JP4829108B2 - Method and apparatus for creating shape of ligand molecule - Google Patents

Description

  The present invention relates to a method and an apparatus for creating a shape of a ligand molecule that binds to a biopolymer for use in molecular design of physiologically active compounds such as pharmaceuticals and agricultural chemicals.

  In order for a ligand molecule such as a drug or a physiologically active substance to exhibit a physiological activity, the ligand molecule must be a target biopolymer (pharmacological receptor, enzyme, cytokine protein or other protein related to signal transduction between cells, Strong binding to nucleic acids and their complexes). In order to form a strong bond between the biopolymer and the ligand molecule in this way, the biopolymer and the ligand molecule must have complementary shapes, and there must be an affinity between the two molecules. is important. Above all, the shape of the ligand molecule is known as the key and key of the keyhole, and its importance is known.

  When the shape of a ligand molecule for a certain biopolymer is clarified, it is very useful for designing a chemical structure of a novel compound or for screening a compound on a computer (so-called in-silico screening). Drug discovery chemists can design compounds that fit well based on the shape of the ligand molecule. In addition, a drug discovery chemist can perform design based on comparison of ligand shapes of a plurality of biopolymers in order to obtain selectivity between biopolymers. In In-Silico screening, a drug discovery chemist can perform screening using the shape of a ligand molecule. Thus, the shape of the ligand molecule is extremely useful particularly when designing a new drug or physiologically active substance based on the three-dimensional structure of the biopolymer.

  The 3D structure of biopolymers is expected to increase dramatically in the future through genome decoding, 3D structural analysis of biopolymers by national projects such as protein 3000, and homology modeling based on the results. . Accordingly, it is considered that many three-dimensional structures of biopolymers with unknown ligand molecules can be obtained. When the ligand molecule is an unknown biopolymer, it is very difficult to find a ligand that binds to such a biopolymer. In such a case, if the shape of the ligand can be predicted only from the three-dimensional structure of the biopolymer, it is extremely useful for searching for the ligand molecule.

  The shape of the ligand is generally determined by superimposing these ligand molecules based on the structure of a plurality of ligand molecules. However, in such a method, the obtained results differ depending on the conformation or superposition method of the ligand molecules used, and the shape of the ligand cannot always be uniquely predicted. Further, since information on a plurality of ligands is used, it cannot be applied to a biopolymer whose ligand itself is not known.

  When searching for a ligand from the three-dimensional structure of a biopolymer, a method of simulating a stable binding state between the biopolymer and an arbitrary ligand molecule (so-called docking study) is often performed (for example, see Non-Patent Document 1). ). However, since the docking study is a method for searching for a ligand having a possibility of activity and not a method for predicting the shape of the ligand molecule, the shape of the ligand cannot be predicted only from the protein structure.

  Even when the structures of the biopolymer and the ligand are known, it is not easy to predict the shape of the ligand. When trying to predict the shape of a ligand, the structure of a complex of a plurality of ligands and proteins is clarified using the results of, for example, X-ray crystal structure analysis, and these ligands are commonly occupied Must decide. Such work requires a great deal of time and effort.

  When the three-dimensional structure of the biopolymer is known, another method is to roll a sphere with a radius on the surface of the biopolymer and express the shape of the binding site of the biopolymer from the locus drawn by the center of the sphere. Widely known (see, for example, Non-Patent Document 2). However, the shape required by this method merely represents the depression on the surface of the biopolymer and is different from the shape of the ligand.

  As a method using the three-dimensional structure of another biopolymer, there is a method (so-called GRID method) in which a lattice point is created in the biopolymer and a virtual probe is placed there to calculate the interaction with the biopolymer. It is known (for example, see Non-Patent Document 3). When this GRID method is used, the position of a part important for the interaction between the ligand and the biopolymer can be known, but it is difficult to predict the shape of the ligand. Further, as a disadvantage of the GRID method, a number of stable positions of the probe appear on the surface of the biopolymer, and it is not possible to determine which of them is important for binding with the biopolymer. (For example, refer nonpatent literature 4).

Yamada M. et al., J. Mol. Biol., 1994, Vol.243, p.310 Kuntz I D. et al., J. Mol. Biol., 1982, Vol.161, p.269 Goodford P.J. et al., J. Med. Chem., 1985, Vol.28, p.849 Mattos C. et al., Nat. Biotech., 1996, Vol. 14, p.595 Word J M. et al., J. Mol. Biol., 1999, Vol.285, p.1735 Perun T.J., Propst C.L., translated by Yoshiaki Kiso, Akiko Itai, "Chemistry and Biological Experiment Line (Volume 44), Computer Assisted Drug Design", Yodogawa Shoten, 1999, p.289-295)

  As described above, the shape of the ligand is very useful information in drug design (structure-based drug design) based on the three-dimensional structure of biopolymers. A general technique for creating molecular shapes has not been established.

  It is an object of the present invention to provide a method and apparatus for creating a shape of a ligand molecule that binds to a biopolymer.

  Van der Waals force, Coulomb force, and hydrogen bond force are known as interaction forces acting between a biopolymer and a ligand molecule. Coulomb force is a charge interaction, and its value is greatly influenced by the point charge value of the molecule and the dielectric constant of the solvent. The hydrogen bonding force is a force acting between an electrically positive atom and a negative atom, and its value is direction-dependent and is also affected by the dielectric constant of the solvent. On the other hand, the van der Waals force is a force that works universally regardless of the polarity or non-polarity of the molecule and is not affected by the dielectric constant of the surrounding solvent. Give stable results. Therefore, the present inventors paid attention to the van der Waals force, which is a universal interaction, and the use of its potential.

  Furthermore, the present inventors conducted extensive research on a method for creating a ligand shape based on only the three-dimensional structure information of the biopolymer without using the structure information of the ligand molecule. A virtual atom was placed in the region, and a method for calculating the van der Waals potential between the biopolymer and the virtual atom was considered. As a result of further detailed research on the obtained calculation results, it is obtained by removing a specific range of potential energy regions from the van der Waals potential between a biopolymer and a virtual atom, as will be described later. Surprisingly, it has been found that the three-dimensional space agrees very well with the ligand molecule in the biopolymer, and the present invention has been completed.

  The method and apparatus of the present invention predicts the shape of a ligand molecule that binds to the biopolymer based only on the information of the biopolymer without using the information of the ligand molecule, and among the shapes of the ligand molecules, A three-dimensional space that is particularly important for binding with biopolymers can be determined. Therefore, it can be applied not only to biopolymers whose ligand molecules are known, but also to biopolymers whose ligand molecules are not known. The shape of the molecule can be predicted, and among the shapes of the ligand molecules, a three-dimensional space that is particularly important for binding to a biopolymer can be determined.

  In addition, since the method and apparatus of the present invention can create the shape of a ligand molecule using only van der Waals force without considering the interaction due to Coulomb force and hydrogen bond, The shape of the ligand molecule can be predicted without considering the solvent, the type of the biopolymer, and the like, and the three-dimensional space particularly important for binding to the biopolymer can be determined from the shape of the ligand molecule.

  Furthermore, using the predicted shape of the ligand molecule obtained by the method and apparatus of the present invention, In-Silico screening is performed to determine whether the library compound group and the newly designed compound group conform to the shape of the ligand molecule. By doing so, molecular design studies of drug active substances or physiologically active substances can be conducted efficiently.

1 is a flowchart for a first embodiment of the method of the present invention. It is a figure which shows the structure of the apparatus used for the method of this invention. 3 is a flow chart for a second embodiment of the method of the present invention. It is a schematic diagram which shows the relationship between a biopolymer, the shape of a ligand molecule, a ligand binding site, and a ligand binding region. It is a figure which shows the relationship between a biopolymer and a ligand binding area | region in the case of designating a ligand binding area | region as a rectangular parallelepiped. It is a figure which shows the prediction shape of the ligand molecule couple | bonded with the dihydrofolate reductase obtained by Example 1. FIG. It is the figure which piled up the shape of the ligand molecule obtained by Example 1, and the known ligand molecule (methotrexate). FIG. 3 is a diagram in which a three-dimensional space particularly important for binding of a ligand molecule and a known ligand molecule (methotrexate) obtained by Example 1 are superimposed. It is a figure which shows the prediction shape of the ligand molecule | numerator couple | bonded with the retinoic acid receptor gamma obtained by Example 2. FIG. It is the figure which piled up the prediction shape of the ligand molecule obtained by Example 2, and the known ligand molecule (BMS961). It is a figure which shows the prediction shape of the ligand molecule couple | bonded with bacteriorhodopsin obtained by Example 3. FIG. It is the figure which piled up the prediction shape of the ligand molecule obtained by Example 3, and the known ligand molecule (retinal).

Explanation of symbols

10, 20, 30, 40, 50, 60, 70, 80, 90 Step 41, 42, 71, 72 Substep 100 Input unit 210 Molecular data holding unit 220 Parameter holding unit 230 Virtual atom data holding unit 310 E _vdW (fan Delwals potential) calculation unit 320 comparison unit 330 display figure creation unit 400 display unit

  In the present invention, the following terms have the following meanings.

  “Biopolymer” means a protein (for example, a receptor, an enzyme, a cytokine, etc. involved in signal transduction between cells), a nucleic acid, and a complex thereof, and is preferably a protein.

  “Ligand” or “ligand molecule” means a low molecular weight compound (eg, endogenous physiologically active substance, drug, drug candidate compound, etc.) that binds to a biopolymer.

  “Ligand molecule shape” means a three-dimensional space occupied by a ligand molecule when the ligand molecule and a biopolymer targeted by the molecule form a strong bond. That is, it is a three-dimensional space occupied by a ligand molecule when the ligand molecule is strongly bound to a biopolymer.

  A “ligand binding site” is a space where a biopolymer and a ligand bind. However, not all the spaces are filled with ligand molecules (see FIG. 4).

  “Ligand binding region” means a three-dimensional space of any shape that encompasses a ligand binding site.

FIG. 2 is a diagram showing an example of the configuration of a computer device used in the method for creating a shape of a ligand molecule that binds to a biopolymer according to the present invention. The computer device includes an input unit 100, a molecular data holding unit 210, a parameter holding unit 220, a virtual atom data holding unit 230, a van der Waals potential calculation unit 310 (hereinafter referred to as an E _vdW calculation unit), a comparison unit 320, a display A graphic creation unit 330 and a display unit 400 are included.

  The input unit 100 includes a keyboard, a touch panel, a paper tape reader, a paper card reader, a disk (CD, DVD) reader, a magnetic tape reader, a floppy disk reader, and the like, and includes a molecular data holding unit 210, a parameter holding unit. 220, the comparison unit 320, the display graphic creation unit 330, and the like.

  The molecular data holding unit 210 stores the types and coordinates of atoms constituting a biopolymer described later. The molecular data holding unit 210 may be configured by a computer main storage device or an external storage device (a readable / writable recording medium such as a hard disk, a CD, a DVD, a flash memory, a magnetic tape, or a floppy disk).

  The parameter holding unit 220 stores parameters that express biopolymer atoms, which will be described later, and parameters that express virtual atoms. The parameter holding unit 220 may be configured by a computer main storage device or an external storage device (a readable / writable recording medium such as a hard disk, a CD, a DVD, a flash memory, a magnetic tape, and a floppy disk).

  In the virtual atom data holding unit 230, the three-dimensional coordinates of the three-dimensional lattice points generated in the ligand binding region, which will be described later, the van der Waals potential at each three-dimensional lattice point, and each three-dimensional lattice point are removed by the comparison unit. A flag indicating this is stored. The virtual atom data holding unit 230 may be configured by a computer main storage device or an external storage device (a readable / writable recording medium such as a hard disk, a CD, a DVD, a flash memory, a magnetic tape, or a floppy disk).

The _EvdW calculation unit 310 includes a CPU, a computer program stored in a storage medium, and a working storage area. Here, the computer-readable “storage medium” storing the computer program may be a CD, DVD, flash memory, magnetic tape, floppy disk, or the like in addition to the hard disk. The working storage area is mainly composed of a main storage device of a computer, but a readable / writable recording medium such as a hard disk, a CD, a DVD, a flash memory, a magnetic tape, or a floppy disk may be used in combination. The E _vdW calculation unit 310 includes the types and coordinates of each atom constituting the biopolymer stored in the molecular data holding unit 210, parameters representing each atom of the biopolymer stored in the parameter holding unit 220, and virtual The parameters representing the atoms and the three-dimensional coordinates of the three-dimensional lattice points stored in the virtual atom data holding unit 230 are read out, and the information is applied to the formula (I) to obtain the van der Waals of the three-dimensional lattice points. Calculate the potential. The calculation result is recorded in the virtual atom data holding unit 230 in association with the coordinates of the three-dimensional lattice point.

  The comparison unit 320 includes a CPU, a computer program stored in a storage medium, and a working storage area. The working storage area is mainly composed of a main storage device of a computer, but a readable / writable recording medium such as a hard disk, a CD, a DVD, a flash memory, a magnetic tape, or a floppy disk may be used in combination. The comparison unit 320 compares the van der Waals potential of the three-dimensional lattice points stored in the virtual atom data holding unit 230 with a threshold value, and sets a flag indicating whether or not each three-dimensional lattice point should be removed as a virtual atom. Record in the data holding unit 230. Alternatively, the data of the three-dimensional lattice points to be removed may be deleted from the virtual atom data holding unit 230. The threshold value may be read from the input unit 100.

  The display graphic creation unit 330 includes a CPU, a computer program stored in a storage medium, and a work storage area. The working storage area is mainly composed of a main storage device of a computer, but a readable / writable recording medium such as a hard disk, a CD, a DVD, a flash memory, a magnetic tape, or a floppy disk may be used in combination. The display graphic creation unit 330 creates a predicted shape of the ligand molecule using the 3D coordinates of the 3D lattice points stored in the virtual atom data holding unit 230 and displays the predicted shape on the display unit 400. Further, the three-dimensional structure of the biopolymer may be displayed together with the predicted shape of the ligand molecule using the three-dimensional coordinates of the atoms of the biopolymer held in the molecular data holding unit 210. Further, R1 used when creating the predicted shape of the ligand molecule may be read from the input unit 100.

Here, the CPUs configuring the _EvdW calculation unit 310, the comparison unit 320, and the display graphic creation unit 330 may be the same or different.

  The display unit 400 may be configured by any visualization means known in the art, such as a display, a printer, or a plotter.

  Note that the apparatus shown in FIG. 2 may be implemented in a distributed computing environment. In this case, the input unit 100 may be configured as a communication interface and various data may be acquired via a network.

  Next, a ligand molecule shape creation method (first embodiment of the present invention) executed by the apparatus configured as described above will be described with reference to the flowchart of FIG. 1 and FIG.

(First step)
In the first step 10 of FIG. 1, parameters representing biopolymer atoms are read from the storage device. The “storage device” in the present invention includes a reading device for a computer-readable recording medium such as a hard disk, CD, DVD, flash memory, magnetic tape, and floppy disk, and a buffer memory including the input unit 100 such as a keyboard. Including. The parameter expressing the atoms of the biopolymer may be determined by selecting a set of descriptors of the atoms of the biopolymer, generally known as a force field, or may be determined uniquely. . In the present invention, the CHARMm force field or the Amber force field is selected and described as a parameter expressing the atoms of the biopolymer, but the present invention is well known to those skilled in the art instead of these force fields. The same can be done using other force fields. The atomic parameters of the biopolymer read in this step are stored in the parameter holding unit 220 in FIG.

(Second step)
In the second step 20 of FIG. 1, the virtual atom parameters corresponding to the force field are read from the storage device. The Van der Waals potential (E _vdW ) in the CHARMm force field and the Amber force field is expressed by the following formula (I):

(In the formula, v represents a virtual atom, i represents the i-th atom of the biopolymer, r represents the distance between the center of the virtual atom and the center of the biopolymer atom, and VR represents each atom. (Represents the van der Waals radius (angstrom). Vε represents the depth (Kcal / mol) of the van der Waals potential well of each atom.)
Is calculated using

In order to calculate this potential energy, parameters Vε and VR of virtual atoms are defined. As a result of extensive research, the present inventors have found that the parameters of the virtual atom can be defined with reference to the parameters of the methyl group of the selected force field. Specifically, the Vipushiron _v virtual atoms the same value as the Vipushiron methyl group, 85% to 100% of the VR of the methyl groups of the VR _v virtual atoms, preferably it is desirable to 90% to 95% in I found. Furthermore, as a result of intensive studies, the present inventors have found that it is not necessary to change these values regardless of the biopolymer. That is, in the present invention, it is not necessary to consider indeterminate conditions surrounding the biopolymer in the living body, and it is necessary to change the parameter of the virtual atom unless the parameter (force field) expressing the biopolymer is changed. The second step is basically performed once. In the case of using two or more virtual atoms, each VR _v can be selected appropriately different values from the range of 85% to 100% of the VR methyl group. The virtual atom parameters Vε _v and VR _v may be input from the input unit, or may be automatically generated from the parameters stored in the parameter holding unit 220 in the first step 10. The virtual atom parameters Vε _v and VR _v defined in this step are also stored in the parameter holding unit 220.

  In addition, the first and second steps are performed in advance, and parameters expressing the biopolymer atoms and virtual atoms are recorded on a read-only medium (CD-ROM, DVD-ROM, ROM, EPROM, EEPROM, etc.), This may be used as the parameter holding unit 220.

(Third step)
In the third step 30, the types of atoms constituting the biopolymer and their coordinates are read from the storage device. Information on the atoms constituting the biopolymer can be input according to various methods well known to those skilled in the art. For example, a method of calculating from a three-dimensional structure obtained by X-ray crystal analysis or NMR analysis; a method of obtaining using a protein crystal database or the like; or a biopolymer model (homology modeling) constructed based on such information A model) can be used. In the information obtained according to these methods, the coordinates of hydrogen atoms may be unknown. In this case, information is added to the coordinates of hydrogen atoms using known software such as Insight II (registered trademark), Reduce, etc. (see, for example, Non-Patent Document 5). In addition, information on cofactors that bind to biopolymers and play an important role (eg, coenzymes, water molecules, metal ions, etc.) is further input as biopolymer information, and the subsequent steps are performed. Also good.

  Reading in this step is performed at the input unit and may be performed manually using a keyboard or by reading information recorded in advance on a removable recording medium (floppy disk, CD, DVD, magnetic tape, etc.). Or may be performed by downloading information via a communication line such as the Internet. The read data of the types and coordinates of the atoms constituting the biopolymer are stored in the molecular data holding unit 210.

(4th process)
The fourth step 40 includes the following two sub-steps 41 and 42.

(4-1)
In the first substep 41, a ligand binding region is designated. This ligand binding region can designate any part of the biopolymer. For example, it can be specified as an arbitrary shape (a sphere, a rectangular parallelepiped, or an arbitrary polyhedron) including a ligand binding site (see FIG. 4). In the present invention, hereinafter, the ligand binding region is described as being specified as a rectangular parallelepiped. However, in the present invention, as long as the ligand binding region can be specified in relation to the atomic coordinate information of the biopolymer, the ligand binding region specifying method is used. There is no limit. For example, the region including the entire biopolymer input in the third step may be automatically specified by a computer, or information (atomic species, interatomic bonds, coordinates, etc.) of the biopolymer is displayed on the display device. The ligand binding region may be designated interactively. The designated ligand binding region is stored in the molecular data holding unit 210.

(4-2)
Next, in the second sub-step 42, a plurality of three-dimensional lattice points are set in the ligand binding region specified in the sub-step 41 using the coordinate information of the biopolymer atoms input in the third step 30. generate. A three-dimensional lattice point means a point composed of three-dimensional coordinates generated at regular intervals. The three-dimensional coordinates for designating the three-dimensional lattice points may be based on an arbitrary coordinate system such as an orthogonal coordinate system (xyz), a polar coordinate system (rθφ), or a cylindrical coordinate system (rθz). It is convenient to use coordinates based on the system. An interval between generated three-dimensional lattice points (a distance from a specific lattice point to the closest adjacent lattice point) is at least equal to or less than VR _{v of} the virtual atom read in the second step, and VR _v is When it is within the preferred range described in the second step (within 85% to 100% of the VR of the methyl group of the force field used), the spacing between the three-dimensional lattice points is preferably 0.5 to 1.0 angstrom. . The generated three-dimensional coordinates of the plurality of three-dimensional lattice points are stored in the virtual atom data holding unit 230.

(5th process)
In the fifth step 50, a virtual atom is placed at one of a plurality of three-dimensional lattice points generated in the fourth step (sub-step 42), and the biopolymer atoms read in the first step 10 are represented. Then, the van der Waals potential of the virtual atom is calculated using the parameter of the virtual atom read out in the second step 20. The calculated value is stored as van der Waals potential of each three-dimensional lattice point. This calculation is repeated for all the three-dimensional lattice points generated in the fourth step (sub-step 42). In this step, each three-dimensional lattice point is made to coincide with the center of the virtual atom. Specifically, the E _vdW calculation unit 310 calculates the type and coordinates of each atom constituting the biopolymer stored in the molecular data holding unit 210 and each atom of the biopolymer stored in the parameter holding unit 220. The parameters (VR _i , Vε _i ) to be represented, the parameters (VR _v , Vε _v ) representing the virtual atom, and the three-dimensional coordinates of the three-dimensional lattice point stored in the virtual atom data holding unit 230 are read out, and information about this is read out. _Is applied to the formula (I) to calculate the van der Waals potential E _vdW of the three-dimensional lattice point. Then, the calculated E _vdW is recorded in the virtual atom data holding unit 230 in association with the coordinates of the three-dimensional lattice point.

(Sixth step)
In the sixth step 60, among the plurality of three-dimensional lattice points generated in the fourth step (sub-step 42), a three-dimensional lattice point having a van der Waals potential _{equal to} or higher than the first threshold E _th1 shown below is used. Remove. By performing this step, the position of the virtual atom that is not stable, that is, the position of the virtual atom that is not desirable for bonding with the biopolymer can be removed. The present inventors have found that in order to remove such not stable virtual _atom, an _{E th1,} -7 to-11 times the absolute value of Vipushiron _v, even more preferably -9 absolute value of Vipushiron _v It has been found that it is preferable to set to -10 times. This setting is effective when the atoms of the biopolymer are expressed in the CHARMm force field or the Amber force field in the first step 10 and the parameters of the virtual atoms are within the range described in the second step 20. In this step, _V ε _v stored in the parameter holding unit 220 may be read to generate E _th1 automatically.

In this step, the comparison unit 320 compares the set E _th1 with the E _vdW calculated for each three-dimensional lattice point, and searches for a three-dimensional lattice point that satisfies E _th1 <E _vdW . The comparison unit 320 receives the input from the virtual atom data holding unit 230 and compares E _vdW with E _th1 . At this time, when a plurality of virtual atoms having different VR _v are used, there may be a plurality of inputs from the virtual atom data holding unit 230, but the smallest value, that is, the most stable value is represented by a fan on the three-dimensional lattice point. It is desirable to have a Delwars potential. Regarding the removal of the three-dimensional lattice points that satisfy the above condition, the data that is stored in the virtual atom data holding unit 230 is flagged (recorded) to indicate whether or not the above condition is satisfied. It is convenient to show the data. Alternatively, data that satisfies the above conditions may be deleted from the virtual atom data holding unit 230.

(Seventh step)
The seventh step 70 includes the following two sub-steps 71 and 72.

(7-1)
In the first sub-step 71, the display graphic creation unit 330 generates a sphere having a radius Rl around each of the three-dimensional lattice points that remain without being removed in the sixth step. When obtaining the shape of the ligand molecule, including hydrogen atoms, if VR _v is within the range described in the second step, setting the R1 to 100-90% of VR _v (preferably 100%) Is preferred. When obtaining the shape of the ligand molecule that do not include a hydrogen atom, VR If _v is in the range described in the second step, is the R1 to 1.0 Å to 0.9 Å (preferably from VR _v 0. It is preferable to set to a value obtained by subtracting 95 angstroms). It may be configured to read VR _v from the parameter holding unit 220 and automatically generate R1. The display graphic creation unit 330 receives the input from the virtual atom data holding unit 230 and generates a spherical aggregate having the radius R1 as described above.

(7-2)
Next, in the second sub-step 72, the display graphic creation unit 330 makes the overlapping portion of the spherical aggregate generated in the sub-step 71 unique. That is, hidden line processing is performed to obtain a three-dimensional outline of a spherical aggregate. As a result, the resulting three-dimensional space is a prediction result of the shape of the ligand molecule that binds to the biopolymer.

  Then, the obtained prediction result can be displayed on the display unit 400. The display unit 400 may be configured to display the three-dimensional structure of the biopolymer using the information on the biopolymer held in the molecular data holding unit 210 together with the prediction result of the shape of the ligand molecule. Alternatively, the input unit 100, the display graphic creation unit 330, and the display unit 400 may be linked to display the prediction result and / or the shape of the biopolymer viewed from the viewpoint designated by the operator.

In the above description, the case where the first threshold value _Eth1 used for comparison in the sixth step is set in advance has been described. However, it goes without saying that the value used as E _th1 can be input from the input unit 100. Alternatively, a plurality of E _th1 is set, and the sixth and seventh steps are performed for each to display the results. Based on the display and chemical considerations, an appropriate E _th1 for the biopolymer is determined by the operator. May be selectable. The operator may select an appropriate one from the plurality of displayed _Eth1 , or may input an appropriate value from the input unit 100. Then, using _Eth1 selected or input by the operator, the sixth and seventh steps can be performed again to obtain a predicted shape of the ligand molecule.

  Furthermore, using the prediction result of the shape of the ligand molecule obtained by the above method, screening of candidate compounds that match the shape can be performed. Specifically, by superimposing the three-dimensional structure of an arbitrary compound on the predicted shape of the ligand molecule and evaluating its fitness using various indices (for example, shared volume, etc.), the compound becomes a candidate compound. It can be screened for appropriateness. Such superposition of the three-dimensional structure and evaluation of the fitness can be performed using known software capable of three-dimensional search using the shape as a search condition. For example, when using Catalyst (registered trademark), candidate compounds can be screened by the three-dimensional search function of Catalyst (registered trademark) using the shape of the ligand molecule as a prediction result as a search condition. At this time, if necessary, general chemical characteristics (for example, hydrogen bonding region, hydrophobic region, etc.) based on chemical considerations are added to the shape of the ligand molecule obtained according to the method of the present invention, and the degree of fitness. Can also be evaluated. By applying such screening to a library compound group or a newly designed compound group, it becomes possible to efficiently screen candidate compounds that bind to a biopolymer.

  FIG. 3 shows a method for determining a three-dimensional space particularly important for binding to a biopolymer, in addition to predicting the shape of a ligand molecule that binds to the biopolymer, which is a second embodiment of the present invention. It is the shown flowchart. The method of the second embodiment can also be performed using the apparatus shown in FIG. Hereinafter, the second embodiment of the present invention will be described in detail.

  In the method of the second embodiment, first, the first to seventh steps of the first embodiment are performed. Next, the following steps are further performed to determine which region of the ligand molecule shape is particularly important for creating a strong bond when the ligand molecule and the biopolymer are bound. .

(8th step)
In the eighth step 80, the three-dimensional lattice points having van der Waals potential _{equal to} or higher than the second threshold _Eth2 shown below are further removed from the three-dimensional lattice points remaining without being removed in the sixth step 60. To do. This step makes it possible to determine a three-dimensional space that is particularly important for binding with biopolymers. If the atoms of the biopolymer are expressed by the CHARMm force field or the Amber force field in the first step 10 and the parameters of the virtual atom are in the range described in the second step 20, E _th2 (<0) is expressed as Vε _v -12 times the absolute value or less, preferably set below -14 times (i.e., the absolute value of E _th2, more than 12 times the absolute value of Vipushiron _v, preferably set to more than 14 times the ) Was found to be preferable.

This step is performed in the comparison unit 320, and the set E _th2 is compared with the E _vdW calculated for each three-dimensional lattice point to search for a three-dimensional lattice point that _satisfies E _th2 > E _vdW . The comparison unit 320 receives the input from the virtual atom data holding unit 230 and compares E _vdW with E _th1 . At this time, when a plurality of virtual atoms having different VR _v are used, there may be a plurality of inputs from the virtual atom data holding unit 230, but the smallest value, that is, the most stable value is set to the fan on each three-dimensional lattice point. It is desirable to have a Delwars potential. For the removal of the three-dimensional lattice points that satisfy the above conditions, the data stored in the virtual atom data holding unit 230 is flagged (recorded) to indicate whether or not the above conditions are satisfied, and the removed data Is convenient.

(9th step)
Next, the ninth step 90 is performed. In this step, the display graphics creation unit _330, E _th2> for three-dimensional grid point as a _{E vdW} repeat the same sub-steps and sub-steps 71 and 72 described in the seventh _step, the E _{th2> E vdW} A three-dimensional outline of a collection of spheres having a radius R1 centered on each of the three-dimensional lattice points is obtained. The value of R1 in this step is preferably the same as the value of R1 used in the fifth step. As a result, the resulting three-dimensional space is a particularly important three-dimensional space for the binding between the biopolymer and the ligand molecule.

  A three-dimensional space that is particularly important for binding to the obtained biopolymer can be displayed on the display unit 400. In addition to the three-dimensional space, the display unit 400 may be configured to display the prediction result of the shape of the ligand molecule and / or the three-dimensional shape of the biopolymer held in the molecular data holding unit 210. When displayed in combination with the prediction result of the shape of the ligand molecule, the prediction result of the three-dimensional space particularly important for the binding to the biopolymer is an attribute (brightness, color, Display with different degrees, hue changes, display by different hatching, blinking display, highlight display, and the like. In addition, the input unit 100, the display figure creation unit 330, and the display unit 400 cooperate to predict the shape of the ligand molecule from the viewpoint specified by the operator. A particularly important three-dimensional space and / or the shape of a biopolymer may be displayed.

In the above description, the case where the second threshold E _th2 used for the comparison in the eighth step 80 is set in advance has been described. However, it goes without saying that the value used as _Eth2 can be input from the input unit. Alternatively, a plurality of E _th2 is set, and the eighth and ninth steps are performed for each to display the results. Based on the display and chemical considerations, the operator determines an appropriate E _th2 for the ligand molecule. You may make it selectable. The operator may select an appropriate one from the plurality of displayed _{Eth 2} , or may input an appropriate value from the input unit 100. Then, the eighth and ninth steps are performed again using _Eth2 selected or input by the operator so as to obtain the shape of the three-dimensional space particularly important for the binding to the biopolymer among the shapes of the ligand molecules. can do.

  The method and apparatus of the present invention described above predicts the shape of a ligand molecule that binds to the biopolymer based on only the information of the biopolymer without using the information of the ligand molecule, and Among the shapes, it is possible to determine a three-dimensional space that is particularly important for binding to biopolymers. Therefore, it can be applied not only to biopolymers whose ligand molecules are known, but also to biopolymers whose ligand molecules are not known. The shape of the molecule can be predicted, and among the shapes of the ligand molecules, a three-dimensional space that is particularly important for binding to a biopolymer can be determined.

  In addition, since the method and apparatus of the present invention creates the shape of the ligand molecule using only van der Waals forces, the ligand molecule can be formed without considering the solvent around the biopolymer and the type of the biopolymer. One can predict the shape and / or determine the three-dimensional space of the ligand molecule that is particularly important for binding to the biopolymer.

  Furthermore, using the shape of the ligand molecule obtained by the method and apparatus of the present invention, in-silico screening is performed to determine whether the library compound group or the newly designed compound group matches the shape of the ligand molecule. This makes it possible to efficiently conduct molecular design studies on drug active substances or physiologically active substances.

  The method of the present invention will be described in more detail with reference to examples, but the present invention is not limited to these contents.

(Shape of ligand molecule that binds to dihydrofolate reductase)
A specific example for carrying out the method of the present invention will be described using dihydrofolate reductase. Dihydrofolate reductase is one of the enzymes that have been studied for a long time, and several X-ray crystal structures have been analyzed. For the atomic coordinates of dihydrofolate reductase (hereinafter referred to as DHFR), information of entry ID “1RX3” of the protein data bank was used. Further, as a descriptor of these atoms, a general CHARMm Version 22 (hereinafter, CHARMm) was used as a force field for a biopolymer.

In CHARMm, the van der Waals radius (VR) of a methyl atom is defined as 2.165 angstroms, and the depth (Vε) of a potential well is defined as −0.1811 Kcal / mol. Therefore, the virtual atom VR _v for calculating the van der Waals potential is 1.95 angstrom (about 90% of the VR of the methyl atom), and Vε _v is −0.1811 Kcal / mol (100% of the methyl atom Vε). did.

  Since the three-dimensional coordinates of the DHFR do not have coordinate information of hydrogen atoms, the coordinates of hydrogen atoms were added using software Reduce. The ligand binding site of DHFR is a rectangular parallelepiped (X coordinate = 28.0 to 37.0, Y coordinate = 37.0 to 47.0) including the peripheral amino acid residues based on information known in the literature. , Z coordinate = 0.0 to 14.0) (see FIG. 5).

  Next, three-dimensional lattice points with a 1 angstrom interval are generated in the ligand binding region of this rectangular parallelepiped, and the van der Waals potential on each lattice point is calculated using virtual atoms, and the van der Waals force used in CHARMm is calculated. Calculated according to formula (I).

Subsequently, as the first threshold value _{E th1} set the -1.8Kcal / mol (about -10 times the absolute value of Vε _v), to remove the three-dimensional grid point with _{E th1} larger van der Waals potential, A sphere having a radius of R1 = 1.0 angstrom (VR _v −0.95 angstrom) was generated at each of the remaining three-dimensional lattice points, and the shape of the ligand molecule was created with unique overlapping portions. The shape of the obtained ligand molecule is shown in FIG.

  In addition, in order to examine the validity of the shape of the ligand molecule obtained according to the method of the present invention, the three-dimensional coordinates of methotrexate obtained from a known DHFR / methotrexate co-crystal (protein data bank entry ID “1RX3”) When this methotrexate was overlapped with the shape of the ligand molecule obtained by the method of the present invention without moving the coordinates, it was confirmed that the methotrexate agreed very well as shown in FIG.

Further, as a second threshold value _{E th2} set the -2.53Kcal / mol (about -14 times the absolute value of Vε _v), to remove the three-dimensional grid point with _{E th2} larger van der Waals potential, remains In addition, a sphere having a radius R1 = 1.0 angstrom was generated at each of the three-dimensional lattice points, the overlapping portion was made unique, and a three-dimensional space important for coupling with DHFR was determined. FIG. 8 is a diagram in which the obtained three-dimensional space and methotrexate are similarly superimposed without moving the coordinates. The three-dimensional space thus determined coincides with the region occupied by the pyrimidine ring of methotrexate. From the results of X-ray crystallographic analysis, this pyrimidine ring moiety plays a major role in forming a stable DHFR-methotrexate complex by forming a number of hydrogen bonds and hydrophobic bonds with DHFR, and lacks the activity of methotrexate. It is known that the partial structure cannot be used (see Non-Patent Document 6). Thus, pyrimidine which is a partial structure important for the binding to DHFR among the three-dimensional space particularly important for the binding to the biopolymer obtained according to the method of the present invention using the second threshold and the molecular structure of methotrexate. It was confirmed that the ring portion was in good agreement.

(Shape of ligand molecule binding to retinoic acid receptor gamma)
As a specific example of creating a shape of a ligand molecule that binds to a nuclear receptor, an example using retinoic acid receptor gamma will be described. Retinoic acid receptor gamma (hereinafter RAR) is one of the nuclear hormone receptors and is known to be involved in skin diseases and cancer. For the RAR atomic coordinates, information of the entry ID “4LBD” of the protein data bank was used. Further, as a descriptor of these atoms, a general CHARMm Version 22 (hereinafter, CHARMm) was used as a force field for a biopolymer.

In this embodiment, 1.95 Å VR _v virtual atoms for calculating the van der Waals potential (methyl atomic about 90% VR), the Vε _v -0.1811Kcal / mol (methyl atomic Vipushiron 100%).

  Since there is no hydrogen atom coordinate information in the three-dimensional coordinates of RAR, the coordinates of hydrogen atoms were added using software Reduce as in Example 1. The ligand binding site of 4LBD is a rectangular parallelepiped (X coordinate = -31.0-47.0, Y coordinate = 10.0-21. 0, Z coordinate = 78.0-90.0).

  Next, three-dimensional lattice points with a 1 angstrom interval are generated in the ligand binding region of this rectangular parallelepiped, and the van der Waals potential on each lattice point is calculated using virtual atoms, and the van der Waals force used in CHARMm is calculated. Calculated according to formula (I).

Subsequently, the first threshold E _th1 is set to −1.8 Kcal / mol (approximately −10 times the absolute value of Vε of the virtual atom), and three-dimensional lattice points having a van der Waals potential greater than E _th1 are removed. Then, a sphere having a radius of 1.0 angstrom (VR _v -0.95 angstrom) was generated at each of the remaining three-dimensional lattice points, and the shape of the ligand molecule was created by making the overlapping portion unique. The shape of the obtained ligand molecule is shown in FIG.

  Further, in order to examine the validity of the shape of the ligand molecule obtained according to the method of the present invention, a known RAR / RAR agonist (BMS961: Chemical name 3-fluoro-4- [2-hydroxy-2- (5,5 , 8,8-Tetramethyl-5,6,7,8-tetrahydronaphthalen-2-yl) acetylamino] benzoic acid) co-crystal (protein data bank entry ID “4LBD”) tertiary of RAR agonist When the original coordinates were used and this RAR agonist was superimposed on the shape of the ligand molecule obtained by the method of the present invention without moving the coordinates, it was confirmed that the RAR agonists were in excellent agreement as shown in FIG. .

(The shape of the ligand molecule that binds to bacteriorhodopsin)
As a specific example of creating a shape of a ligand molecule that binds to a seven-transmembrane protein, an example using bacteriorhodopsin will be described. As the coordinate information of the bacteriorhodopsin, the information of the entry ID “1BRR” of the protein data bank is used. In addition, Amber 91, which is a general force field for biopolymers, was used as a descriptor for these atoms.

In Amber 91, the van der Waals radius (VR) of a methyl atom is defined as 2.165 angstroms, and the depth (Vε) of a potential well is defined as 0.181 Kcal / mol. Therefore, the virtual atom VR _v for calculating the van der Waals potential is 1.95 angstrom (about 90% of the methyl atom VR) and Vε _v is 0.181 Kcal / mol (100% of the methyl atom Vε). .

  Since there is no coordinate information of hydrogen atoms in the three-dimensional coordinates of the bacteriorhodopsin, the coordinates of hydrogen atoms were added using software Reduce as in Example 1. The ligand binding site of “1BRR” is a rectangular parallelepiped (X coordinate = 9.0 to 17.0, Y coordinate = −15.0 to 0.0, Z coordinate = 19.0-29.0).

  Next, three-dimensional lattice points with an interval of 1 angstrom are generated in the ligand binding region of the rectangular parallelepiped, and the van der Waals potential on each lattice point is calculated using virtual atoms, and the van der Waals force used in Amber is calculated. Calculated according to formula (I).

Subsequently, the first threshold value E _th1 is set to −1.8 Kcal / mol (about −10 times the absolute value of Vε _v ), and three-dimensional lattice points having a van der Waals potential greater than E _th1 are removed, A sphere having a radius of 1.0 angstrom (VR _v −0.95 angstrom) was generated at each of the remaining three-dimensional lattice points, and the shape of the ligand molecule was created with the overlapping portion unique. The shape of the obtained ligand molecule is shown in FIG.

  Further, in order to examine the validity of the shape of the ligand molecule obtained according to the method of the present invention, the three-dimensional coordinates of the retinal obtained from a known bacteriorhodopsin / retinal co-crystal (protein data bank entry ID “1BRR”) When this retinal was superimposed on the shape of the ligand molecule obtained by the method of the present invention without moving the coordinates, it was confirmed that the retinal matches very well as shown in FIG.

By using the method and apparatus of the present invention, the shape of a ligand molecule that binds to a biopolymer can be created. Therefore, the method and apparatus of the present invention are extremely useful for molecular design of compounds having physiological activity such as pharmaceuticals and agricultural chemicals.

Claims (8)

A method for creating a shape of a ligand molecule that binds to a biopolymer comprising:
(1) a first step of reading out a parameter expressing atoms of a biopolymer from a storage device;
(2) a second step of reading parameters of one or more virtual atoms from the storage device;
(3) a third step of reading the atomic coordinates of the biopolymer;
(4) a fourth step of generating a plurality of three-dimensional lattice points in the ligand binding region of the biopolymer read out in (3);
(5) When the virtual atom whose parameter is read in (2) is placed on each of the plurality of three-dimensional lattice points generated in (4), van der Waals between the virtual atom and the biopolymer The fifth step of calculating the potential;
(6) Sixth step of removing three-dimensional lattice points whose van der Waals potential calculated in (5) is larger than the first threshold; and (7) The shape of the ligand molecule from the three-dimensional lattice points remaining in (6). A method for creating a shape of a ligand molecule that binds to a biopolymer, characterized in that the seventh step of creating a biomolecule is performed. A method for predicting the shape of a ligand molecule that binds to a biopolymer, and to determine the three-dimensional space of the ligand molecule shape that is particularly important for binding to the biopolymer:
(1) a first step of reading out a parameter expressing atoms of a biopolymer from a storage device;
(2) a second step of reading parameters of one or more virtual atoms from the storage device;
(3) a third step of reading the atomic coordinates of the biopolymer;
(4) a fourth step of generating a plurality of three-dimensional lattice points in the ligand binding region of the biopolymer read out in (3);
(5) When the virtual atom whose parameter is read in (2) is placed on each of the plurality of three-dimensional lattice points generated in (4), van der Waals between the virtual atom and the biopolymer The fifth step of calculating the potential;
(6) A sixth step of removing a three-dimensional lattice point whose van der Waals potential calculated in (5) is larger than a first threshold;
(7) A seventh step of creating the shape of the ligand molecule from the three-dimensional lattice points remaining in (6);
(8) Eighth step of removing a three-dimensional lattice point whose van der Waals potential calculated in (5) is larger than a second threshold; and (9) Living body of a ligand molecule from the three-dimensional lattice point remaining in (8). The ninth step of determining a three-dimensional space that is particularly important for binding to the polymer is further performed, and the shape of the ligand molecule that binds to the biopolymer is characterized in that the second threshold value is smaller than the first threshold value. A method of predicting and determining a three-dimensional space that is particularly important for binding to a biopolymer among ligand molecule shapes.   The virtual atom parameters read in the second step include the depth of the virtual atom van der Waals potential well, and the first threshold is the absolute depth of the virtual atom van der Waals potential well defined in the second step. The method of claim 1, wherein the method is in the range of −7 to −11 times the value.   The virtual atom parameters read in the second step include the depth of the virtual atom van der Waals potential well, and the first threshold is the absolute depth of the virtual atom van der Waals potential well defined in the second step. 3. A method according to claim 2, characterized in that it is in the range of -7 to -11 times the value.   The method according to claim 4, wherein the second threshold value is −12 times or less the absolute value of the depth of the van der Waals potential well of the virtual atom. A device for creating a shape of a ligand molecule that binds to a biopolymer:
(1) a parameter holding unit that holds parameters representing atoms and virtual atoms of a biopolymer;
(2) a molecular data holding unit for holding the coordinates of the atoms of the biopolymer;
(3) By using the parameters held in the parameter holding unit and the coordinates of the biopolymer atoms held in the molecular data holding unit, 1 is added to a plurality of three-dimensional lattice points in the ligand binding region of the biopolymer. A van der Waals potential calculator that calculates a van der Waals potential between the virtual atom and the biopolymer when a seed or two or more types of virtual atoms are placed;
(4) A virtual atom data holding unit that holds the three-dimensional coordinates of the plurality of three-dimensional lattice points and the van der Waals potential calculated by the van der Waals potential calculation unit;
(5) The van der Waals potential held in the virtual atom data holding unit is compared with a first threshold value, and a three-dimensional lattice point having a van der Waals potential larger than the first threshold value is determined as the virtual atom data. A comparison unit to be removed from the holding unit; and (6) a display figure creation unit for creating a shape of a ligand molecule from a three-dimensional lattice point held in the virtual atom data holding unit and not removed by the comparison unit. A device characterized by that. The comparison unit compares the van der Waals potential held in the virtual atom data holding unit with a second threshold value, and determines a three-dimensional lattice point having a van der Waals potential larger than the second threshold value as the virtual value. A second function of removing from the atomic data holding unit;
The display graphic creation unit is held in the virtual atom data holding unit, and from the three-dimensional lattice points that are not removed by the second function, out of the shape of the ligand molecule that binds to the biopolymer, The apparatus according to claim 6, further comprising a function of creating a three-dimensional space particularly important for the combination of the two. A method for screening candidate compounds that bind to a biopolymer comprising:
(1) a first step of reading out a parameter expressing atoms of a biopolymer from a storage device;
(2) a second step of reading the virtual atom parameters from the storage device;
(3) a third step of reading the atomic coordinates of the biopolymer;
(4) a fourth step of generating a plurality of three-dimensional lattice points in the ligand binding region of the biopolymer read out in (3);
(5) When the virtual atom whose parameter is read in (2) is placed on each of the plurality of three-dimensional lattice points generated in (4), van der Waals between the virtual atom and the biopolymer The fifth step of calculating the potential;
(6) A sixth step of removing a three-dimensional lattice point whose van der Waals potential calculated in (5) is larger than a first threshold;
(7) A seventh step of creating the shape of the ligand molecule from the three-dimensional lattice points remaining in (6);
(8) A candidate for binding to a biopolymer, characterized in that the eighth step of superimposing the shape of the ligand molecule obtained in (7) and the three-dimensional structure of the candidate compound and evaluating its fitness is performed. Compound screening method.

Images