Methods of de novo drug design by computer-aiding (part two)

  1. Molecular fragmentation

The molecular fragment approach is nicknamed the chemical number holding algorithm or the fragment joining algorithm, and was proposed by Qu et al. in 1991. The basic structure applied by this method is fragments. Each fragment is composed of a single functional group (such as a hydroxyl group, a phenyl group, or a benzene ring, etc.). It is divided into fragment connection methods and fragment growth methods according to the different connection and growth methods of each fragment.

(1) Fragment connection method (or linked-fragment approach): first, there must be a fragment library storing various fragments and a linker library of various linkers. Linkers are -CH2-, -CH2CH2-, -CH2CH2CH2-, -CH = CH-, -C00-, -CONH-, -O-. During the operation, a network is first created in the receiving point region of the receptor. The surface properties of the receiving point are analyzed with probe atoms, such as hydrophobicity, fluorine bonds, static electricity, and van der Waals’ gravitation. Subregions that can accommodate a fragment, such as hydrogen-bonded donors, hydrogen-bonded acceptors, fat hydrophobic, aromatic hydrophobic, and electrostatic interaction zones. Then search the debris library to find the matching shape and quality of the debris. Then search the linker library, find the appropriate linker, connect the fragments of each sub-region, and you can get a complete molecule. A series of molecules produced is optimized by molecular mechanical calculations to select the best ones The structure is for further study.

The software of the fragment connection method includes CAVFAT, SPLICE. HOOK, NEWLEAD and PRO-LIGAND.

(2) Fragment growth (fragment build): similar to the atom build, but at the receptor receiving point according to the nature and shape of the requirements, using fragments instead of atoms to grow one by one to build molecules. The starting point for fragment generation can be the seed atom specified on the recipient point, or the core fragment. Core fragments are fragments that can be attached to the acceptor point and can be obtained from the fragment library, or a fragment of the ligand molecule is selected. Then, fragments are grown according to the amount of energy, and a complete molecule is finally obtained. From a series of structures obtained by optimization of molecular mechanics, a reasonable structure was selected for further research.

Since the fragmentation method uses chemically reasonable fragments as the basic unit for design, the obtained new molecules are easy to accept in structure, so this method has become the mainstream of new drug design.

There are many design software for the fragment growth method, such as LUDI, LEAPFROG, SPROUT, CLIX, GROW, SPLICE, GROMOL and GEMINI.

LUDI (Biosym), which distinguishes four types of interaction sites between receptors and ligands:

(I) Hydrogen-bonded donors and hydrogen-bond acceptors can find sites where hydrogen bonds can be formed; (2) Aliphatic lipophilic sites and aromatic lipophilic sites, which can find suitable positions for hydrophobic interactions.

After determining the site of action, fill in the site with structural debris. Use a certain distance as a search criterion to search for suitable receiving points in groups of 2, 3, and 4 groups. A suitable fragment was retrieved from the library and filled in. If the fragment overlaps with the recipient without Fan Lihua and is electrically rejected, only the distance is appropriate, the fragment can be accepted. “The fragments are connected to form a fragment that is not connected with a real bond, and it is enough to simply lead the compound in the form of one or more points connected, and then artificially interfere, design and optimize the structure of the compound to obtain the lead compound.

LEAPFROG (Tripos) is similar to LUDI, but it can quickly calculate the energy of a large number of candidate compounds and eliminate compounds with inappropriate energy. It has three main functions for post-processing: (1) Structural optimization (Optimize), The lead compound is optimized for structure; (2) Dream, a new lead compound is fictional; (3) Guide, which guides the user’s proposed structure optimization scheme.

4.Homologous protein modeling

Direct drug design must know the three-dimensional structure of the receptor. However, so far, the X aromatic number receptor protein only knows the sequence of amino acids, and its three-dimensional structure is unknown, and there is no method to predict the three-dimensional structure of a protein from the primary structure of the protein.

Homologous protein maintains structural guidance during evolution, that is, homology.

For example, aspartic protease and peniclllopepsin, rhizopuspcpsin, fungal enzymes such as pepsin and cndothiapcpsin, and mammalian enzymes such as porcine pepsin, have a certain degree Homology. From the comparison of their structural sequences, it is not difficult to see that there are structurally conserved regions (SCRs) and variable regions (VRs) different from known structures, which will occupy the same positions of the structurally conserved regions. Arranged to get the structural sequence diagram of the aspartic protease.

Homologous proteins have similar spatial folding patterns and similar biological functions. Because the spatial structure of proteins is more conservative than that of sequences in evolution, sometimes two proteins with a certain sequence difference are still homologous. Therefore, homology can be identified based on the similarity of the spatial structure between proteins. Using the homology of proteins, using homologous proteins with known three-dimensional structure as a template can be used to model homologous proteins. This method is called homologous model building method, comparative molecular simulation method (comparative molecular modeling), homology modeling, or protein homology. Although the homologous molecular method has some errors, it is still a very practical method to obtain the three-dimensional structure of homologous proteins.

Homologous molecules: A homologous protein with a known secondary structure is used as a template. The structure of the target protein with a known sequence and a three-dimensional structure is estimated and modeled.

The basic steps of homologous source protein modeling:

(1) Report the amino acid sequence of an unknown protein. Look for one or several homologous proteins in the protein database. Use as a template for estimating and modeling unknown target proteins

(2) Display and overlap the primary sequence of the target protein and the template protein;

(3) Find structurally conserved regions of the template through structural comparison and sequence analysis;

(4) Find the unknown target molecule and the template of the known structure and divide the common secondary structure region to construct the local main chain structure of the target molecule:

(5) Connect the secondary structure fragments in the target molecule according to the known structure;

Molecular dynamics and molecular dynamics methods were used to optimize the initial model of the model to obtain the structure of the unknown protein.

Software for homologous protein modeling is available from many companies, such as Homology from MSI, Consensus from Biosym, Composer from Tripos, and Insight II, Modeler.

Creative Biolabs provides professional antibody modeling assessment services for antibody modeling. With the company’s most devoted scientists, it is confident to deliver researchers the comprehensive and professional assessment to meet various project development.

Antibody Homology Model

The three-dimensional (3D) structure of antibodies offers an understanding of their function and evolution, and assists in drug design and optimization. When an experimental structure is unavailable, the antibody’s 3D structure is usually obtained through comparative modeling, also known as homology modeling, as well as via de novo computational methods. Commonly, there are four steps to construct a homology model: template selection, template–target sequence alignment, model building, and model evaluation. Besides, many sequence alignment tools and protein structure databases are available to meet the task, such as the Protein Data Bank (PDB), which has more than 1000 crystal structures of antibody fragments (Fab or Fv).