including potential target identification, therapeutic agent identification, and adverse effects prediction. Creative Biolabs has internationally advanced compound libraries and a variety of software systems that can predict and analyze active sites for tens of thousands of compounds.

How does Compound Active Site Prediction Work?

Taking the active site prediction of traditional Chinese medicine compounds as an example, Creative Biolabs provides a one-stop service platform covering new drug development services in various target and disease fields, especially, active site-based drug activity screening (including enzymatic targets, GPCR targets, ion channel targets, nuclear receptor targets) and in vitro and in vivo experiments.

• Obtaining active ingredients from Chinese medicine to be studied
• Predicting active sites based on active ingredients
• Obtain the corresponding target site based on disease/phenotype
• Intersection of the active site corresponding to the active ingredient and the target site corresponding to the disease phenotype
• Molecular binding validation (validation of chemical composition and molecular targets, molecular and molecular interactions)

Fig 1. Graphical abstract for synthesis, docking, ADMET prediction, cytotoxicity and antimicrobial activity of oxathiadiazole derivatives. (Yadav, 2018)

Applications in Preclinical Research

Recently, scientists have used 70,000 molecules and 130 research-researched chemical reactions to build a total of 170 million virtual compound libraries (the compounds can be synthesized on demand), including a total of 10.7 million unseen skeleton types. Using AmpC β-lactamase (AmpC) and D4 dopamine receptor as binding sites, 44 and 549 highest scoring compounds were predicted based on the results of molecular docking virtual screening. Subsequent predictions indicate that the hit rate (percentage of compounds that meet the required activity) is 11% and 24%, respectively. Starting from these predicted compounds, new types of AmpC high activity non-covalent inhibitors and D4 dopamine receptors with high activity and high selectivity have been successfully discovered.

Fig 2. Electron density maps for AmpC-inhibitor complexes. (Lyu, 2019)

Our Highlights

• Target prediction: quickly find potential targets for newly synthesized compounds or natural products.
• Drug repositioning: revitalize existing clinical drugs in the treatment of disease.
• Pharmacological analysis: elucidate the active molecules of the drug and its mechanism of action, constructing a component-target-disease network map.
• Molecular mechanism studies: explain the potential targets of small molecule compounds with phenotypic activity.

Creative Biolabs owns a variety of software systems which are based on the ligand similarity method to predict the pharmacological target of the compound, and provides a reference for the determination of synthetic compounds or natural product targets. If you are interested in the compound active site prediction, please contact us and we will happy to help you.

References

1. Yadav, M.; et al. Synthesis, docking, ADMET prediction, cytotoxicity and antimicrobial activity of oxathiadiazole derivatives. Computational Biology and Chemistry. 2018,77: 226-239.
2. Lyu, J.; et al. Ultra-large library docking for discovering new chemotypes. Nature. 2019, 566: 224–229.

For Research Use Only.