JP4688467B2 - Method for searching structure of receptor-ligand stable complex - Google Patents

Description

  The present invention relates to a search method using a computer for finding a stable complex structure of a receptor molecule and a ligand molecule that can be used for molecular design of physiologically active substances such as pharmaceuticals and agricultural chemicals. The present invention also relates to a program for executing the method.

Many drugs and physiologically active substances exhibit their pharmacological activity or toxicity by binding to a biopolymer (protein, nucleic acid and complex containing them) serving as a receptor at a specific site.
With regard to the structure of biopolymers that are candidates for receptors, the progress of analytical equipment such as X-rays and NMR and the software that analyzes the results enables structural determination to be made in a shorter period of time than ever before. It is registered and published in a database such as a data bank. On the other hand, deciphering the entire base sequence of a gene becomes a reality, and due to advances in molecular modeling technology using computers, a certain degree of three-dimensional structure is estimated even for biopolymers whose structure has not been experimentally determined It is possible.

Compounds such as relatively low molecular weight drugs and inhibitors that bind to such biopolymers are generally called ligands, and they efficiently form ligands that can form stable conjugates with target receptors. This is an important foundation for drug discovery research. Enter the coordinates of the biopolymer and ligand, specify the approximate binding site (binding pocket) in the biopolymer, find the appropriate binding mode with the receptor using a computer, and By applying the calculation to multiple ligands in succession, the calculation method to discover ligands that may bind or to screen out compounds that do not bind is a drug discovery research using a computer. It is widely known as a method for improving efficiency. These calculation methods are generally called docking studies, and are also called virtual screening when a large number of ligands are compared at high speed at the expense of a certain degree of accuracy. In addition, such research methods for designing ligand molecules based on the three-dimensional structure of the receptor are collectively referred to as structure-dependent drug discovery. In order to apply these methods to actual molecular systems, it is necessary to develop an algorithm that satisfies the following two criteria: (i) efficient search of coordinate space, and (ii) evaluation criteria for evaluating the validity of binding modes. It is essential and many methods have been proposed. However, considering not only the complex conformation of the ligand molecule but also the degree of freedom including the relative position that the molecule can exist in the receptor, the degree of freedom of the space to be searched in (i) is enormous. Therefore, it is not practical to conduct a complete exhaustive search even with the current computational capability. In addition, there is not yet a complete evaluation function for quantifying the criterion (ii). In this way, although it is clear that a complete search by calculation is theoretically limited, by utilizing such a method at the initial stage of drug discovery research, actual synthesis experiments and pharmacological activity tests can be performed. Reducing the number is extremely effective for shortening the development period and reducing the cost of R & D, so much research has been done (Patent Document 1 or Non-Patent Document 1), and the main ones are published as software packages. Or commercially available (Tripos FlexX, Accelrys C ² -LigandFit, Schlodinger Glide, etc.).
Patent No. 2621842 DesJarlais RL, et al., J. Med. Chem. 29, 2149-2153 (1986)

As mentioned above, docking studies generally have the same effect as exploring all coordinate spaces, effectively simplifying the infinite degree of freedom that consists of a combination of ligand conformations and the coordinates in which the ligand can exist. It is an indispensable requirement to increase the calculation speed and accuracy.

In Patent No. 2621842, as shown in FIG. 1, clustering of dummy atoms set at the position of a heteroatom that can be a hydrogen bonding partner of a hydrogen bonding functional group of a receptor (biopolymer) is performed. And a receptor-ligand characterized in that it is determined based on the hydrogen bonding mode and conformation whether or not this matches with a hydrogen bonding part consisting of a hydrogen bonding hetero atom of a ligand that has been conformationally analyzed. A stable complex structure search method is disclosed.
However, this method is (i) the calculation time becomes very long in a system with many matching hydrogen bonding moieties, and (ii) the classification is focused on the hydrogen bonding mode, so it is estimated from the receptor structure. The hydrogen bonding mode is slightly different from that of the ligand, and may actually miss the ligand structure that may exist as a complex, or may generate a structure in which atoms other than hydrogen-bonding heteroatoms collide with the acceptor atom. . These are essential problems due to assumptions based on the chemical characteristics of the receptor (simplification of the receptor structure). It is necessary to select or redefine only the structure, and there is a limit to the scope of application as a fully automatic docking study by a computer.
Given this situation, rather than limiting the intermolecular interaction between the ligand and the receptor to the hydrogen bonding mode, it would be possible to efficiently find a candidate ligand structure that simply conforms to the shape of the binding pocket. It is considered to be a desirable method for conducting a fully automatic docking study.

  In view of the situation as described above, the present inventors, as a result of intensive studies, estimated the position of the binding pocket in the biopolymer and determined the candidate structure of the ligand that can be bound in the binding pocket as a ligand atom in the pocket. A virtual dummy atom is defined at the coordinates where the can exist and the position of the surrounding acceptor atom, and various conformations of the ligand generated by a computational chemical method in advance are adapted to this dummy atom model. In this way, we have succeeded in developing a method for accurately estimating the arrangement in which a ligand may exist as a stable complex in a receptor in a short time. Furthermore, the present inventors have found that the conformation and the binding mode can be determined simultaneously by continuously performing this configuration estimation on the conformations of various ligands that have been generated in advance, leading to the completion of the present invention. .

That is, the present invention is as follows.
(1) A method for searching for a stable complex structure of a receptor and a ligand using a computer, wherein a combination of a receptor and a ligand is received by executing the following first to seventh steps. A method characterized by identifying a structure in which a body and a ligand can exist stably:
(A) a first step of comprehensively generating conformations that can be taken by a ligand;
(B) In a three-dimensional structural model of the receptor, a coordinate method is used to comprehensively search for coordinates where a sphere with a radius of 0.5 to 5 接触 can be in contact with at least four heavy atoms in the receptor molecule. A second step of arranging a dummy atom S having a radius of 1 to 3 Å at the coordinate position,
(C) a third step of setting a bond pocket including at least one dummy atom by clustering the dummy atoms S on the basis of coordinates;
(D) In the binding pocket specified in the third step, select the acceptor heavy atom existing in the range of 3 to 20 mm around the dummy atom S, and set −3 Å <r <−0.5 に at the coordinate position. A fourth step of setting a dummy atom X having a virtual radius r in the range of
(E) Randomly rotate the ligand molecule so that one heavy atom in the ligand conformation generated in the first step overlaps one randomly selected from the dummy atoms S defined in the second step And a fifth step of translating,
(F) calculating the degree of overlap between the ligand atom and the dummy atoms S and X as a score, and rotating and translating the entire ligand molecule to select a conformation that maximizes the score;
(G) A seventh step in which the fifth to sixth steps are repeated to select the relative conformation of the receptor and the ligand that form a stable complex.
(2) In the third step, furthermore, all the distances between the dummy atoms S in the binding pocket are calculated, and when there are a plurality of dummy atoms close to each other within 0.1 to 2.0 mm, one of them is left. 2. The method for searching a receptor-ligand stable complex structure according to (1) , wherein other adjacent dummy atoms are deleted.
(3) The method for searching a receptor-ligand stable complex structure according to (1) , wherein in the sixth step, only heavy atoms in the ligand are considered as ligand atoms.
(4) Further, the arrangement of the ligands obtained in the seventh step as an initial structure, and obtains the interaction energetically stable structure of the receptor by molecular mechanics calculations, acceptance (1) Body-ligand stable complex structure search method.
(5) A program for causing an information processing apparatus to execute processing for searching for a stable complex structure of a receptor and a ligand, and for executing the following procedure:
(A) a first procedure for comprehensively generating possible conformations of a ligand;
(B) In a three-dimensional structure model of the receptor, a coordinate method is used to comprehensively search for coordinates where a sphere with a radius of 0.5 to 5 mm can be in contact with at least four heavy atoms in the receptor molecule. And a second procedure for arranging a dummy atom S having a radius of 1 to 3 に at the coordinate position,
(C) a third procedure for setting a bond pocket including at least one dummy atom by clustering the dummy atoms S on the basis of coordinates;
(D) In the binding pocket defined by the third procedure, select the acceptor heavy atom in the range of 3 to 20 Å around the dummy atom S, and set −3 Å <r <− A fourth procedure to set a dummy atom X with a virtual radius r in the range 0.5 、
(E) Randomize the ligand molecules so that one heavy atom in the ligand conformation generated by the first procedure overlaps one randomly selected from the dummy atoms S defined by the second procedure. The fifth step of rotating and translating to
(F) a sixth procedure for calculating the degree of overlap between the ligand atom and the dummy atoms S and X as a score, and selecting the conformation that maximizes the score by rotating and translating the entire ligand molecule;
(G) A seventh procedure in which the treatment by the fifth procedure and the sixth procedure is repeated, and the relative conformation of the receptor and the ligand forming a stable complex is selected.
(6) The program according to (5) , further including a procedure for obtaining a structure that is stable in terms of interaction energy with a receptor by molecular mechanics calculation using the arrangement of the ligand obtained by the seventh procedure as an initial structure.

As described above, in the docking study using a computer, the present invention automatically specifies a stable complex structure in a short time from only the structure of a receptor and a ligand. Moreover, despite the exhaustive search of the structure without relying on the user's experience and knowledge, the speedup applicable to virtual screening has been achieved, and a versatile and powerful docking technique is provided.
By using the novel search method disclosed in the present invention, it is possible to specify a stable complex structure in a short time even for a receptor whose binding site is not specified. Therefore, not only the development of new bioactive substances, but also the possibility of finding new pharmacological activities or causing side effects and toxicity by examining whether known ligands form stable complexes with other receptors. Therefore, the cost and time of drug discovery research can be greatly improved.

  In the present invention, the receptor is not particularly limited as long as it is a substance capable of binding a ligand, and examples thereof include high molecular compounds such as proteins, nucleic acids, and sugars. The receptors may be natural ones, modified ones thereof, or artificial ones synthesized by imitating them. The receptor preferably has a known three-dimensional structure or can easily estimate a three-dimensional structure based on the structure of a receptor having a known three-dimensional structure. When a protein is used as a receptor, the three-dimensional structure of the protein can be obtained from, for example, Protein Data Bank Japan (PDB, http://pdbj.protein.osaka-u.ac.jp/). In the present invention, one or more receptors may be used for analysis.

A ligand is generally a small molecule compared to a receptor, and is a concept that includes all compounds that specifically bind to a specific site in a receptor, regardless of whether it is a natural product or a chemically synthesized substance. Ligands include, for example, drugs or enzyme substrates or inhibitors. In general, a ligand sometimes refers to a substance that binds to an active site of a receptor and exerts a pharmacological action. However, in the present invention, a ligand may be any substance that can form a stable complex with a receptor. However, it is not always necessary to bind to the active site and exert a pharmacological action.
In addition, the dummy atom has a size, mass, etc. in order to use it as a standard for processing such as movement and superposition of molecules to relative positions in space when handling molecular structures on a computer. It is a virtual atom that is arbitrarily set and generally has no potential parameter.
The binding pocket refers to an inner hole or depression that is observed on the molecular surface of the biopolymer and has a size that allows ligand molecules to bind.

The method of the present invention will be described step by step with reference to the flowchart of FIG. In the first step, first, the conformation that the ligand can take is comprehensively generated. A single ligand may be used as the ligand, or a plurality of ligands may be used.
Various known methods can be used as a method for efficiently generating various conformations that can be taken by the ligand. For example, for a rotatable bond in a ligand, it is sequentially rotated at a certain angle to comprehensively generate all structures, randomly changing dihedral angles and atomic coordinates, and then structure by molecular mechanics calculation Stochastic search for optimization (Ferguson, DM, Raber, DJ, J. Am. Chem. Soc. 111, 4371-4378 (1989)), structural optimization by short-term molecular dynamics and molecular mechanics calculations at high temperatures And simulated annealing (Nilges, M., et al., FEBS Lett., 239, 129-136 (1988)) that samples various conformations that may be obtained by long-term observation Can be mentioned. In addition, commercial programs that extend these methods, such as the Conformation Import function in MOE (trade name) manufactured by Canada Chemical Computing Group, and the polling method (Smellie, A., Teig, SL, and Towbin, P., J. Comp. Chem., 16, 171-187 (1995)) can be used to search various conformational spaces in a short time.

In the second step, in the three-dimensional structural model of the receptor, coordinates that a sphere of a certain size can exist in contact with at least four heavy atoms in the receptor molecule are comprehensively determined by a geometric method. Search and place a dummy atom S with a radius of 1 to 3 Å at the coordinate position.
As a geometric method for exhaustively searching for the coordinates of a hypothetical sphere that may exist in contact with any four heavy atoms in the receptor, the method by Edelsbrunner et al. (Edelsbrunner, H., et al. , Proceedings of the 28th Hawaii International Conference on Systems Science. 256-264 (1995)) and MOE manufactured by Chemical Computing Group based on this algorithm
The Alpha Site Finder installed in the.
The size of a hypothetical sphere called an alpha sphere used in this method is preferably a radius of 0.5 to 5 mm, more preferably 1 to 2 mm. Since the center coordinates of the sphere thus determined are close to at least four atoms in the biopolymer, it is in the depression of the biopolymer and sufficient for the heavy atoms of the ligand to be present. It can be regarded as the center of a simple space.
Next, a dummy atom S having a radius of 1 to 3 に is placed at this coordinate position. The dummy atoms S can be arbitrarily set at the coordinate position, or a plurality of dummy atoms may be set.

In the third step, a bond pocket including at least one dummy atom is set by clustering the dummy atoms S on the basis of coordinates. Clustering can be performed based on the coordinates of the dummy atoms. For example, clustering can be performed by grouping together dummy atoms within a distance of 1.0 to 5.0 mm, preferably 1.5 to 3.5 mm.
In the third step, for a bond pocket containing a plurality of dummy atoms S, all distances between the dummy atoms S in the bond pocket are calculated, and when there are a plurality of dummy atoms close to each other within 0.1 to 2.0 mm, It is preferable to delete other dummy atoms in the vicinity, leaving one.

  In order for a receptor and a ligand to form a stable complex, it is important that the atoms constituting the ligand can adopt a conformation that does not overlap with the receptor. In order to take this into consideration, in the fourth step, acceptor heavy atoms existing in the range of about 3 to 20 周 囲 around the dummy atom S in the binding pocket defined in the third step are selected, and their coordinate positions are selected. To set a dummy atom X with a virtual radius r in the range −3 Å <r <−0.5 Å.

  Next, the ligand shape conformed in the first step is superimposed on the dummy atom group set in the second to fourth steps, and the degree of the overlap is estimated, thereby matching the ligand shape to the binding pocket. Assess sex.

Specifically, in the fifth step, one heavy atom in the ligand conformation generated in the first step overlaps with one randomly selected from the dummy atoms S defined in the second step. Rotate and translate the ligand molecule randomly.
In the sixth step, the degree of overlap between the ligand atom and the dummy atoms S and X is calculated as a score, and the conformation that maximizes the score is selected by rotating and translating the entire ligand molecule. In this step, it is preferable to consider only heavy atoms in the ligand as ligand atoms.
In the seventh step, the fifth to sixth steps are repeated to select the relative conformation of the ligand that forms a stable complex with the receptor.

ここで、m、nは、それぞれダミー原子とリガンド原子の数、r_i, r_jは、それぞれ、ダミー原子、リガンド原子の原子半径、 α は Gauss 型関数の減衰定数、dは、ダミー原子とリガンド原子の中心間距離を表す。 As a way to maximize the electrostatic charge and volume overlap for different molecules, Kearsley and
Smith's SEAL method (Kearsley, S., & Smith, GM, Tetrahedron Computer Methodology, 3, 615-633, (1990)).
The degree of overlap can be easily calculated by defining, for example, the following function as a function of the distance d between the dummy atom and the ligand atom and the radius r of the atom.

Here, m and n are the number of dummy atoms and ligand atoms, r _i and r _j are the atomic radii of dummy atoms and ligand atoms, respectively, α is the attenuation constant of the Gaussian function, and d is the dummy atom. Represents the distance between the centers of the ligand atoms.

In this function, the effect of overlap between distant atoms becomes smaller as α is smaller. The value of α is generally 0.1 to 1, preferably 0.3 to 0.7. Also, the smaller d is, the larger the value is given. When d = 0, the product of the set atomic radii, r _i × r _j , is taken. Therefore, f
By moving the ligand molecules in parallel and rotationally so that the value of is maximized, a more overlapping arrangement can be found. In actual programming, it is also possible to employ a method for searching for a condition for minimizing a value obtained by multiplying f by −1 by using a general numerical analysis method.
Furthermore, here, since the dummy atom X is stipulated to have a virtually negative radius, if there is a ligand atom overlapping with X, negative contribution to the value of f depends on the degree of overlap. Will do. Therefore, by calculating the sum of the degree of overlap of each atom constituting the ligand molecule to the dummy atom S and the dummy atom X, and maximizing the value, it is adapted to the shape of the active pocket in the receptor, and It is possible to simultaneously find the arrangement of ligands that do not hit the wall.
The numerical analysis methods used to minimize -f used here include the Steepest Descent (SD) method, the Conjugate Gradients (CG) method, the Newton Rapthon method (NR), and the Truncated Newton (TN) method. There is a minimum value can be obtained in a short time by continuously executing these alone or in combination of two or more.
The ligand structure that is a candidate for a stable complex that is output by this calculation takes a shape that fits well into the active pocket as a whole of the ligand molecule, even if there is a slight overlap with the wall surface of the active pocket.

The structure of the receptor is not necessarily fixed, and the structure of the receptor may change somewhat due to complex formation with the ligand. It is also a well-known fact that the coordinates of the receptor obtained by techniques such as X-ray and NMR are normal proteins and have an error of about 1 to 3 Å.
Therefore, the ligand structure obtained by the method of the present invention is considered to be appropriate as an initial structure for finding the most stable complex structure. However, if necessary, it can be fine-tuned together with the receptor structure by molecular mechanics calculation. can do. Specifically, a stable complex structure can be determined by optimizing the ligand-receptor complex structure by molecular mechanics calculation on the list of ligand structures obtained in the seventh step. it can. Since the accuracy in molecular mechanics calculation depends on the force field to be used, it is necessary to use a force field having parameters corresponding to both the biopolymer and the low-molecular ligand in the docking calculation. Commonly available force fields include CVFF (P. Douber-Osguthorpe et al ,, Proteins: Struct. Funct. Genet., Vol. 4, p31, 1988), CFF (MJHwang et al, J. Am. Chem. Soc , vol 116, p2515, 1994), OPLS-AA (Jorgensen WL, et al ,, J. Am. Chem. Soc., vol. 117, p11225-11236, 1996), MMFF94, MMFF94s, MMFF94x (all TAHalgren , J. Comp. Chem. Vol. 17, No. 5 & 6, p490, 1996). In addition, molecular mechanics calculation engines including these parameters include Discover (registered trademark, Accelrys Inc. (USA)), CHARMM (Cornell WD, et al ,, J. Am. Chem. Soc., Vol. 117, p5179. , 1995), MOE-Minimizer (trademark, Chemical Computing Group Inc. (Canada)), Direct Force Field (trademark, Aeon Technology, Inc. (US)), etc., but are not limited thereto.
Furthermore, the first step to the seventh step or the subsequent molecular mechanics calculation step are continuously performed for a plurality of ligands and / or a plurality of receptors, so that the receptors and the ligands exist stably. By specifying a possible structure, the biological activity of the ligand can be estimated from the discovery of a candidate compound for a novel physiologically active substance or the known function of a receptor to which the ligand can bind.

The present invention usually applies the process defined by the present invention to an appropriate computer language for a file containing sufficient data to reproduce the three-dimensional structure of a molecule on a computer, such as receptor and ligand coordinates, element names, etc. This is implemented by executing the program described in. The computer language is not particularly defined as long as it is a language normally used for scientific calculation, such as C, C ++, FORTRAN, JAVA, SVL (CCG Scientific Vector Language). In addition, as a program for carrying out the present invention, not only a consistent program that describes all the processes is used, but also a combination of published programs that have a function of performing each process or a part thereof, It is also possible to use a script that calls a function that implements the above.

In order to show the effectiveness of the method of the present invention, a program capable of continuously executing the steps of the present invention is created, and a receptor and a ligand are separated from a complex having a known three-dimensional structure. By applying this program, it was confirmed that a known ligand-receptor complex structure can be reproduced in a short time.
Here, the program was created using the language SVL and functions installed in the computational chemistry system MOE (2004.03 version) manufactured by Chemical Computing Group, Inc. In molecular mechanics calculations, we used Merck force field (MMFF94x), which has parameters for a wide range of ligands as well as biopolymers and has high structure reproducibility. For parameters that are still insufficient, a rule-dependent force field that complements this Merck force field can be used to automatically calculate any intermolecular interaction.

Of the complexes registered in the PDB, entries 1AQW, 1C5C, 1CBS, 1CVU, 1D0L, 1D3D, 1D3P, 1DG5, 1EI1, 1BXO, 1CKP, 1DD7, 1F0R, 1F0S, 1LIC, 3ERT were conducted. Here, the procedure is explained using a complex of 3ERT estrogen receptor and ligand OHT (4-Hydroxy-tamoxifen) as an example.
The three-dimensional structure registered in the protein data bank is determined by X-ray and does not include the coordinates of hydrogen atoms. Before applying the method of the present invention, this structure is
After reading into MOE, adding hydrogen atoms, the position of hydrogen atoms was determined by molecular mechanics calculation. At this time, in order to maintain the structure obtained by the experiment, the coordinates of atoms other than hydrogen were fixed, and only the position of the hydrogen atom was optimized by molecular mechanics calculation.

  In the first step, the conformation_import function installed in the MOE was used to comprehensively generate conformations that can be taken by the ligand OHT, and 100 types were stored in ascending order of energy (Fig. 2).

The search for bond pockets, dummy atom placement, and clustering in the second and third steps are performed using the MOE-Alpha Site Finder function, which is an extension of Edelsbrunner et al. (Edelsbrunner, H., et al., 1995). Coordinates that can enter a sphere with a radius of 1.4 to 1.8 入 り were determined, and dummy atoms with the same radius as the carbon atoms, without energy, were placed at all positions. All the distances between the dummy atoms were measured, and those within a distance of 2.5 に し て were grouped into a binding pocket. In this system, it was divided into 21 groups (clusters) (Fig. 3). Of these, all but the dummy atoms contained in the largest pocket were deleted.
In the third step, all the distances between the dummy atoms were also measured, and the extra dummy atoms were deleted so that all the dummy atoms had no other dummy atoms within a distance of 0.8 mm.

  In the fourth step, as the excluded volume region, the coordinates of atoms other than the hydrogen atom of the acceptor that are within a distance of 4.5 か ら from all dummy atoms are detected, and on each coordinate, the same absolute coordinates as those of the heavy atoms are detected. A virtual dummy atom having a radius whose value is negative is arranged (FIG. 4).

The conformation with the lowest energy generated in the first step was overlaid with this dummy atom model according to the fifth and sixth steps, and 50 arrangements that could become stable complexes were found (FIG. 5). .
Next, RMSD (square root of mean square distance) of the ligand structure and other conformations obtained in the first step was calculated, and the fifth and sixth steps were similarly performed for the conformations having the most different structures. This operation was repeated until 200 initial structures were output.

  This list of initial structures was rearranged according to the size of the score f, and only 100 structures with good scores were sequentially placed in the receptor, and the structure was optimized by molecular mechanics calculation. The smaller the total of the internal energy of the ligand and the interaction energy of the ligand and the receptor, the more stable the complex, so the structures were rearranged in order from the lowest total energy. In this system, the structure with the lowest total energy has an RMSD of 0.59 compared to the data determined by X-ray crystal structure analysis, confirming that the experimental data is well reproduced (Fig. 5). ).

In addition, combinations of other ligands and receptors were also analyzed by the same method. The results are shown in Table 1. Best RMSD is the best square root of the root mean square of the distance between the ligand's heavy atoms in the calculated results and the X-ray crystal structure. Etotal is the sum of the total interaction energy of the structure and the internal energy of the ligand. , Score is the score of the structure f, Best RMSD No. in Score is the rank in the score of the structure with the best RMSD, Best RMSD No. in Etotal is the rank in the Etotal of the structure with the best RMSD, and Time is Shows the time taken to calculate on an Intel Xeon 3.06 GHz x 2 machine.
According to this, the structure of the Best RMSD obtained by the method of the present invention is ranked higher in the scores f and Etotal, and the complex structure of the ligand and the receptor can be reduced in a short time by using the method of the present invention. It was found that it can be predicted accurately.

The figure which shows the flowchart of the method of this invention. The figure which shows the example of the various ligand conformation of the ligand OHT which generate | occur | produced at the 1st process of Example 1. FIG. The figure which shows the molecular model of the dummy atom and receptor which were obtained at the 2nd process of Example 1. FIG. The solid line represents the receptor estrogen receptor, the sphere represents the alpha sphere replaced with a carbon atom in a space-filling model, and the set represents the binding candidate pocket. The figure which shows the ASE model which consists of the alpha sphere obtained at the 3rd process of Example 1, and the exclusion volume obtained at the 4th process. White is the alpha sphere and black is the excluded volume. A part of the excluded volume in front of the alpha sphere is not shown. The figure which shows the example of the coupling | bonding structure candidate (sphere bar model) which rotated and translated the initial arrangement | positioning of the ligand obtained at the 5th process of Example 1 so that a score might become the maximum at a 6th process. White globules are alpha spheres and black globules indicate excluded volume. A part of the excluded volume before and after the alpha sphere is not shown. The ligand atoms are in the region of the alpha sphere and do not overlap the excluded volume. The stereo figure of the structure (bar model) of the most stable complex which minimized the interaction energy about the score maximum arrangement | positioning of the ligand obtained at the 6th process of Example 1. FIG. The solid line is the receptor, the sphere model is the X-ray crystal structure of the ligand, and the surface is the pocket surface of the receptor in contact with the alpha sphere. Receptor atoms in front of the ligand and the back side of the pocket surface are not shown.

Claims (6)

A method of searching for a stable complex structure of a receptor and a ligand using a computer, wherein the receptor and the ligand are performed by performing the following first to seventh steps for the combination of the receptor and the ligand. A method characterized by identifying structures in which can exist stably:
(A) a first step of comprehensively generating conformations that can be taken by a ligand;
(B) In a three-dimensional structural model of the receptor, a coordinate method is used to comprehensively search for coordinates where a sphere with a radius of 0.5 to 5 接触 can be in contact with at least four heavy atoms in the receptor molecule. A second step of arranging a dummy atom S having a radius of 1 to 3 Å at the coordinate position,
(C) a third step of setting a bond pocket including at least one dummy atom by clustering the dummy atoms S on the basis of coordinates;
(D) In the binding pocket defined in the third step, select the acceptor heavy atom in the range of 3 to 20 周 囲 around the dummy atom S, and set −3 Å <r <−0.5 at the coordinate position. A fourth step of setting a dummy atom X having a virtual radius r in the range of Å
(E) Randomly rotate the ligand molecule so that one heavy atom in the ligand conformation generated in the first step overlaps one randomly selected from the dummy atoms S defined in the second step And a fifth step of translating,
(F) calculating the degree of overlap between the ligand atom and the dummy atoms S and X as a score, and rotating and translating the entire ligand molecule to select a conformation that maximizes the score;
(G) A seventh step in which the fifth to sixth steps are repeated to select the relative conformation of the receptor and the ligand that form a stable complex. In the third step, all the distances between the dummy atoms S in the binding pocket are further calculated, and when there are a plurality of dummy atoms that are close to each other within 0.1 to 2.0 mm, one of them is left and the other The method for searching a receptor-ligand stable complex structure according to claim 1 , wherein the dummy atoms are deleted. The receptor-ligand stable complex structure searching method according to claim 1 , wherein in the sixth step, only heavy atoms in the ligand are considered as ligand atoms. The receptor according to claim 1 , further comprising obtaining a structure that is stable in terms of interaction energy with the receptor by molecular mechanics calculation using the arrangement of the ligand obtained in the seventh step as an initial structure. -Ligand stable complex structure search method. A program for causing an information processing apparatus to execute processing for searching for a stable complex structure of a receptor and a ligand, and for executing the following procedure:
(A) a first procedure for comprehensively generating possible conformations of a ligand;
(B) In a three-dimensional structural model of the receptor, a coordinate method is used to comprehensively search for coordinates where a sphere with a radius of 0.5 to 5 接触 can be in contact with at least four heavy atoms in the receptor molecule. And a second procedure for arranging a dummy atom S having a radius of 1 to 3 に at the coordinate position,
(C) a third procedure for setting a bond pocket including at least one dummy atom by clustering the dummy atoms S on the basis of coordinates;
(D) In the binding pocket defined by the third procedure, select the acceptor heavy atom in the range of 3 to 20 Å around the dummy atom S, and set −3 Å <r <− A fourth procedure to set a dummy atom X with a virtual radius r in the range 0.5 、
(E) Randomize the ligand molecules so that one heavy atom in the ligand conformation generated by the first procedure overlaps one randomly selected from the dummy atoms S defined by the second procedure. The fifth step of rotating and translating to
(F) a sixth procedure for calculating the degree of overlap between the ligand atom and the dummy atoms S and X as a score, and selecting the conformation that maximizes the score by rotating and translating the entire ligand molecule;
(G) A seventh procedure in which the treatment by the fifth procedure and the sixth procedure is repeated, and the relative conformation of the receptor and the ligand forming a stable complex is selected. Furthermore, the program of Claim 5 which has the procedure which calculates | requires the structure stable in interaction energy with a receptor by molecular mechanics calculation by making arrangement | positioning of the ligand obtained by the said 7th procedure into an initial structure.

Images