JPS63223917A - Processing method for chemical reaction information - Google Patents

Abstract

PURPOSE:To automatically derive a reaction type peculiar to a chemical reaction by executing an operation for extracting a specific effective part flag, to chemical reaction information which has been inputted as a virtual transition structure and/or a coupling table. CONSTITUTION:A processing method for this chemical reaction information is to process the information related to a chemical reaction generated in at least one starting substance. The information related to this chemical reaction contains a virtual transition structure which has discriminated and shown (1) a coupling between nodes existing in common the starting substance and a generated portion substance, (2) a coupling between nodes existing in only the starting substance, and (3) a coupling between nodes existing in only the generat ed portion substance, respectively between a structure of this starting substance and a structure of the generated portion substance which have been superposed topologically, and information related to the coupling between these nodes and the node. Also, (2) and (3), and a series of nodes coupled by its coupling are contained, and by extracting an effective part flag consisting of a carbon atom. and a non-carbon atom. from said structure and coupling table, chemical information is processed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of the Invention] The present invention relates to a method for processing information on chemical reactions, and more particularly to a method for processing information regarding structural changes of substances involved in chemical reactions.

[Technical Background of the Invention] In recent years, with the development of computers, various methods for recording structural information of chemical substances, particularly organic compounds, have been proposed and are being used□. The amount of organic compounds and organic chemical reactions that have been studied and elucidated to date is enormous, and this known information can be used effectively to search for known chemical substances or chemical reactions in a short amount of time. Furthermore, it is desired to find a method for synthesizing new substances having desired properties. To this end, instead of using conventional chemical structural formulas, which are easy for engineers to grasp their structural characteristics, as a form of representation of chemical substances and chemical reactions, it is necessary to use formulas that can be processed by computers (i.e., can be ) is required to develop and utilize forms of expression.

The method for recording chemical substances is W L N (Wisw
Typical methods include linear notations such as esser Linear Notation (Esser Linear Notation) and methods using join tables.
, R,l1eller, R,J,Feldsann,
E, )lyde (Eds): Computer Re
Presentation and Manipulation
ion of Chemical Information
on'' (John Wiley and 5ons,
New York, 1974) [Wipke, Heller, Feltman, Hyde (eds.): [Computer Representation and Handling of Chemical Information] (John Wiley & Sons)]. concatenation table
For example, a cation table is a table that lists the type of each atom in the structural formula of a chemical substance, the atom to which it is bonded, and the type of bond. It has the advantage of being able to be searched by unit.

In addition, methods for recording information regarding structural changes (chemical reactions) of chemical substances have also been proposed, but so far no satisfactory method of expression is known. for example,
One way to record information about chemical reactions is to use reaction codes, specifically as described by J. Valls, 0.
5cheiner: Chemical Informa
tion Systems'', E, Ash, E,
Hyde (Eds), (Ellis llorwoo
d Limlted, 1975) p, 241-25
8 [Method described in [Barr, Shina 12 "Chemical Information Systems", Asch, Hyde ed. (Ellis Horaud)], M. A., Lobeck, Angew, C.
Hew, Intern.

Ed, Engl., '4.578 (1970)
[Robetsk, Angewante Hiemi International Edition in English] and H, J, Ziegler, J, C
hem-1nf, Comput.

Sci., 19.141 (1979) [Ziegler, Journal of Chemical Information and Commercial Science]. This method has a drawback in that it cannot record new chemical reactions when they are discovered, since the viewpoint of chemical reaction expression is fixed. Furthermore, since the structural information of chemical substances and the information on changes thereof are recorded in separate formats, there is a drawback that effective information retrieval cannot be performed.

Separately, a recording method devised from the standpoint of designing synthetic routes for chemical substances is also known. For example, E, J, Cor
ey, R.D., Cramer, W.J., How.
e, J, As, Chem, Soc,,
94.440 (1972) [Kohli, Kramer, Houe, Journal of the American Chemical Society], LIJgi, J., Bauer, J.
8raudt.

J, Fr1edrich, J, Gasteiger,
L., Jochum, W., Chubert, A.
ngew, Chew, Intern, Ed
, Engl., 18. 111 (1979)
There is a method described in [Ugi, Haumer, Braut, Friedrich, Gasteiger, Jochum, Schubert, Angewante Hiemi International Edition in English]. However, this method is not suitable for recording individual chemical reactions.

The applicant has proposed a reaction line in which (2) bonds between nodes that exist only in the starting material and (3) bonds between nodes that exist only in the product material are alternately connected as a type of reaction type. We have already filed a patent application for a method for processing chemical reaction information, which consists of extracting an abstracted reaction graph from an imaginary transition structure and/or a bonding table, and a method for processing chemical reaction information, which consists of enumerating the reaction graphs. (Japanese Patent Application No. 61-20437) [Summary of the Invention] An object of the present invention is to provide a novel processing method for deriving a reaction type from chemical reaction information.

Another object of the present invention is to provide a processing method for recording, storing or displaying reaction types based on chemical reaction information in an expression form that can be processed by a computer.

A further object of the present invention is to provide a processing method for classifying and enumerating reaction types of chemical reactions.

That is, the present invention provides a method for processing information regarding a chemical reaction that produces at least one product from at least one starting material, wherein the information regarding the chemical reaction is obtained by processing information about the topologically superimposed structure of the starting material. and the structure of the product: (1) bonds between nodes that exist in common in the starting material and product, (2) bonds between nodes that exist only in the starting material, and (3) bonds that exist only in the product. and (2) nodes that exist only in the starting material. A bond between and/or (
3) Contains bonds between nodes that exist only in the product substance and a series of nodes connected by the bonds, and the series of nodes includes one or more adjacent carbon atoms and the carbon atoms adjacent to the series of nodes. A method for processing chemical reaction information comprising extracting an effective subgraph consisting of non-carbon atoms from the heavy transition structure and/or bonding table, and [2] Chemistry that produces at least one product from at least one starting material. A method for processing information about a reaction; comprising a bond between nodes that exists only in the starting material and/or a bond between nodes that exists only in the product material and a series of nodes connected by the bond, and Based on the combination of bonds, the number of carbon atoms, the number and type of non-carbon atoms for an effective subgraph consisting of one or more adjacent carbon atoms and non-carbon atoms adjacent to the carbon atoms in a series of nodes. The present invention provides a method for processing chemical reaction information, which consists of hierarchizing chemical reaction information.

According to the first method of the present invention, a reaction type specific to a chemical reaction can be automatically derived by performing suitable processing on chemical reaction information input as a heavy transition structure and/or a bond table. .

In the present invention, a heavy transition structure (i
Ansition 5structure (hereinafter abbreviated as ITS) is a structural change in a substance involved in a chemical reaction that includes (1) bonds that exist only in the starting material, (2)
It refers to a two-dimensional or three-dimensional structural diagram (figure) that is differentiated into three types, consisting of bonds that exist only in the product substance and (3) bonds that exist in common in both. This structural diagram is based on conventional structural formulas and three-dimensional structural diagrams of compounds, and can represent chemical reactions in a form that is visually familiar to engineers and easy to understand.

In addition, a connection table is a simple and clear list of combinations such as the types of nodes in a chemical reaction, the partner nodes that connect to the nodes, and the connections between these nodes that are differentiated into the three types listed above. This combination table allows chemical reaction information to be stored on a recording medium without requiring a large capacity.

In the heavy transition structure and bond table above, chemical reactions are basically nodes (nodes) consisting of atoms, atomic groups, etc.
It is expressed in a simple way about the bonds between and nodes, and the bonds between nodes in the reaction system are expressed by distinguishing them into the above three types. To this end, by performing simple graphical or arithmetic processing on the multiple transition structures or bond tables stored (registered) in a computer, focusing on the distinction between these bonds, we can obtain information on bond changes specific to reactions, i.e. “Effective subgraph J (effecti
ve subgraph) can be obtained.

In the present invention, an effective subgraph representing a reaction type is
A partial reaction structure (subgraph) containing one or more adjacent carbon atoms, a non-carbon atom bonded to it, and a bond between these atoms, in which the bond between the atoms is (2).
(3) A bond between nodes that exists only in the starting material and/or (3) a bond between nodes that exists only in the product material. That is, the effective subgraph is a partial reaction structure unique to a reaction in which a carbon atom is the center, and similar reactions have a common effective subgraph.

In the field of organic synthesis, the ability to classify and register individual reactions using this effective subgraph is of great value in searching for, studying, and applying individual reactions or a series of reactions.

By the method of the present invention, the reaction type becomes an effective subgraph,
Because it is obtained in the form of a two-dimensional figure following the heavy transition structure of a reaction, it is easy for engineers to visually understand, and since it follows the usual method of representing compounds, the information can be used directly. be able to. The effective subgraph can also be expressed in the form of a join table, or in the form of characters such as codes, symbols, or a combination of these. In this case, the information of the effective subgraph can be expressed very simply. I can do it. Therefore, it is easy to register the information in the computer, and the registration does not require a large amount of storage space.

Furthermore, according to the second method of the present invention, the effective subgraphs of each reaction can be automatically classified and systematized. The effective subgraph is classified according to the bonding state of nodes, the number of carbon atoms, the type of non-carbon atoms, etc., and this is referred to as rWi layering 1 in the present invention. Hierarchical effective subgraphs can be associated with conventional reaction names, and by coating them, chemical reactions can be easily searched.

Therefore, reactions involving carbon atoms (organic reactions) are
It can be classified and systematized based on its effective subgraph. Furthermore, it is possible to search for similar reactions and register individual chemical reactions according to this classification via the effective subgraph. A huge amount of chemical reactions can be classified and processed easily in a short time, and the utilization of reaction information registered in a computer can be increased.

The figures, connection tables, and/or codes obtained regarding the information regarding these effective subgraphs can be further stored in a computer, recorded on blank paper, or displayed on a screen such as a cathode ray tube.

Based on the registered effective subgraph information, it is possible to efficiently search for information on chemical substances, chemical reactions, etc. in a short time. It is possible to shorten the time and increase the density of information obtained, making it possible to proceed with research efficiently.

In addition to these advantages, by using the information on the effective subgraph obtained by the method of the present invention in combination with the chemical reaction information that has already been input, it is possible to further improve the chemical substance information obtained from the chemical reaction information. By using this together, structural analysis and molecular design (
molecular modeling), organic synthesis route design (heuristic analysis)
It is possible to carry out organic 5 synthesis). In addition, chemical substructure searches,
Structure-activity relationships, automatic structure determination of unknown compounds, and prediction of reaction mechanisms and reaction products when complex compounds are reacted under certain conditions.
aluation of organic react
ion) etc. can be carried out within a sufficiently practical range in a short period of time.

[Structure of the Invention] In the chemical reaction information processing method of the present invention, firstly,
An imaginary transition structure and/or an imaginary transition structure in which bonds between nodes are differentiated into three types: bonds that exist only in the starting material, bonds that exist only in the product material, and bonds that exist in common to both.
Alternatively, for chemical reaction information expressed as a bonding table, an operation is performed to extract an effective subgraph in which nodes involved in the reaction including carbon atoms are connected by specific bonds, and thereby Information regarding bond changes is obtained as shapes, bond tables and/or coats.

The method of the present invention will be specifically explained below, taking the bromination reaction of naphthalene as an example.

The bromination reaction of naphthalene is represented by r (reaction formula l). The imaginary transition structure (ITS) of this reaction is expressed, for example, as follows.

Below 7; Mortar, where: i) The symbol → represents a bond that exists in common in the starting material and the product material.

ii) The symbol represents a bond that is present only in the starting material, and 1ii) The symbol represents a bond that is present only in the product material.

In other words, an imaginary transition structure (ITS) is a two-dimensional or three-dimensional structure in which the structure of the starting material and the structure of the generated material are topologically superimposed, and the bonds between each node are differentiated into the three types i) to 1ii) above. Refers to the original structure. Note that "r topologically superposed" specifically refers to matching nodes appearing in the structure of the starting material and nodes appearing in the structure of the product material, and combining these structures into one.

In the imaginary transition structure (ITS) according to the present invention, the nodes of substances involved in chemical reactions may be expressed in units of atoms contained in the filament (starting material group) and the production system (product material group). Alternatively, a methine group [the above node (
3), (7), etc., which are simply brominated and the node representation is abbreviated as C] may be represented by an atomic group unit such as a functional group. Further, when expressing a chemical reaction, some nodes appearing in the yarn and the production system may be omitted.

Furthermore, the distinction between the three types of bonds is not limited to the display using symbols such as i) to 1ii) above. , red, green) can be determined by the user's five senses, such as display, and any means may be used as long as it can be processed by a computer.

Hereinafter, in the present invention: i) A bond (symbol -) common to the starting material and the product is referred to as a "colorless bond" or a constant bond (par-tiand).

ii) Bonds (symbols) that exist only in the starting material are called out-bonds, and 1ii) Bonds that exist only in the product materials (symbols) are called r-man bonds J (1n-
The outgoing bonds and incoming bonds are collectively called the r-colored bonds, and these three types of bonds that appear in the imaginary transition structure are collectively called the imaginary bonds 1.

Specifically, Table 1 summarizes the types of bonds that appear in the imaginary transition structure in the present invention. Note that in Table 1, the horizontal numbers mean indicators of the inflow and outflow of bonds.

In Table 1 below, for example, the bond represented by the symbol r/1 is a single 1n-bond, and can be represented by a pair of numbers (0+1). Here, 0 means that no bond exists in the original system before the reaction, and +1 means that a single bond is present in the product system after the reaction. Similarly, the bond represented by +1 is a single out-bond and is represented by (1-1), meaning that a single bond exists in the original system before the reaction, but in the product system after the reaction. This means that a single bond has disappeared.Also, the bond represented by (2-1) is a double bond si
ngly cleaved) and is expressed as r±1.

Thus - a type of bond, a pair of numbers: (a,
b) [where a is an integer representing the bond multiplicity in the starting material and b is an integer representing the change in bond multiplicity in the chemical reaction], which can be expressed as r complex bond number 1 bromination (complex bond num
ber) or r imaginary multiplicity J (imaginary m
ultiplicity). In addition,(
In the notations a and b), the comma (1) may be omitted.

According to this notation, even if the connection multiplicity is two or more, it can be expressed concisely. In addition, it is particularly preferable for storing records of chemical reactions because it does not require much storage capacity and can be directly processed by computer, and is a preferred expression method for creating the bond table described below.

Alternatively, the chemical reaction may be performed using a connection table containing information about each node, the binding partner node that binds to the node, and the bonds between the nodes.
It is expressed as

Table 2 shows the bonding table for the naphthalene bromination reaction.

As shown in Table 2, the bond table is a list of all nodes, all nodes bonded to each node, and the types of bonds between these nodes, listed in order of node number, for the fibers and production system involved in the reaction. It is a table.

Note that it is possible to create a bonding table based on the imaginary transition structure of a reaction, and conversely, if the bonding table includes position information of each node, it is possible to create an imaginary transition structure from the bonding table. Ru. In other words, it can be said that the imaginary transition structure and the bond table are two sides of the same coin as a form of registration and display of chemical reaction information.

The connection table may also include input information regarding the position coordinates of each node. In addition, the imaginary transition structure and/or bond table may contain additional information, if desired, including information on the charge and stereochemistry of each node; various physical property values and spectral information of chemical substances involved in the reaction; and enthalpy and temperature of the reaction. 1 hour, information regarding the catalyst used, atmosphere, reaction phase, reaction yield, presence or absence of by-products, etc. may be described. Furthermore, reaction names, reaction numbers, etc. may be attached to each imaginary transition structure and/or bond table in order to facilitate the accumulation, management, and search of chemical reaction information.

For details of the above imaginary transition structure and bond table, see
Japanese Patent Application No. 177345/1986 filed by the applicant
No. 180875. For details of the imaginary transition structure (DITS) containing charges and the bond table, and the three-dimensional imaginary transition structure (ITSS) and the bond table, please refer to Japanese Patent Application No. 60-298603 and No. 60 filed by the same applicant.
-298604.

Furthermore, in the above ITS and connection table, the operation of extracting the yarn and production system [respectively, projection J (pr
injection to the starting
abbreviated as stage%P S), “Projection J
(projection t.

the product stage, abbreviated as PP)], a starting material and a product are obtained. Here, the PS operation for ITS is an operation that deletes the incoming bond (+) from ITSI and considers the outgoing bond (+) to be a constant bond (-), as shown below, and the PP operation is performed from ITSI. This is an operation that deletes outgoing bonds and considers incoming bonds to be permanent bonds.

In addition, by extracting colored bonds, that is, parts where outgoing bonds (+) and incoming bonds (now) are alternately connected, from the imaginary transition structure (ITSI) of the above reaction, a reaction equation representing a partial reaction structure can be obtained. [reaction string, graph of reaction
(also called inters)] is obtained.

In other words, it is obtained by removing the colorless bond (-) from ITSI. However, if a colorless bond and a colored bond are combined (for example, earth, shibe)
Depending on the purpose, colorless bonds may be left without being deleted.

For details on ps, PP operations and reaction procedures, see Japanese Patent Application Nos. 60-185386 and 60-19 filed by the present applicant.
No. 7463.

The effective subgraph of the above reaction is determined from ITSI by nodes consisting of one or more adjacent carbon atoms connected by colored bonds [outgoing bonds (+) and incoming bonds (÷)];
It is also obtained by extracting nodes consisting of non-carbon atoms bonded to the carbon atom with colored bonds, and these colored bonds.

In the effective subgraph, a subgraph consisting of adjacent carbon atoms and the bonds between them is defined as a reaction nucleus (rea(ji□n
The part consisting of a non-carbon atom (this is called a terminal atom) and the bond between the non-carbon atom and the reaction nucleus is called a terminal descriptor (terminal descriptor).
6 Also, among the effective subgraphs,
As mentioned above, a graph that contains only a particularly reactive tree and the sum of the number of carbon atoms and non-carbon atoms is n is called an n-nodal subgraph (where n is is an integer greater than or equal to 3). Generally, a chain n-node subgraph contains n-2 carbon atoms and two non-carbon atoms. The above 1sg is a five-node subgraph (abbreviated as 3NSG).

This five-node subgraph is also a subgraph of the reaction graph (1rc).

In addition, for the five-node subgraph, after detecting colored bonds (bonds that are b-#O) from the bond table shown in Table 2, atom by atom is created in the order of the nodes for the detected bonds.
A partial bond table is obtained by alternately tracing out and in bonds using methods such as , and leaving only adjacent carbon atoms and adjacent non-carbon atoms among the nodes.

Another example of the five-node subgraph of the bromination reaction is shown in ITS2 below.
~15. Further, examples of five-clause subgraphs for chlorination reactions and iodination reactions are shown in ITS16 to 18 and ITS19 and 20, respectively. The five-node subgraph of each reaction is the area surrounded by a broken line.

An example of a five-clause subgraph for a nitration reaction is shown in ITS21.22.
Furthermore, an example of a five-node subgraph of the nitrosation reaction is shown in ITS23. In both cases, the same H10+N is obtained as a five-clause subgraph.

Nitration and nitrosation can be distinguished by the number of oxygen atoms substituted in the pentabar subgraph.

Note that atoms that are included in the reaction chain, such as hydrogen atoms and nitrogen atoms, are called intrastring atoms, and atoms that are not included in the reaction chain, such as oxygen atoms, are called extrastring atoms. .

Examples of five-node subgraphs for oxidation reactions are shown in ITS24, and examples of five-node subgraphs (H+ (: +S)) for reactions involving sulfur are shown in ITS25-27. Sulfonylation,
The distinction between sulfonation and chlorosulfonation is made by extra atoms.

Second blank space below: In the chemical reaction information processing method of the present invention, the effective subgraph is processed based on the number of reaction lines included in the effective subgraph, the combination of bonds, the number of carbon atoms, the type of non-carbon atoms, etc. Operations to classify and hierarchize are performed.

Prior to hierarchizing the five-node subgraph described above, first, the hierarchization of non-carbon atoms is determined.

Table 3 shows the hierarchy of non-carbon atoms according to the present invention.

Table 3 As shown in Table 3, non-carbon atoms bonded to carbon atoms are collectively expressed as hyperato■,
Among them, electron-donating atoms such as H, Na, and Mg are collectively referred to as superhydrogen (HH), and the remaining atoms are referred to as Z (Z atom). Among the Z atoms, a distinction is made between halogens (abbreviation: Hal) such as F, CfL, Br, and ■ and other heteroatoms (abbreviation: Het) such as N, O, S, and P.

Based on this hierarchy of non-carbon atoms, the above ITS1-2
The five-node subgraph shown in 7 is classified and stratified as shown in Table 4. H+Ce-F, H+ (: + (f
l, l (+(:+9r and)l + C+ generate ll+ C-e Hal as upper descriptor of l,)I
+C+N, H+C÷0 and H+C now generates Hl(+C+ Het) as a higher descriptor of S, etc.
As the upper descriptor of HH+C Hal and ll+C Het) lH+c Z is generated.

Margin below Ro Ro Tatami Tatami opening K1
″ At each level of the hierarchy, phylum, class (
class), group (division), section (sec)
tion), order, family
), genus, and species (spec 1es). However, the level distinction is not limited to these predicates, and arbitrary characters and codes such as combinations of alphanumeric characters and other codes can be used as long as they can be distinguished.

As shown in Table 4, species is the lowest descriptor level and corresponds to the ITS of each reaction. Again, subgenus (
subgenus) corresponds to a subgraph in which an extra atom is included in a five-section subgraph, and can be associated with a reaction name (predicate in natural language) that has conventionally been generally called. A genus corresponds to a five-section subgraph. Subfamily (subf)
amily) is a five-node subgraph that integrates one of the terminal atoms, such as a hydrogen atom, with a higher-order descriptor (hyperhydrogen),
A family is a five-section subgraph in which the other end atom is also integrated by a higher-order descriptor (Hat and Bet). HH+C
+Hal family corresponds to halogen family reaction, HH+C+He
The t family corresponds to heteroatom substitutions. The eyes are further (1a
! It is a five-node subgraph that integrates and Het by its higher-level descriptor (Z), and this
tive 5ubstition).

Tables 5 and 6 show Z-a-C joint HH and 2+
The hierarchization of the five-node subgraph is shown for each of the C and Z positions. Lower levels below genus are omitted.

Z+C+HH is a reductive substitution reaction (reductive
Examples are shown in ITS28-30. In addition, Z + C + Z is a substitution that does not change the oxidation state (isohypsic 5ubsti
tition) reaction, an example of which is shown in ITS3
1 to 35.

Below is the margin from table 5 C rooster EE) (3J3yo EO (i positive ball D Q from table 6 C3D ■yo IDC four rooster ko E2) G below ITS36
38 shows an example of a five-clause subgraph containing an identity connection.

When the five-clause subgraph is hierarchized with respect to terminal descriptors, a list as shown in Table 7 is obtained. Here, Ko means a reaction nucleus containing an odd number of carbon atoms. In Table 7, for example, ARK, synthesis A (however, Ko is not included in the terminal descriptor but is listed for convenience) is a super substitution reaction (super substitution reaction).
rsubstition) and is a higher-level descriptor at the clause level. The eye level terminal descriptors shown in Tables 4 to 6 belong to this lower level.

An example of a five-section subgraph in Table 7 is shown below in ITS40 (the area surrounded by a broken line). The reaction graph 40rc also includes a five-node subgraph. The five-section subgraph 405g can be described separately into a reaction nucleus and a terminal descriptor.

40, c5NSG 4059 The following margin five-node subgraph (40sg) is converted into the five-node subgraph (1
sg), it is clear that they have similar terminal descriptors. This is a natural consequence of the alternating nature of the reaction line (outgoing bonds and incoming bonds appear alternately). For example, at 405g the terminal descriptor is Na KO-e-
H, but this is llH+KO÷ in the list in Table 7.
) is the lower descriptor of IH.

That is, an odd-numbered n-node subgraph can be divided into a reaction nucleus containing n-2 carbon atoms and a terminal descriptor containing two terminal non-carbon atoms.

Further, as shown in Table 7, the terminal descriptor is common when n is an odd number.

Other examples of five-clause subgraphs are shown in ITS41-50 (portions surrounded by broken lines).

An example of the upper section subgraph is shown in ITS51 (the part surrounded by a broken line).

Below, the reaction nucleus containing m total white carbon atoms is defined as C1 (reaction) nucleus (C, r
action kernel or C,kernel
) (where m is an integer of 1 or more). C3
Table 8 shows a list of nuclei. Further, a part of the list regarding C5 is shown in Table 9. These are basic C3
Nucleus C+C Minute C1 Basic CS Nucleus C+C÷C+C+C It can be regarded as an isomer in which a small bond (-) and/or a double constant bond (=) is replaced with each side (bond).

Therefore, the reaction nucleus containing an odd number of m carbon atoms is the basic C,
Isomerism in which the nucleus C10 [+C10], now C (where p is an integer satisfying p≧O) is replaced with a small bond (-) and/or a double constant bond (=) It is the body. As a result, an odd-numbered n-node subgraph has a Cn-2 kernel and a terminal descriptor (
(see Table 7).

List 8 of all songs below. +C+C substitution C±C±Cπ-rearrangement Table 9 (-e (-++-(-6-C-1)-(C standing C north C standing C amount C This will be explained by taking as an example a reaction that produces a double bond by oxidation.

This reaction can be expressed as ITS52.

A reaction graph 52rc is obtained from the ITS52. ITS
52 and the reaction graph 52rc both include a four-node subgraph 525g (the portion surrounded by a broken line). A four-node subgraph can be divided into a reaction nucleus and a terminal descriptor in the same way as above. The terminal descriptor is Cl+, +C
It is represented by 1, corresponds to the dechlorination reaction, and H31
+l (lower descriptor of e+la1 (see Table 11)
. Moreover, the reaction nucleus is C±C.

A list of the C2 nucleus and C4 nucleus is shown below in Table 1O in the margin below.

Also, another example of a four-node subgraph is shown in ITS53-56,
Examples of hole subgraphs are shown in ITS58 to ITS61, and an example of a node subgraph is shown in ITS62. The part surrounded by the broken line is the nm subgraph.

As is clear from these examples, the terminal descriptors in even-numbered nm subgraphs are common and are selected from the lists shown in Tables 11 and 12. In addition, No. 11.
In Table 12, lower levels below genus are omitted.

In addition, in general, the reaction nucleus of an even-numbered nm subgraph contains n-2 carbons, and C4) [C-@-(], ÷c or C
+[C min C1p+C (where p is an integer satisfying p≧0), a small value combination (-
) and/or an isomer in which the double permanent bond (=) is substituted.

Below is the margin Table 10 Detachment type (Touge Table 11) (-e-(: C”#CC all C C÷C Kano CC town de Kano CC → C2 c-ecC sum+:C-e
CC±C-+C Mi C standing C two C Mi C meeting C+CeCC-
eCIC-eC Giant C:! :C ÷ cc To Kano C meeting cc Standing C Main C → CC number tC
±CC world C2C±C HiC→C spiritual cc association CiC both C addition type (→Table 12) c-+c c+-c c±CC+C-e-
c+c, C=C-eC+Cc-41-c-"c+
cCz difference C+CC': C→C±cc±Cmi voice CC old (+
C-? (C-? (Including 0C CfC'eC-1?c (fQ6(t(C+C-&C
+ CC main C town ec c = c spirit c + c C position ()C 竺cc ccmain c9: cmain C Table 11 - Margin Table 12 Below Margin Examples of reaction nuclei containing cyclic structures are shown in Table 13.

These reaction nuclei are combined with the terminal descriptors listed in Table 11.12. An example is ITS63.
64 (reaction graphs 63rc, 64rc, reaction nucleus 6
35g, 64sg).

In addition, there are also n-clause subgraphs that do not contain terminal descriptors and consist only of reaction nuclei (containing n carbon atoms); for example, in the case of a four-membered ring, ITS65.66
The cases of six-membered rings are shown in ITS67 to ITS69, respectively.

Margin below Table 13 Margin below 6868 ro 68 rk Margin below Next, the effective subgraph of a reaction including two reaction results (two reactions) will be explained, taking as an example the reaction of reducing a carbonyl group to a methylene group.

The imaginary transition structure of this reaction is expressed as ITS71, and the reaction graph is 71rc. An effective subgraph 71ge can be extracted from the ITS 71, and 71ge is also included in the reaction graph 71rc.

71ge, O(2) + C(1) combination H(3)
and O(2) +C(1)÷H(14) each correspond to one reaction result. When the effective subgraph 71ge, in which the reaction nucleus is the central carbon atom, is hierarchized based on the terminal atoms according to the method described above, the family (71 "a)" and the order (
71or) and clause (71se) level J higher descriptors are obtained.

Below is another example of the effective subgraph of the full white 719e 71fa 71゜r715e margin two cotton reaction ITS72~7
4, and the reaction graphs 72rc to 74rc,
Also listed are the results of hierarchizing the effective subgraphs 72ge to 74ge.

729e72fo72or725e Below margin 373rc 73ge73fa 73o, 73se474
rc ○*(JN Heti″C3Het zsc*z
A$c$A74 g (s 741 a
In 74°r745eITS75.76, the effective subgraphs 75ge and 76se are reaction nuclei (C
, nucleus) and a terminal descriptor, and by layering the terminal descriptors, the upper terminal descriptor 75 fa
, 75o", 75se and 76 fa, 760r, 7
6se is obtained.

Margin below 75　　　　　　75,.

59e 75,. 75or755e 76 76rc769e ↓ 76fo76°, 765e In this way, even an effective subgraph containing Nikio can be divided into a reaction nucleus and a terminal descriptor. Typical examples of node-level terminal descriptors when the carbon atom of the reaction nucleus is common to two reaction lines are shown in Tables 14 and 15.
Shown in the table.

As mentioned above, from the effective subgraph derived from the ITS and/or bonding table, higher descriptors can be generated using the terminal atoms as clues, and furthermore, the number of carbon atoms in the reaction nucleus and the reaction line can be generated. Classified by number. Table 16 summarizes the hierarchy of effective subgraphs.

Below, in the margin Table 14 A8 to PA A'' You can add something.

This allows the organic reaction to be performed in the first order based on the effective subgraph.
It is classified and systematized as shown in Table 6. Furthermore, it is preferable that each level of hierarchy be associated with a reaction name (predicate in natural language). If there is no corresponding existing predicate, the subgraph is recorded and saved as is or coated.

The obtained effective subgraph and its hierarchical list may be stored (registered) in a computer, or may be displayed or recorded via a suitable display/recording means. In this case, the effective subgraph may be blocked by the ITS or connection table of the reaction, and the list thereof may be
It may be associated with the ITS or bond table of the reaction.

Registration in a computer is performed by recording and storing information in the main storage device within the computer or via a suitable recording medium (magnetic disk, optical disk, magnetic tape, etc.). Further, display recording is performed by recording on various recording materials such as plain paper using a suitable recording device, or by displaying on a color cathode ray tube connected to a computer or a one-column display.

When registering valid subgraphs in a computer, numbers may be assigned to the valid subgraphs to facilitate information storage, management, and retrieval, or a common reaction name may be registered together for each valid subgraph. You may.

If the reaction information registered in the computer contains an effective subgraph as additional information, it can be used in chemical substance structure search systems, chemical reaction search systems, organic synthesis design systems, etc. based on this effective subgraph. It can be widely used. In addition, integrating reactions that span multiple stages and representing them as multiple effective subgraphs is also an integral part.

That is, it is possible to display not only individual reactions but also entire complex reaction routes such as those in organic synthesis, or extract only a part of the reactions and display them as reaction types.

By combining the effective subgraph information obtained by the method of the present invention with already registered chemical reaction information or chemical substance information involved in the reaction, molecular design focusing on specific properties of substances, organic It can be effectively used in a wide range of computer-based fields in chemistry, such as determining synthetic routes for compounds and automatically determining the structure of unknown compounds.

Claims (1)

[Scope of Claims] 1. A method for processing information regarding a chemical reaction that produces at least one product from at least one starting material; the information regarding the chemical reaction is based on a topologically superimposed structure of the starting material; and the structure of the product: (1) bonds between nodes that exist in common in the starting material and product, (2) bonds between nodes that exist only in the starting material, and (3) bonds that exist only in the product. (2) nodes that exist only in the starting material; (2) nodes that exist only in the starting material; and/or (3
) Contains bonds between nodes that exist only in the product substance and a series of nodes connected by the bonds, and the series of nodes includes one or more adjacent carbon atoms and non-containing atoms adjacent to the carbon atom. The effective subgraph consisting of carbon atoms is
A method of processing chemical reaction information comprising extracting from the imaginary transition structure and/or bond table. 2. Consisting of the above (2) bonds between nodes that exist only in the starting material and/or (3) bonds between nodes that exist only in the product material, and n adjacent nodes that are connected by the bonds. , and the two terminal nodes are non-carbon atoms, and the remaining n-2 nodes are carbon atoms (where n is an integer of 3 or more). 2. The method for processing chemical reaction information according to claim 1, wherein the chemical reaction information is extracted from a combination table. 3. In the above connection table, nodes are indicated by their types and consecutive numbers, and connections between nodes are indicated by a pair of numbers (
a, b) [wherein a is an integer representing the bond multiplicity in the starting material, and b is an integer representing the change in the bond multiplicity in the chemical reaction]. A method for processing chemical reaction information according to scope 1. 3. The method for processing chemical reaction information according to claim 1, wherein the effective subgraph is obtained as a two-dimensional figure, a connection table, and/or a code. 4. A method for processing chemical reaction information according to claim 1, characterized in that the effective subgraph is recorded and saved. 5. A method for processing chemical reaction information according to claim 1, characterized in that the effective subgraph is displayed and recorded on a recording material or on a screen of a display device. 6. A method for processing information regarding a chemical reaction that produces at least one product from at least one starting material; bonds between nodes that exist only in the starting material and/or bonds between nodes that exist only in the product. and a series of nodes connected by the bond, and the series of nodes consists of one or more adjacent carbon atoms and a non-carbon atom adjacent to the carbon atom, the bond A method of processing chemical reaction information consisting of classification based on combination, number of carbon atoms, number and type of non-carbon atoms. 7. Contained in the above effective subgraph, 1) A reaction sequence consisting of bonds between nodes that exist only in the starting material, bonds between nodes that exist only in the product material, and nodes that are alternately connected by the bonds. 2) the number of adjacent carbon atoms, 3) the type of bond between carbon atoms and non-carbon atoms, 4) the type of non-carbon atoms. A method for processing chemical reaction information according to scope item 6. 8. The method for processing chemical reaction information according to claim 6, characterized in that chemical reaction information can be searched based on the hierarchical effective subgraph. 9. Claims characterized in that chemical reaction information can be searched by a reaction nucleus comprised of one or two or more adjacent carbon atoms and bonds between them, included in the effective subgraph. The method for processing chemical reaction information according to item 8. 10. Chemical reaction information can be searched by a terminal descriptor consisting of a non-carbon atom and a bond between the non-carbon atom and the carbon atom, which is included in the effective subgraph. How to process chemical reaction information. 11. The method for processing chemical reaction information according to claim 6, characterized in that the classified effective subgraphs are associated with reaction names. 12. The method for processing chemical reaction information according to claim 6, characterized in that the classified effective subgraph is encoded. 13. The method for processing chemical reaction information according to claim 6, characterized in that the classified effective subgraphs are recorded and stored. 14. A method for processing chemical reaction information according to claim 6, characterized in that the classified effective subgraphs are displayed and recorded on a recording material or on a screen of a display device.