AI-assisted Structure and Function Prediction Services
Bioinformatics is an interdisciplinary field of computer science, statistics, and life sciences. Based on the algorithms and professional software, bioinformatics can mine and interpret tremendous biological data. As a powerful tool, in silico bioinformatics can help scientists to understand the biological processes with computationally techniques, for data mining, machine learning algorithms, pattern recognition, and visualization. Creative Biolabs has established the AI-assisted structural and functional analysis platform to support our customers in molecular docking, molecular dynamics simulation, virtual screening, and quantitative structural activity relationship (QSAR) analysis.
Fig.1 Summary of a classical structure-based drug design approach. (De, et al., 2016)
Molecular docking is a structure-based drug design technique that has been widely used in protein-ligand docking, protein-protein docking, and protein-nucleic acid docking. Scientists utilize this method to predict the most probable 3D conformations of small-molecule ligands within target binding sites. The molecular docking method can provide the binding energetics of the docked compounds and help researchers find promising molecules from compound databases that show a high affinity with targets. By calculating the active site, shape, and energy, Creative Biolabs uses cutting-edge molecular docking software to identify the binding conformation of the target and ligand, and predict their binding pattern and affinity.
A computer simulation technique known as molecular dynamics (MD) simulation enables one to predict the time-dependent evolution of a system of interacting particles using the principles of physics. Under predetermined physiological parameters, such as temperature and pressure, MD simulation can offer atomic-level details regarding protein conformational changes and binding thermodynamics, enabling the study of a protein system at a timescale not yet possible with experimental methods. Creative Biolabs uses state-of-the-art software tools to study the molecular dynamics of protein systems. Based on the MD simulation technology, our AI-assisted planform can refine experimentally determined structures, characterize protein flexibility, study biocatalysts, evaluate protein-ligand binding, and monitor the protein folding process.
Fig.2 A schematic showing how a molecular dynamics simulation is performed. (Durrant & McCammon, 2011)
Virtual screening can be defined as a set of computational methods that select some promising compounds from compound databases for further experimental activity. This method requires a combination of drug design theory, computer technology, and professional application software. Based on the in silico virtual screening, drug researchers can screen out novel leads from dozens or even millions of molecules. Compared with the traditional high-throughput screening (HTS) approach, virtual screening is reliable and relatively inexpensive. Moreover, the hit rate can be greatly improved by combining the results of virtual screening with experimental screening.
Quantitative structure-activity relationship (QSAR) is statistical empirical model for the prediction of the biological effect of chemical compounds based on mathematical and statistical relations. QSAR analysis is a critical step in the optimization process of lead compounds to correlate molecular structure, which has speeded up the lead optimization process by multiple degrees in the last two decades. A ligand-based strategy utilizing the QSAR model may offer guidance for drug design for targets with little structural knowledge. Creative Biolabs offers AI-assisted QSAR analysis to guide drug discovery programs and explore significant factors associated with the activity of drug molecules.
De, Ruyck.J.; et al. Molecular docking as a popular tool in drug design, an in silico travel. Advances and applications in bioinformatics and chemistry: AABC. 2016, 9: 1.
Durrant, J.D.; McCammon, J.A. Molecular dynamics simulations and drug discovery. BMC biology. 2011, 9(1): 1-9.