JP2001319179A - Device for generating molecular orbital initial value - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a molecular orbital initial value generating system where an atom characteristic is reflected on a large molecule and additional trouble is removed. SOLUTION: A standard molecule data base 7 stores the corresponding information of an atom and a standard molecule and coordinate position information of the constituting atom of the standard molecule. An input means 101 inputs the constituting the constituting atom type information 20 of a target molecule and atomic orbital base information 30 which is to be used. A standard molecule selection means 102 selects the standard molecule from the standard molecule data base in accordance with constituting atom type information. An incorporated molecular orbital calculation means 103 calculates charge distribution or a density matrix by using atomic orbital base information on all selected standard molecules. A density matrix synthesis means 104 extracts the one relating to the atom constituting the target molecule from the calculated results and embedding it as the partial element of the density matrix of the target molecule. Thus, an approximate density matrix is generated. A conversion means 105 generates a molecular orbital initial value form the approximate density matrix.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to the generation of initial values of molecular orbitals, and more particularly to the generation of initial values suitable for large molecules in an ab initio molecular orbital calculation method using a general atomic orbital basis.

[0002]

2. Description of the Related Art All theoretical methods in molecular chemistry are based on quantum mechanical equations (Schrodinger's wave equations).
Is closely related to solving Quantum mechanical equations are many-body problems with few exceptions, so there are no analytical solutions, and their solutions are based on numerical calculations.

[0003] In the field of theoretical chemistry, a method called molecular orbital calculation method (also called molecular orbital method) has been established and used as a numerical solution of quantum mechanical equations (Seijiro Yonezawa et al., "Introduction to Quantum Chemistry", Chem. Doujin, 1983). According to the molecular orbital calculation method, solving the quantum mechanical equation
It turns into a problem of determining a set of functions of one-electron coordinates called molecular orbitals.

That is, in the molecular orbital calculation, the set of molecular orbitals {φi} is determined by solving a system of nonlinear equations. Specifically, the molecular orbital φi is determined by the determined atomic orbital base ｛χ
Approximate by a linear combination of p｝, and let φi = Σ (p) Cpiχp. Here, p is any number from 1 to an integer n, and n is determined by a computer. The linear coupling coefficient in this equation, that is, the molecular orbital coefficient {Cpi}, is determined by repeatedly solving the nonlinear simultaneous matrix equations, C′FC = e and C′SC = E. Here, C is a molecular orbital coefficient matrix having the molecular orbital coefficient Cpi as a matrix element, C ′ is its complex transpose, F is a Fock matrix, e is an eigenvalue matrix of the Fock matrix F, and S is the atomic orbital basis. The inner product matrix, E, represents an identity matrix. The Fock matrix F is nonlinear because it depends on the molecular orbital coefficient matrix C. This iterative solution is a procedure of estimating the initial value C0 of the molecular orbital coefficient matrix C by some method and repeating the improvement to converge on the solution. Convergence means that when the number of repetitions is N, the difference between C (N) pi and C (N + 1) pi falls within a predetermined value for all p and i. The method of estimating the initial value C0 where N = 0 is the molecular orbital initial value calculation method.

[0005] Conventional methods for calculating molecular orbital initial values can be classified into the following three types. (1) This is a semi-empirical method in which the calculation elements for obtaining the molecular orbital initial value are set by empirical parameters, and the molecular orbital initial value is obtained by diagonalization of the element matrix, etc., and many proposals have been made so far. (Patent No. 25000009, etc.). (2) The atoms constituting the target molecule are subjected to the atomic orbital calculation of the atoms alone in accordance with the given atomic orbital basis information, and the density matrix of the molecule is synthesized by using the atom density matrix or the charge distribution of the calculation result. How to get the orbit initial value. (3) The target molecule is divided into parts, molecular orbital calculations are performed in accordance with the atomic orbital basis information given for each part, and the density matrix or charge distribution of the calculation result is used to synthesize a molecular density matrix, and the molecular orbital initial is calculated. How to get the value.

[0006] In recent years, the need for analysis prediction for larger molecules has been increasing. For example, in order to develop a drug with higher efficacy, molecular orbital calculations for macromolecules such as proteins that make enzymes that cause chemical reactions in the human body are required. The same applies to the case where the electronic behavior of a highly integrated semiconductor is analyzed.

[0007]

The above-mentioned prior art is insufficient for such needs. First, there are a number of methods in the prior art belonging to the first category, but the common problem is that they are effective for small molecules, but become less effective for larger molecules. . The reason is that the calculation elements for obtaining the molecular orbital initial values are fixedly determined by empirical values or measured values for small molecules.

Another problem of the prior art belonging to the second category is that it is not possible to obtain an initial value reflecting the electric charge attractive force of atoms in a molecule. The reason is that the exchange of charges between atoms is not considered at all because the calculation result of atoms alone is used.

Further, a problem of the prior art belonging to the third category is that the target molecule (hereinafter, referred to as “target molecule”) as a target of molecular orbital initial value calculation is additionally provided with a designation for dividing into appropriate portions. Is extra.

A first object of the present invention is to provide an apparatus and a method for generating an initial value of a molecular orbital effective for a large molecule.

A second object of the present invention is to provide an apparatus and a method for generating a molecular orbital initial value reflecting the charge characteristics of atoms contained in a molecule.

A third object of the present invention is to provide an apparatus and a method for generating an initial value of a molecular orbital which eliminates an additional time for dividing a target molecule into a standard molecule.

[0013]

Means for Solving the Problems To solve the above problems, the molecular orbital initial value generation system of the present invention converts the state of an atom in a standard molecule contained in a target molecule into a given atomic orbital basis. Using information to calculate without estimation,
Due to the diversion, the estimation of the density matrix of the target molecule is automatically performed without any additional specification.

According to a first aspect of the present invention, there is provided a molecular orbital initial value calculating device for selecting a standard molecule according to the type of an atom contained in a target molecule, and a molecular orbital for the standard molecule using given atomic orbital basis information. Means for executing calculation, means for diverting a partial density matrix or a charge distribution index, which is a result of the molecular orbital calculation, to form an approximate density matrix for the target molecule according to the type of the atom, Means for converting the initial value into the molecular orbital value of the target molecule.

According to a second aspect of the present invention, there is provided a molecular orbital initial value calculating apparatus in an ab initio molecular orbital calculation method using atomic orbital basis information, wherein a standard molecule and a standard molecule which are included for each atom are included. A standard molecule database storing coordinate position information of atoms in the standard molecule, input means for inputting constituent atom type information and atomic orbital basis information of the target molecule and storing them in a memory, and the target molecule according to the molecular constituent atom type information A standard molecule selecting means for selecting a set of standard molecules from the standard molecule database so that all kinds of atoms constituting the standard molecules are included in any of the standard molecules, and the standard molecule selected using the atomic orbital basis information. A built-in molecular orbital calculation means for performing a molecular orbital calculation for all of them, and an element for forming the target molecule from the molecular orbital calculation result Density matrix synthesis means for extracting an approximate density matrix of the target molecule by extracting a partial density matrix or a charge distribution with respect to the target molecule and embedding the partial density matrix as a partial element of the density matrix of the target molecule; Conversion means for converting into molecular orbitals.

Specifically, in FIG. 1, the molecular orbital initial value generating apparatus 10 of the present invention comprises constituent atom type information 20 and atomic orbital base information 3 to be used, which constitute a target molecule.
0 is input by the input means, the standard molecule is selected from the standard molecule database 7 by the standard molecule selection means 2 according to the constituent atom type information 20, and the built-in molecular orbital calculation means 3 calculates the atomic orbital basis for all of the selected standard molecules. The charge distribution or density matrix is calculated using the information 30. Next, an approximate density matrix is generated by the density matrix synthesizing means 4, an initial molecular orbital value is generated from the approximate density matrix by the conversion means 5, and an initial molecular orbital value for the target molecule is output by the output means 6. .

As described above, according to the present invention, since the charge distribution or the density matrix is calculated in small standard molecular units, it can be applied to a large target molecule. Further, since the calculation unit is a standard molecule, the standard charge characteristics of atoms in the molecule can be reflected. Further, since the target molecule is automatically divided into standard molecules, no additional labor is required.

[0018]

Next, embodiments of the present invention will be described with reference to the drawings.

Referring to FIG. 1, a molecular orbital initial value generating apparatus 10 according to one embodiment of the present invention includes information on the types of atoms (constituting atomic type information) 20 constituting the target molecule and basic information on atomic orbitals 30 to be used. Input to the memory (not shown), a standard molecule selecting unit 2 for selecting a standard molecule from the standard molecule database 7 according to the constituent atom type information 20, and a molecule using the atomic orbital bottom information 30 for the standard molecule. Built-in molecular orbital calculation means 3 for orbital calculation
A density matrix synthesizing means 4 for synthesizing an approximate density matrix for the target molecule by using the calculation result of the standard molecule, a converting means 5 for converting the approximate density matrix to a molecular orbital, a molecular orbital for the target molecule The output means 6 outputs an initial value 40.

Each of these means operates as follows. The input means 1 inputs the molecular constituent atom type information 20 and the atomic orbital basis information 30 in a predetermined data format and stores them in a memory (not shown) in the molecular orbital calculation initial value generating device 10.

The standard molecule selecting means 2 reads the standard molecule database 7 from the standard molecule database 7 in accordance with the molecular constituent atom type information stored by the input means 1 so that all kinds of atoms constituting the target molecule are included in any of the standard molecules. Choose a set of standard molecules.

The built-in molecular orbital calculation means 3 performs the molecular orbital calculation for all the selected standard molecules using the input atomic orbital basis information 30. At this time, the built-in molecular orbital calculation means 3 also needs the molecular orbital initial value for the standard molecule. However, since the standard molecule is a small molecule, a conventionally known molecular orbital initial value generating method is used.

The density matrix synthesizing means 4 extracts a density matrix or a charge distribution for the atoms constituting the target molecule from the molecular orbital calculation result for the set of standard molecules, and embeds the density matrix as a partial element of the density matrix of the target molecule. Synthesize an approximate density matrix of the target molecule. The conversion means 5 converts the density matrix into molecular orbitals according to a normal method, and the output means 6 outputs the molecular orbital initial value 40 in a predetermined data format.

Next, the operation of this embodiment will be described in detail with reference to the block diagram of FIG. 1, the flowchart of FIG. 2, and FIGS.

First, the molecular type information 20 and the atomic orbital basis information 30 are inputted by the input means 1 (procedure A1 in FIG. 2). The molecular constituent atom type information 20 has a format listing element symbols or atomic numbers of atoms constituting the target molecule. For example, if the target molecule is CaH200C100N5
When it is described as 0O50 and described by element symbols, a total of 401 element symbols are listed as shown in FIG.
In FIG. 3, the same atomic species are continuously arranged, but the same atomic species may appear in a random manner. The atomic orbital basis information 30 is a normal configuration including the shell type of the orbital, the number of primitive orbitals constituting the atomic orbital, and the definition of each primitive orbital for each type of atom contained in the target molecule. The details are omitted here.

Next, the types of atoms constituting the molecule are identified by the standard molecule selecting means 2 (procedure A2 in FIG. 2). The atoms to be washed out are five atomic species of Ca, H, C, N and O in the example of CaH200C100N50O50 molecule.
Then, a set of standard molecules is selected from the standard molecule database 7 so as to include all these atomic species (procedure A3 in FIG. 2).

The information of the standard molecule database 7 includes type information and coordinate information of atoms constituting the standard molecule. For example, as shown in an example shown in FIG. Has definition information. When the target molecule is CaH200C100N50O50, FIG.
In the example of, from the index reference, Ca → standard molecule 1, H
→ Standard molecule 1, C → Standard molecule 1, N → Standard molecule 3, O →
The standard molecule 3 is selected, and finally the set of standard molecules is {Ca}
A set of two molecules of H6C2O and H5CNO is selected. FIG.
Shows the procedure for selecting standard molecules CaH6C2O and H5CNO from the target molecules CaH200C100N50O50.

Next, the molecular orbital calculation is performed for all the sets of standard molecules by the built-in molecular orbital calculation means 3 (FIG. 2).
Procedures A4 and A5). For example, in the above example, the molecular orbital calculation is performed for the standard molecule CaH6C2O and the standard molecule H5CNO. Although the detailed configuration of the built-in molecular orbital calculation means 3 is not included in the scope of the present invention, for example, CaH200C100N5
Even if it is impossible to calculate a large molecule such as 0O50, it is possible to calculate a small standard molecule such as CaH6C2O and H5CNO. Such an arrangement is known.

The calculation result for the standard molecule is stored as a density matrix or Mulliken charge. As shown in the configuration example of FIG. 6, the density matrix information is represented by a two-dimensional matrix subscripted by an atomic orbital p that describes a molecular orbital. The density matrix P is defined as P = COC 'where C is a molecular orbital coefficient matrix and O is an occupancy matrix. The Mull of the atomic orbital p
The iken charge λp is defined as λp = Σ (q) PpqSpq.

Next, the density matrix synthesizing means 4 extracts partial density matrix (atomic density matrix) information or Mulliken charge information necessary for the target molecule from the calculation results of the standard molecules (procedure A6 in FIG. 2, FIG. 6). For example, in the above example, as shown in FIG.
And the information of H and C are extracted, and N and O are extracted from the H5 CNO molecule.
Extract the information of When extracting Ca information from a CaH6C2O molecule, extract a submatrix in which both the atomic orbitals p and q belong to the same atom in the element Ppq of the density matrix, or make the atomic orbital p belong to the same atom in Ppp or λp. Extract diagonal elements (see FIG. 6).

For a standard molecule containing a plurality of the same atomic species, an average value or a representative value is taken, and the element values relating to the sub-orbitals belonging to the same shell are replaced with the average values in the sub-orbitals (A7 in FIG. 2). .

Then, according to the constituent atom type information of the target molecule, as shown in FIG. 7, a partial density matrix or a Mulliken charge extracted from the standard molecule for each atom is arranged in a diagonal term as a partial matrix, thereby obtaining the target molecule. Is synthesized (A8 in FIG. 2).

Further, in the conversion means 5, F0 = SP0S
And solving the simultaneous matrix equations, C0'F0C0 = e and C0'SC0 = E, to obtain the initial molecular orbital value C0 (procedure A in FIG. 2).
9).

Finally, the output means 6 outputs an initial value C0 of the molecular orbital coefficient matrix C in a predetermined data format (procedure A10 in FIG. 2). The molecular orbital initial value information is represented by a two-dimensional matrix subscripted by the molecular orbital s and the atomic orbital t as in the configuration example shown in FIG.

Thus, according to the present invention, a good molecular orbital initial value can be generated even for a large molecule for which a good initial value cannot be generated by the existing molecular orbital initial value generation method. Become like

A program for causing a computer to execute the method for generating the molecular orbital initial value described above is stored in a semiconductor memory, a floppy (registered trademark) disk, or a CD-ROM.
Alternatively, the program may be recorded on a recording medium such as the above, and read and executed by a computer. The program controls the computer to function as a molecular orbital initial value generation device.

[0037]

A first effect of the present invention is that an initial molecular orbital value reflecting the charged attraction of atoms can be formed.
The reason for this is that not only the atoms alone but also the atoms in the standard molecule containing the atoms are calculated using the given atomic orbital basis, which reflects the standard charge characteristics of the atoms in the molecule.

The second effect of the present invention is that a good molecular orbital initial value can be provided even for a large molecule which has been difficult to obtain conventionally, so that the applicable range of the molecular orbital calculation is expanded. That is. In particular, the ability to apply the molecular orbital method to a macromolecule such as a protein, which has been difficult to converge with the conventional molecular orbital initial value generating method, is improved.

Further, a third effect of the present invention is that, since the division of the target molecule into standard molecules is automated, no additional labor is required.

[Brief description of the drawings]

FIG. 1 is a configuration diagram showing one embodiment of a molecular orbital initial value generation device of the present invention.

FIG. 2 is a flowchart illustrating an example of a method for generating a molecular orbital initial value according to the present invention;

FIG. 3 is a diagram showing a configuration example of molecular constituent atom type information of a target molecule for explaining the present invention.

FIG. 4 is a diagram showing a configuration example of a standard molecule database used in the present invention.

FIG. 5 is a diagram showing a standard molecule selection procedure in the present invention.

FIG. 6 is a diagram exemplifying a correspondence relationship between a configuration example of atomic density matrix information and extraction of a partial matrix according to the present invention using a standard molecule example;

FIG. 7 is a diagram illustrating a procedure for synthesizing an approximate density matrix of a target molecule in the present invention.

FIG. 8 is a diagram showing a configuration example of molecular orbital initial value information for explaining the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Input means 2 Standard molecule selection means 3 Built-in molecular orbital calculation means 4 Density matrix synthesizing means 5 Conversion means 6 Output means 7 Standard molecular database 10 Molecular orbital initial value generation device 20 Composition atom type information 30 Atomic orbital base information 40 Molecular orbital initial value

Claims (7)

[Claims] 1. A means for selecting a standard molecule according to the type of atoms contained in a target molecule; a means for performing a molecular orbital calculation on the standard molecule using given atomic orbital basis information; Means for diverting a partial density matrix or a charge distribution index that is an average density matrix for the target molecule in accordance with the type of the atom, and means for converting the approximate density matrix into an initial molecular orbital value of the target molecule. A molecular orbital initial value calculation device used for ab initio molecular orbital calculation method characterized by comprising: 2. A molecular orbital initial value calculation apparatus in an ab initio molecular orbital calculation method using atomic orbital basis information, wherein a standard molecule containing the atom and coordinate position information of the atom in the standard molecule are stored for each atom. A standard molecule database; input means for inputting constituent atom type information and atomic orbital basis information of the target molecule and storing the information in a memory; and all of the types of atoms constituting the target molecule according to the molecular constituent atom type information are standard molecules. A standard molecule selecting means for selecting a set of standard molecules from the standard molecular database to be included in the standard molecular database; and a built-in molecular orbital calculation for performing a molecular orbital calculation for all of the selected standard molecules using the atomic orbital basis information Means, a partial density matrix or a charge distribution for atoms constituting the target molecule from the molecular orbital calculation result. Density matrix synthesizing means for extracting and embedding the approximate density matrix of the target molecule by embedding as a partial element of the density matrix of the target molecule; and converting means for converting the approximate density matrix into molecular orbitals of the target molecule. A molecular orbital initial value calculation device characterized by comprising: 3. The standard molecule database includes a reference index indicating, for each atom, a standard molecule that includes the atom, and standard molecule definition information for displaying position coordinates of an atom included for each standard molecule. The molecular orbital initial value calculation device according to claim 2, characterized in that: 4. The density matrix synthesizing means, when the selected standard molecule contains a plurality of the same atoms, takes an average value or a representative value thereof. 3. The molecular orbital initial value calculation device according to 3. 5. The density matrix synthesizing means extracts a partial density matrix or a Mulliken charge for a corresponding atomic species from a molecular orbital calculation result of each standard molecule, and extracts a partial density matrix or a Mulliken charge from the partial density matrix or the Mulliken charge in a shell of an atomic orbital. Of the sub-orbital in the shell is replaced by the average value, and the partial density matrix or Mulliken is used for each primitive of the target molecule according to the molecular constituent atom type information.
5. The molecular orbital initial value calculation apparatus according to claim 2, wherein an approximate density matrix for the target molecule is synthesized by embedding charges as a partial matrix. 6. A method for calculating an initial value of a molecular orbit in an ab initio molecular orbital calculation method using atomic orbital basis information, comprising: a standard molecule containing each atom and coordinate position information of an atom in the standard molecule; A procedure for storing in a database, a procedure for inputting constituent atom type information and atomic orbital basis information of a target molecule and storing them in a memory, and a procedure for storing all types of atoms constituting the target molecule according to the molecular constituent atom type information as standard Selecting a set of standard molecules from the standard molecule database to be included in any of the molecules; performing molecular orbital calculations for all of the selected standard molecules using the atomic orbital basis information; A partial density matrix or a charge distribution for atoms constituting the target molecule is extracted from the molecular orbital calculation result, and the target is extracted. A molecule having a procedure of synthesizing an approximate density matrix of the target molecule by embedding as a partial element of the density matrix of the molecule, and a step of converting the approximate density matrix into a molecular orbital of the target molecule. Orbit initial value calculation method. 7. A computer equipped with a standard molecule database which stores therein standard molecules containing atoms for each atom and coordinate position information of atoms in the standard molecules. And a process of storing a set of standard molecules from the standard molecule database such that all types of atoms constituting the target molecule are included in any of the standard molecules according to the molecular constituent atom type information. Performing a molecular orbital calculation for all of the selected standard molecules using the atomic orbital basis information; extracting a partial density matrix or charge distribution for atoms constituting the target molecule from the molecular orbital calculation result; By embedding as a subelement of the density matrix of the target molecule, the approximate density of the And processing for combining the matrix, the approximation density recording a computer-readable recording medium storing a program for executing a process to convert the molecular orbital of the matrix the target molecule.