JP7470437B2 - Structure prediction method and structure prediction device - Google Patents

Description

Patent Law Article 30, paragraph 2 applied May 9, 2019 22nd Theoretical Chemistry Symposium Poster presentation abstracts posted http://www.rkk-web.jp/theochem22/asset/program/poster.pdf May 28, 2019 22nd Theoretical Chemistry Symposium Hokkaido University Academic Exchange Hall May 16, 2019 Journal of the American Chemical Society 2019 Paper posted https://pubs.acs.org/doi/10.1021/jacs. 9b01541 June 5, 2019 Journal of the American Chemical Society, 2019, Vol. 141, No. 22, pp. 8675-8679 September 1, 2019 Symposium "New Frontiers in Chemical Reaction Path Search 2019" Online poster abstract https://iqce.jp/SRPS/SRPS2019/P08.pdf September 16, 2019 Symposium "New Frontiers in Chemical Reaction Path Search 2019" Nagoya University Noyori Memorial Hall

The present invention relates to a technology for predicting molecular structures.

In recent years, with the improvement of computer processing power, research into computational chemistry and quantum chemistry, which use computers to predict the properties of structures such as molecular structures, has become active. In computational chemistry and quantum chemistry, it is possible to predict the structure and properties of a substance on a computer without actually creating the substance.

Various tools have been developed to predict the structures and properties of such materials, and various research and development projects have been carried out using these tools (see, for example, non-patent document 1).

Yukihiro Ohta, "Automated screening of molecular structures with computational chemistry", Research Organization for Information Science and Technology, RIST News No. 62 (2017), 2017, pp. 11-20

The objective of the present invention is to provide a structure prediction technology that efficiently searches for molecular structures having stable structures.

In order to solve the above problem, one aspect of the present invention relates to a structure prediction method having the steps of: a processor acquiring molecular coordinates of a molecule; the processor detecting a first minimum point on the potential energy surface of the molecule based on a gradient from the molecular coordinates and determining a stable structure corresponding to the first minimum point; the processor detecting one or more second minimum points within a search range based on the stable structure and determining a group of stable structures corresponding to the one or more second minimum points; and the processor determining a most stable structure based on the potential energy of each molecular coordinate of the group of stable structures.

According to the present invention, a structure prediction technology can be provided that efficiently searches for molecular structures having stable structures.

FIG. 2 is a schematic diagram showing a molecular structure prediction process according to an embodiment of the present invention. 1 is a block diagram showing a hardware configuration of a structure prediction device according to an embodiment of the present invention. 1 is a flowchart showing a structure prediction process according to an embodiment of the present invention. FIG. 2 is a schematic diagram showing a molecular structure of a prediction target according to an embodiment of the present invention. FIG. 2 illustrates molecular coordinates according to one embodiment of the present invention. FIG. 1 shows a peptide molecule to which metal ions are coordinated based on a priority order according to one embodiment of the present invention. FIG. 2 is a schematic diagram showing a procedure for searching for a stable structure by a structure prediction process according to an embodiment of the present invention. FIG. 13 shows experimental results of conformational search according to one embodiment of the present invention.

In the following examples, a structure prediction method and apparatus are disclosed for searching for a stable structure from a given molecular structure, such as a metal-coordinated molecular structure.

[Structure Prediction Device]
To summarize the embodiment described later, as shown in FIG. 1, in the structure prediction device according to the embodiment described below, a structure optimization calculation (e.g., PM6, density functional theory (DFT), etc.) is performed on the input model molecular coordinates, and a stable structure corresponding to a minimum point near the molecular coordinates on the potential energy surface is searched for. When a stable structure is determined, a conformation search calculation (e.g., GRRM (Global Reaction Route Mapping), PM6, etc.) is performed within a search range based on the determined stable structure, and a stable structure group (multiple stable structures) corresponding to one or more other minimum points is determined. Then, based on the potential energy of each molecular coordinate of the determined stable structure group, the most stable structure having the minimum potential energy is determined.

In this way, conformational search calculations, which require a large amount of computation, are performed by narrowing the search range based on stable structures rather than searching the entire range, making it possible to efficiently predict molecular structures.

Here, the structure prediction device 100 may be, for example, a computing device such as a server or a personal computer, and may have a hardware configuration as shown in Fig. 2. That is, the structure prediction device 100 has a drive device 101, an auxiliary storage device 102, a memory device 103, a CPU (Central Processing Unit) 104, an interface device 105, and a communication device 106, which are interconnected via a bus B.

Various computer programs including programs or instructions for implementing various functions and processes described below in the structure prediction device 100 may be provided by a recording medium 107 such as a CD-ROM (Compact Disk-Read Only Memory). When the recording medium 107 storing the programs is set in the drive device 101, the programs are installed from the recording medium 107 via the drive device 101 into the auxiliary storage device 102. However, the programs do not necessarily have to be installed by the recording medium 107, and may be downloaded from any external device via a network or the like.

The auxiliary storage device 102 stores installed programs as well as necessary files and data. When an instruction to start a program is received, the memory device 103 reads out and stores the program or data from the auxiliary storage device 102. The auxiliary storage device 102 and the memory device 103 are realized as non-transitory computer-readable storage media that store programs or instructions. The CPU 104, which functions as a processor, executes various functions and processes of the structure prediction device 100 according to various data such as the program stored in the memory device 103 and parameters required to execute the program. The interface device 105 is used as a communication interface for connecting to a network or an external device. The communication device 106 executes various communication processes for communicating with an external device.

However, the structure prediction device 100 is not limited to the above-mentioned hardware configuration and may be realized by any other suitable hardware configuration, such as, for example, one or more circuits that realize one or more of the functions and processes performed by the structure prediction device 100.

[Structure prediction processing]
Next, a structure prediction process according to an embodiment of the present disclosure will be described with reference to Figures 3 to 7. The structure prediction process is realized by the above-mentioned structure prediction device 100, and may be realized, for example, by a processor of the structure prediction device 100 executing a program or instructions. Figure 3 is a flowchart showing the structure prediction process according to an embodiment of the present invention.

As shown in FIG. 3, in step S101, the structure prediction device 100 acquires molecular coordinates of a molecule. For example, the molecule may constitute a reactant (peptide structure, coordinated metal, etc.) represented by a molecular formula as shown in the upper part of FIG. 4, a solvent (water, methanol, acetonitrile, etc.) as shown in the middle part of FIG. 4, or a product of a molecular structure as shown in the lower part of FIG. 4 (14-nuclear cyclic complex, etc.). The molecular coordinates may also indicate the three-dimensional positions of each atom constituting the molecule as shown in FIG. 5.

Here, a restriction condition due to coordination may be set. In one embodiment, the molecular coordinates may represent the position of each atom of a molecular structure formed by coordinating a metal ion to the coordination site of a given molecular structure. Specifically, the coordination site of the metal may be fixed, and the positions of the atoms (elements) that bond, including the metal, may be indicated by the coordinates. In another embodiment, the molecular coordinates may represent the position of each atom of a molecular structure formed based on distance information between atoms. Specifically, without coordinating a metal, distance information between atoms other than the metal may be set using CONFLEX or the like.

In one embodiment, the coordination site and coordination priority of the metal ion coordinated to the molecule may be determined by the selected metal ion. That is, the type of metal ion, the coordination site, and the coordination priority may be specified as the conditions related to the provision. Here, any metal is possible as the type of metal ion, and the coordination site and the coordination priority can be determined according to the selected metal ion. For example, a selected metal ion (e.g., nickel) may be coordinated as a ligand to a peptide molecule as a receptor as shown in FIG. 6 according to the coordination priority "1", "2", ...

In one embodiment, if the coordination priority is unknown, the complexing ability of a model compound may be examined and entered. Compounds may be synthesized through experiments, and their complexing abilities may be obtained as a result of analysis and specified as conditions.

In step S102, the structure prediction device 100 detects a minimum point on the potential energy surface of the molecule based on the gradient from the molecular coordinates, and determines a stable structure corresponding to the minimum point. Specifically, the structure prediction device 100 may determine a stable structure (or an equilibrium structure) by a structural optimization calculation such as PM6 or DFT. For example, when the molecular coordinates acquired in step S101 correspond to the input molecular structure on the potential energy surface shown in FIG. 7, the structure prediction device 100 may detect a minimum point based on the gradient from the molecular coordinates by a structural optimization calculation such as PM6 or DFT, and determine a stable structure (before conformation search) from the optimized molecular coordinates corresponding to the detected minimum point.

In step S103, the structure prediction device 100 detects one or more other minimum points within a search range based on the stable structure, and determines a group of stable structures corresponding to the one or more other minimum points. Specifically, the structure prediction device 100 uses conformational search calculations such as GRRM and PM6 to search for other minimum points on the potential energy surface within a predetermined search range from the stable structure determined in step S102, and obtains a group of stable structures (multiple stable structures) composed of molecular structures corresponding to each detected minimum point. For example, the structure prediction device 100 may determine a stable structure after multiple conformational search by conformational search calculations such as GRRM and PM6 from the stable structure determined in step S102, as shown in FIG. 7.

The determination of a stable structure can be made based on the energy value obtained by calculation. The energy threshold value in this case is, for example, 5 kJ/mol, but may be changed appropriately depending on the conditions. The determination may also be made from the Boltzmann distribution at a specified temperature. For example, all structures that are present in 10% or more of the Boltzmann distribution may be determined to be stable structures.

In one embodiment, the search range may be defined by multiplying the bond length of the stable structure by a predetermined multiplier. For example, the bond length may be, without limitation, 1.2 times or less, and the search range may be defined by multiplying the bond length of the stable structure determined in step S102 by a multiplier of 1.2. This allows, for example, the number of stable structures obtained to be about 3 to 5, making it possible to search for a group of stable structures more efficiently than in a full-range search of GRRM.

In step S104, the structure prediction device 100 determines the most stable structure based on the potential energy of each molecular coordinate of the group of stable structures. Specifically, the structure prediction device 100 calculates the potential energy of each stable structure determined in step S103 from the molecular coordinates of the stable structure by DFT, and determines the stable structure having the smallest potential energy as the most stable structure.

The most stable structure determined in this way may be output as a predicted structure, returning to FIG. 1, or the determined stable structure may be used to predict and construct an aggregate structure such as a ring structure (e.g., a supramolecular complex) by symmetry operations (e.g., rotation operations, spotting operations, etc.). When predicting and constructing a supramolecular complex as an aggregate structure, a symmetry operation with the addition of a metal coordinate bond is performed.

In this way, model molecular coordinates to be used in a new conformation search may be extracted from the obtained ensemble structure, and the above-mentioned procedure may be repeated (see FIG. 3). When creating an ensemble structure, step S103 may not be performed, and instead, an ensemble structure may be created by symmetry operations as described above, and the procedure may be repeated. In this case, the obtained stable structures may be arbitrarily selected.

[Experimental result]
Next, experimental results of conformational search using GRRM will be described with reference to Fig. 8. Fig. 8 is a diagram showing experimental results of conformational search according to an embodiment of the present invention.

As shown in Figure 8, stable structure EQ0 is obtained from the input molecular coordinates by structural optimization calculation. Then, conformational search calculation is performed within the search range based on stable structure EQ0, and stable structures EQ1, EQ2, EQ3, and EQ4 can be obtained via saddle points TS0/1, TS1/2, TS1/3, and TS3/4. Since EQ2 has the smallest potential energy among the stable structure group EQ1 to EQ4, the energy of EQ2 is the most stable at the PM6 level, but EQ1 is determined to be the most stable structure based on a comparison of energies using the highly reliable DFT method.

From the most stable structure obtained in this way, it is possible to predict and construct a molecular structure with a flexible skeleton (e.g., a molecular metal complex with a flexible space). In addition, by searching for an assembly structure using the stable structure obtained, it is possible to predict and construct an assembly structure of a large size (e.g., a giant cyclic metal complex). Therefore, according to this embodiment, it is possible to predict and obtain a flexible and large assembly structure (e.g., a giant cyclic metal complex with a flexible space).

Although the above describes in detail examples of the present invention, the present disclosure is not limited to the specific embodiments described above, and various modifications and variations are possible within the scope of the gist of the present disclosure as set forth in the claims.

This application claims priority to Japanese Patent Application No. 2019-225462, filed on December 13, 2019, the entire contents of which are incorporated herein by reference.

100 Structure prediction device 101 Drive device 102 Auxiliary storage device 103 Memory device 104 CPU
105 Interface device 106 Communication device 107 Recording medium

Claims (8)

A processor obtaining molecular coordinates of a molecule;
the processor detects a first minimum on the potential energy surface of the molecule based on a gradient from the molecular coordinates, and determines a stable structure corresponding to the first minimum;
The processor detects one or more second minimum points within a search range starting from the stable structure, and determines a group of stable structures corresponding to the one or more second minimum points;
determining a most stable structure based on potential energy of each molecular coordinate of the group of stable structures by the processor;
A structure prediction method having the following structure: The structure prediction method of claim 1, wherein the molecular coordinates represent the positions of each atom of a molecular structure constructed by coordinating a metal ion to a coordination site of a given molecular structure. The structure prediction method of claim 1, wherein the molecular coordinates represent the position of each atom in a molecular structure constructed based on distance information between atoms. The structure prediction method according to any one of claims 1 to 3, wherein the coordination sites and coordination priorities of the metal ions coordinated to the molecule are determined by the selected metal ions. A structure prediction method according to any one of claims 1 to 4, wherein the step of determining the stable structure determines the stable structure by a structural optimization calculation. A structure prediction method according to any one of claims 1 to 5, wherein the step of determining the group of stable structures determines the group of stable structures by a conformational search calculation. A structure prediction method as claimed in any one of claims 1 to 6, wherein the search range is defined by multiplying the bond length of the stable structure by a predetermined multiplier. Memory,
a processor coupled to the memory;
having
The processor,
Obtain the molecular coordinates of the molecule,
Detecting a first minimum point on a potential energy surface of the molecule based on a gradient from the molecular coordinates, and determining a stable structure corresponding to the first minimum point;
Detecting one or more second minimum points within a search range starting from the stable structure, and determining a group of stable structures corresponding to the one or more second minimum points;
A structure prediction device configured to determine a most stable structure based on the potential energy of each molecular coordinate of the group of stable structures.

Images