Drug targets refer to the binding sites of drugs in the body, including gene sites, receptors, enzymes, ion channels, nucleic acids and other biological macromolecules. One of key element in novel drug research and development is to screen and determine the drug targets. Nowadays, rational drug design heavily relies on potential drug targets or the chemical structure characteristics of their endogenous ligands and natural substrates revealed in life science research. Target analysis is therefore playing an essential role in discovering new drugs that selectively act on the target.
In common molecular docking calculations, an indispensable step is to define the binding position of ligand molecules (usually small organic molecules), that is, the docking pocket. For the X-ray crystal structure of the protein-small molecule complex, there is a ligand in the pocket indicating the location of the interface pocket. However, there are many structures analyzed by X-ray and NMR technology can not provide ligand structures. In addition, for other molecules such as nucleic acids, peptides, etc., it's a challenge to define the docking pocket. In the calculation of molecular docking, the docking pocket refers to a possible region where the ligand binds in the receptor. If the docking pocket is set at the accurate active binding site, there is a greater probability of finding the correct active conformation and binding mode of the ligand. As the name suggests, the pocket is usually in the shape of a pocket bag which can hold a certain volume of molecular structure, but there are also other shapes, such as a pipe shape, a groove shape and a shallow depression shape. For protein-ligand complexes, large and deep hydrophobic cavities are essential for ligand binding.
Figure 1. Overview of Chinese tree shrew data assembly for drug target analysis. (A) The pipeline of assembled data for drug target analysis. (B) A Venn diagram illustrating the overlap of transcripts obtained from the ab initio and de novo assembly methods. (Feng, Z.; et al. 2014)
At Alfa Chemistry, we select suitable targets for the development of new drugs by screening and analyzing targets. Our experts have developed various methods and models to perform target analysis:
(1) 2D QSAR/Fingerprint Model
Analyzing the structure of the compound can help to find the common structural features for the characteristic target, which can be used in the subsequent compound selection. We use 2D substructure/fragment cluster analysis, privileged fragment analysis, pharmacophore analysis and other methods to conduct the structural analysis. Methods such as cluster analysis can mine the similarities between data, classify the data, and discover the effective information hidden in the data. Based on this useful information, a molecular fingerprint or QSAR analysis model can be established to predict whether other molecular structures have corresponding target activity.
(2) ANN/KNN-based model
The data collected through a large-scale search often contains multiple factors related to the activity of the target, and these factors are not linearly related. Neural network modeling can learn and analyze these non-linear related factors, build optimal models, and perform predictive analysis on data. At Alfa Chemistry, the ANN/KNN-based model can be used to construct a QSAR model and predict the relationship between structure and activity more accurately. In addition, there are other methods such as principal component analysis (PCA) that can be used for modeling and target analysis.
(3) Molecular docking
We apply molecular docking to perform target analysis based on receptor characteristics and the interaction between receptor and drug molecules.
At Alfa Chemistry, we mainly apply computational chemistry methods to identify protein pockets based on structural analysis of geometric features and biochemical and physical features. Geometric features generally include three-dimensional grid, space sphere, α-share theory, and mathematical morphology theory. Biochemical physical characteristics generally include physical, chemical and biological properties such as binding energy and protein sequence conservation.
(1) Grid-based prediction algorithms (e.g. POCKET, LIGSITE, LIGSITEcs, Q-SiteFinder)
The theoretical basis of this algorithm is that small molecules tend to bind to large and deep pockets on the protein surface. We use POCKET, LIGSITE and LIGSITEcs methods to scan the protein-solvent-protein and surface-solvent-surface events in the grid respectively. POCKET and LIGSITE use atomic coordinates, and LIGSITEcs use connolly surfaces. The order of the Pocket site is the sum of the van der Waals interaction energy between the probe and the protein atom.
(2) Sphere-based prediction algorithms (e.g. SURFNET, PASS)
In SURFNET method, a sphere is placed between two atoms and tangent to the van der Waals surface of the two atoms. If this sphere has an intersection with the van der Waals surfaces of other atoms, then the radius of the sphere is reduced so that it cannot contain any atoms. The largest sphere is defined as the largest pocket. The PASS algorithm is a method to initialize the sphere only on the surface of the protein. It wraps the probe ball on the surface of the protein to remove the spheres that cannot meet the burial number. The active site points (ASPs) are determined according to the size of the burial number and the number of surrounding spheres.
(3) Prediction algorithm based on α-share theory (e.g. CAST)
CAST uses three-dimensional theory to describe the three-dimensional structure of a protein as a network of many triangles. The atoms in the protein form the vertices of the triangle. Then clustering is performed by merging small triangles into adjacent large triangles. The binding site is the final set of empty triangles, that is, the interior does not contain other atoms.
We use molecular dynamics (MD) simulation to simulate pocket dynamics. Alfa Chemistry supports various constrained geometry simulation methods to study the dynamics of protein binding pockets caused by large-scale protein movement. The pocket dynamics can be fully sampled within the simulation time with high accuracy and small calculations amount.
Our target analysis and pocket finding services remarkably reduce the cost, promote further experiments, and accelerate the process of drug design for customers worldwide. Our personalized and all-around services will satisfy your innovative study demands. If you are interested in our services, please don't hesitate to contact us. We are glad to cooperate with you and witness your success!