Based on our professional AI-assisted Glycoinformatics Development and AI-assisted Glycan Drug Discovery services, CD BioGlyco analyzes the binding affinity (BA) between ligands and receptors at the atomic and electronic levels through molecular docking, quantum chemical methods, and molecular dynamics (MD) simulations. Our lab uses computers to simulate the structure and behavior of glycan-related molecules at the atomic level of molecular modeling, which in turn simulates the physicochemical properties of the entire system.
Based on big data and artificial intelligence, integrating molecular simulation techniques, CD BioGlyco provides specialized computational modeling services by nuclear magnetic resonance (NMR), crystallographic, or homology modeling data. Our lab provides analysis services of the forces acting on each system atom, with detailed analysis of the forces generated by the interaction of bonded and non-bonded atoms.
NMR data is a very useful data package, and many of the receptor and ligand conformations we sample through MD simulations can be used to predict NMR measurements. Moreover, we provide a direct comparison service between experimental and theoretical NMR data.
Clustering analysis: CD BioGlyco evaluates the different conformations that a glycan or glycoprotein may have by computer simulation. We use conformational searches, saturated ring conformational searches, etc. to traverse as many of the dominant conformations of the ligand as possible to serve as a library of conformations for docking, thus improving docking accuracy.
Dominant conformation recognition: Each ligand has a variety of binding conformations. Our lab offers a specialized structural pattern screening service by combining free energy scoring, molecular stress energy, and professional judgment.
Identifying cryptic and allosteric binding sites: In addition to the well-defined binding pockets revealed by NMR and X-ray crystal structures, our researchers provide potential drug site identification services including cryptic binding sites and druggable allosteric sites.
Technology: BA, Synthetic accessibility (SA), Quantitative estimate of drug-likeness (QED), Natural product-likeness (NP), Evolutionary molecular generation algorithm (EMGA), Neural language model
Journal: Molecules
Published: 2022
IF: 4.927
Results: In this study, researchers designed and evaluated potential drug candidates through AI, MD simulations. The researchers developed a new neurolinguistic model based on the EMGA and further optimized the molecules in terms of SA, QED, NP, TF, and BA metrics. From the EMGA-generated drug candidates, the researchers used MD simulation to analyze and validate the selected 21 effective molecules. Two molecules were finally identified as potential drug candidates by calculating center of mass (COM) distances, root mean square displacement (RMSD) values, and binding free energies between Mpro and the 21 ligands.
Fig.1 COM distance analysis between ligands in complex simulations. (Elend, et al., 2022)
CD BioGlyco has rich experience in AI-assisted glycan drug discovery services. Our staff have mature industry experience, effectively solving the doubts in molecular simulation. Please feel free to contact us.
Reference
We envision a future where the intricate world of carbohydrate is no longer shrouded in mystery, but rather illuminated by the power of cutting-edge computational tools.