WO2003007187A1

WO2003007187A1 - Method of presuming ligand and method of using the same

Info

Publication number: WO2003007187A1
Application number: PCT/JP2002/007057
Authority: WO
Inventors: Hiroshi Inooka; Yoshio Yamamoto
Original assignee: Takeda Chemical Industries, Ltd.
Priority date: 2001-07-12
Filing date: 2002-07-11
Publication date: 2003-01-23

Abstract

A method for estimating a ligand or a ligand type directly binding (or coupling) to GPCR based on GPCR sequencial data and, in its turn, estimating its function. More specifically, a method of presuming a ligand of a protein with unknown ligand which comprises obtaining classification data of proteins with known ligands wherein the alignments of the proteins with known ligands correspond respectively to the ligands or ligand types, obtaining ligand-determining residue-ligand classification data which shows the correlationships among ligand-determining residues and ligands or ligand types to thereby obtain the alignments of proteins with unknown ligands, applying the data at least concerning the ligand-determining residues in the alignments of the proteins with unknown ligands as described above to the ligand-determining residue-ligand classification data, and thus estimating the ligand or ligand type of the protein with the unknown ligand, etc.

Description

Specification

FIELD OF THE INVENTION

The present invention relates to a method for predicting a binding molecule of an unknown binding molecule protein, a method for producing a medicine using the prediction method, and a computer for predicting a binding molecule of the unknown binding molecule protein. For predicting a binding molecule or a type of a binding molecule of an unknown protein of a binding molecule from information on the amino acid sequence (or sequence alignment) of the protein and the type of the binding molecule or the type of the binding molecule, a medicine using the method, and the method It relates to a computer used for. Furthermore, the present invention relates to a method for predicting the function of a protein whose binding molecule is unknown through prediction of the binding molecule and the type of the binding molecule. Background art

Sequence information of the human genome has been released (Nature Vol. 409, No. 6822 (2001)); Science Vol. 291 and No. 5507 (2001) ), 30,000 to 40,000 genes, including known genes, were reported. However, there are many unknown ones whose relationship between genes and intracellular functions is unknown. In particular, information on genes that cause disease is not only important for clarifying its pathology, but also indispensable for developing drugs used for its diagnosis and prevention and treatment. . Molecules that can be targeted for drug development are those that bind directly or indirectly to other molecules when the drug containing the molecule enters the body, and is most likely to be a protein. Examples of the protein include a protein that binds a low-molecular substance, an enzyme that has catalytic activity for a low-molecular or high-molecular substance, and a protein that binds a high-molecular substance. Among them, there are a very large number of drugs that target proteins that bind low molecular substances. Proteins of particular interest as target molecules include G protein-coupled receptor proteins involved in signal transduction as membrane proteins.

In the case of membrane proteins, signal transduction from extracellular to intracellular involves transmembrane domain It proceeds by causing a conformational change in the transmembrane domain of the in-containing membrane protein. The above G protein-coupled receptor protein having a transmembrane domain

(Groteincouplederecrecint, hereinafter sometimes abbreviated as GPCR) is a membrane protein involved in signal transduction having a seven transmembrane domain in its molecule. GPCRs have an N-terminal outside the cell, have a topology with a C-terminal inside the cell via a seven transmembrane domain, and the ligand binds to the N-terminal region or transmembrane region depending on the type. It is also known that G protein binds to the intracellular C-terminal part and intracellular loop 2. To date, the only crystal structure of oral dopsin, a member of the GP CR family, has been analyzed (Palczewski et al., Science, Vol. 289, pp. 739-745 (2000)). There are no reports on structural analysis. In addition, various biological experiments have revealed that GPCR activation occurs as follows. That is, (i) the ligand binds to the receptor and the structure of the GPCR changes. (Ii) Conjugation changes of the GPCR are transmitted to the conjugated G protein, and a part of the G protein is released. (Iiii) Information (signal) is transmitted into cells. Molecules that specifically bind to the receptor are called ligands. In the case of GPCRs, known ligand molecules include living amines such as dopamine, serotonin, melatonin, and histamine, lipid derivatives such as prostaglandins and leukotrienes, amino acids such as nucleic acid and glutamine, angiotensin, secretin, and somatosutin. And physiologically active peptides such as

The G protein functions as a transducer in the signal transduction system, and is composed of three subunits, i, β, and α. Among these, the α subunit has the activity of binding to GTP and hydrolyzing G Τ 、, and is a subunit unique to each G protein. That is, the function of GPCR is determined by the type of Ga protein to be conjugated. To date, Go; G _s and G or G as proteins. And _Gt have been isolated and purified, and their properties have been investigated.

G protein-coupled receptor protein is present on the surface of each functional cell in living cells and organs, and is a target for molecules that regulate the function of those cells and organs, such as hormones, neurotransmitters, and biologically active substances. Plays a physiologically important role. The receptor transmits a signal into the cell via binding to a physiologically active substance, and this signal causes various reactions such as suppression of activation and activation of the cell.

To clarify the relationship between substances that regulate complex functions in various cells and organs and their specific receptor proteins, especially G protein-coupled receptor Yuichi proteins, it is important to clarify the complex functions in various cells and organs. Will provide a very important tool for drug development closely related to these functions.

In recent years, as a means of analyzing genes expressed in vivo, studies on random analysis of cDNA sequences have been actively conducted, and the resulting cDNA fragment sequences are expressed in Expressed Seauence Tag ( EST) is registered in the database and published. However, many ESTs provide only sequence information, and it is difficult to estimate the function of the sequence.

Not all G protein-coupled receptors have been found, and even at this time, there are many unknown G protein-coupled receptors and so-called orphan receptors for which no corresponding ligand has been identified. It is present, and the search for new G protein-coupled receptors and the elucidation of their functions are eagerly awaited.

With the release of genome sequence information, it has become possible to examine gene regions that can be presumed to have some function by comparing with known sequence information. Prediction of binding molecules and functions of novel proteins has been carried out based on the similarity between the nucleotide sequences of genes or the amino acid sequences of translated proteins. As described above, Clustal W, BLAST, and the like are used as software for calculating sequence similarity and displaying an appropriate correspondence between sequences (sequence alignment). Using sequence alignments obtained from these software and similarity analysis, function prediction is performed based on the concept that a new protein has a function similar to that of the similar protein. Predictions based on such similarities contribute to the prediction of basic function, but have little effect on prediction of binding molecules to the same group of proteins. For example, it is useful for classifying whether the new protein is a GPCR, but it is not suitable for predicting the type of ligand-conjugated G protein that binds to the GPCR unless the similarity is extremely high. To search for and determine the unknown ligand molecule that actually binds to the GPCR, Although it is necessary to measure the activity of many candidate compounds using an extract or the like, there is a problem that candidate compounds of binding molecules having different physical properties cannot be simultaneously evaluated. Therefore, if it was possible to predict the type of ligand molecule without performing complicated activity measurements, it would lead to efficient determination of the ligand molecule. Therefore, it is necessary to understand protein sequence information with a concept different from similarity. If it is possible to predict the molecules that bind to each protein for each member of a protein group (for example, GPCR) classified by similarity, it is useful for identifying ligands and analyzing the functions of the proteins and ligands. is there. For this purpose, studies have been conducted to predict the three-dimensional structure from the one-dimensional sequence information of the protein, and to predict the binding molecule by a binding experiment on a computer. However, at present, it is not practical because of the lack of accuracy of prediction of the three-dimensional structure of the protein and binding experiments on a computer. It is useful in the technical field of the present invention to be able to avoid such low accuracy and complexity and to directly examine binding molecules and functions from sequence information.

Methods for predicting binding molecules and functions from sequence information include, for example, mutations of amino acids (and their mutation positions) that are considered to be involved in determining the type of G protein to be coupled to a limited number of known GPCRs. (Bulseco and Scimerlik, Molecular Pharmacology, 49: 132-141 (1996); Burste in, Spalding and Brann, Biochemistry, 37: 4052-4058 (1998); Kazmi et al., Biochemistry, 39: 3734-3744 (2000)), there is no report that predicts the type of Gα protein coupled to a novel GPCR.

Also, for GPCRs, a method for predicting the site involved in binding of a ligand to its GPCR by analyzing mutations in both GPCR and ligand amino acids, that is, a correlated mutation analysis (CMA) method was developed (Singer et al., Receptors and channels, 3: 89-95 (1995)), there is no report that a GPCR sequence alone was used to select unknown GPCR binding molecules (and / or types of binding molecules).

As described above, for example, in the case of GPCRs, it is not known how to predict molecules or types of molecules that bind (or conjugate) to GPCR directly from the sequence information of GPCR, and predict the function. Not been. Regarding GPCRs, if a new GPCR is discovered, in order to manufacture a drug using the GPCR, it must bind to (or conjugate to) the GPCR after understanding the function of the GPCR. There is a problem that it is necessary to grasp the molecule or the kind of the molecule through experiments or the like, which requires enormous cost and labor. Disclosure of the invention

The present invention relates to a method for simply and accurately predicting a molecule or the like that can bind to sequence information of unknown function, a method for producing a medicine using the method, and a computer used for the method.

At least one of the above-mentioned problems is solved by the following invention.

(1) A binding molecule unknown protein prediction method for predicting a binding molecule that binds to a binding molecule unknown protein, wherein the amino acid sequence and the binding molecule are known. Obtaining a binding molecule known protein classification information in which the sequence alignment of the binding molecule known protein is associated with the type of the binding molecule or the binding molecule; and using the binding molecule known protein classification information, the binding molecule known protein. Specifying one or more binding molecule-determining residue positions that are assumed to be involved in determining a binding molecule among the sequence alignment positions of the amino acid sequence; and aminos at the binding molecule-determining residue positions. By associating acid residues (binding molecule determining residues) with binding molecules or types of binding molecules Obtaining a binding molecule-determining residue-binding molecule classification information indicating a correlation between the binding molecule-determining residue and the binding molecule or the type of the binding molecule; and for a binding molecule unknown protein of the same type as the binding molecule-known protein. Aligning the sequence of the unknown binding molecule protein with respect to the sequence alignment between the known binding molecule proteins, and obtaining a sequence alignment of the unknown binding molecule protein; at least one of the sequence alignments of the unknown binding molecule protein; Applying the information on the determined binding molecule to the classified information on the determined binding molecule-binding molecule to predict the binding molecule or the type of the binding molecule of the unknown binding molecule, and the binding molecule of the unknown binding molecule. Forecasting method.

(2) The binding molecule is any one of a ligand, a regulator, an effector, and a coenzyme. The method for predicting a binding molecule of an unknown protein according to the above (1), wherein

(3) The method for predicting a binding molecule of an unknown protein for a binding molecule according to (1) or (2), wherein the binding molecule is classified into two or more types, and the type of the classified binding molecule is predicted.

(4) the binding protein unknown protein is any one of a G protein-coupled receptor, a kinase, a lipase, a transporter, a protease, and an ion channel according to any of (1) to (3) above; Method for predicting binding molecules of unknown proteins.

(5) In the step of identifying one or more binding molecule-determining residue positions, the one or more binding molecule-determining residue positions are identified from amino acid residues constituting the sequence alignment and types of binding molecules. The method for predicting a binding molecule of an unknown protein according to any one of (1) to (4).

(6) The binding molecule of the unknown binding molecule protein according to any one of (1) to (4) above, wherein the binding molecule-determining residue position is determined using either or both of Formula 1 and Formula 2. Forecasting method.

(7) determining the position of the binding molecule-determining residue using at least one of Formulas 3, 4, and 5; and determining the position of the unknown binding molecule protein according to any one of (1) to (4) above. Binding molecule prediction method.

(8) The step of obtaining the ligand-determined residue-ligand classification information extracts the amino acid residues of the ligand-known protein at the position of the ligand-determined residue with the smallest value of the function f 3 (n). Determining the number (A) of ligand-determined residues that match the extracted ligand-determined residues among the ligand-known proteins listed in the ligand-determined residue-ligand classification information; Obtaining a number (B) of the known ligand proteins which correspond to the extracted ligand-determining residues among the known ligand proteins, wherein the type of the ligand or the ligand matches that of the known ligand protein; and Among the amino acid residues of known proteins, the value of the function f3 (n) is the second smallest or the Xth (where X represents an integer greater than 2 and less than 100). Extracting the one at the position of the smallest ligand-determined residue, and the ligand-known protein listed in the ligand-determined residue-ligand classification information Determining the number (C) of the extracted ligand-determining residues that match the extracted ligand-determining residues; and matching the extracted ligand-determining residues among the ligand-known proteins listed in the ligand-determining residue-ligand classification information. A step of obtaining a number (D) in which the type of ligand or ligand matches that of the known protein of the ligand; a step of obtaining the sum (E) of (A) and (C); ) And (D) for obtaining the sum (F), and further obtaining (E) and (F) ligand-determined residue-one-ligand classification information which further displays (E) and (F). Protein binding molecule prediction method.

(9) For at least two or more known binding molecule proteins whose amino acid sequence and binding molecule are known, binding in which the sequence alignment of the binding protein known protein is associated with the binding molecule or the type of binding molecule A step of obtaining information on the classification of known protein molecules, and a binding molecule which is a position assumed to be involved in determining a binding molecule in the sequence alignment of the known binding molecules using the classification information on known binding molecules. By associating one or more determined residue positions with the amino acid residue (binding molecule determining residue) at the relevant binding molecule determining residue position and the binding molecule or the type of the binding molecule, The binding molecule, which represents the correlation between the determined residue and the binding molecule, and the step of obtaining the binding molecule classification information. Binding molecules prediction method of molecular unknown protein.

(10) The binding molecule determination residue indicating the correlation between the binding molecule determination residue and the binding molecule. The binding molecule classification information includes the binding molecule unknown protein of the same type as the binding molecule known protein. By inputting information on binding molecule-determining residues in the sequence alignment of the unknown binding molecule protein obtained by aligning the sequence of the unknown binding molecule protein with the sequence alignment between the proteins, A binding molecule prediction method for a binding molecule unknown protein that predicts the binding molecule to be bound or the type of the binding molecule.

According to this method, it is possible to predict the binding molecule or the type of the binding molecule only by obtaining information on the amino acid sequence of the protein whose binding molecule is unknown and the sequence alignment obtained using Z or the amino acid sequence. . This makes it much faster than conventional molecular modeling methods that predict even three-dimensional structures. It can predict the binding molecule (ligand etc.) at a low cost. Further, according to the present invention, it is possible to predict the binding molecule or the type of the binding molecule for a protein in which various types of binding molecules are unknown. In addition, it is possible to predict the binding molecule or the kind of the binding molecule easily and quickly than by experimenting whether or not any candidate binding molecule actually binds to the protein whose binding molecule is unknown. Determining residue of binding molecule By obtaining the classification information of one binding molecule, the sequence is applied only to the sequence alignment of the unknown protein, and the information is applied to the table, and the type of the binding molecule or binding molecule that binds to the unknown protein. Can be easily predicted.

Further, at least one of the above problems is solved by the following invention. That is,

(11) Using the method for predicting a binding molecule of an unknown binding molecule described in any of (1) to (10) above, predicting the binding molecule or the type of the binding molecule that binds to the unknown binding molecule A method for producing a medicament comprising the step of:

(12) The above-mentioned (11) wherein the drug is one or both of a preventive agent and / or a therapeutic agent for a central disease, an inflammatory disease, a circulatory disease, a cancer, a metabolic disease, an immune system disease or a digestive system disease. )).

Further, at least one of the above-mentioned problems is solved by the following invention. That is, (13) a method for determining the position of a residue for determining a binding molecule using either or both of the formulas (6) and (7).

(14) a method for determining the position of a binding molecule-determining residue using Formula 8,

(15) Using the amino acid sequence or sequence alignment of the known binding molecule protein and information on the type of the binding molecule or the binding molecule, the binding molecule in the position of the sequence alignment of the protein with the known binding molecule is used. Molecule binding residue, which is an amino acid residue at a position supposed to be involved in the determination of the binding molecule (binding molecule determining residue position), and a binding molecule determination residue indicating the correlation between the binding molecule or the type of the binding molecule. A computer for predicting a binding molecule of a protein whose binding molecule is unknown, which obtains group-binding molecule classification information. A sequence alignment input means for inputting information on the sequence alignment of the protein; an amino acid sequence or sequence alignment of the binding molecule known protein input by the sequence alignment input means; and a binding molecule or a binding molecule. A sequence alignment binding molecule storage means for storing information on the type; an amino acid sequence or sequence alignment of the binding molecule known protein stored by the sequence alignment binding molecule storage means; and a type of the binding molecule or the binding molecule. A binding molecule determining residue position determining means for determining the binding molecule determining residue position using information; an amino acid residue (binding molecule determining residue) at the binding molecule determining residue position; Type and By associating, the binding molecule determination residue-binding molecule classification information obtaining means for obtaining the binding molecule determination residue-binding molecule classification information indicating the correlation between the binding molecule determination residue and the binding molecule or the type of the binding molecule; Information on the sequence alignment of the unknown binding molecule protein obtained by aligning the sequence of the unknown binding molecule protein with respect to the sequence alignment between the unknown binding molecule proteins for the same type of unknown binding molecule protein as the known binding molecule protein Inputting sequence alignment input means, and using the information on the sequence alignment of unknown binding molecule proteins input by the sequence alignment input means as the binding molecule determination residue-binding molecule classification information. Bound molecule unknown protein Binding molecule, or to predict the kind of binding molecules, binding molecules unknown evening protein computer for predicting the binding molecule,

(16) The binding molecule-determining residue position determining means predicts the binding molecule of the binding molecule unknown protein according to (14) using at least one of the functions of Formula 9 and Formula 10 or both functions. Computer,

(17) The binding molecule-determining residue position determining means uses the function represented by Formula 9 to predict the binding molecule of the binding molecule unknown protein described in (15) or (16) above. Computer,

(18) A computer for predicting a binding molecule of a binding molecule unknown protein, wherein the computer is a sequence alignment of a binding molecule known protein having the same type as the binding molecule unknown protein and a binding molecule known. The binding And a binding molecule determining residue position that is assumed to be involved in determining a molecule that binds to a known protein, and a binding molecule determination that is an amino acid residue of the binding molecule known protein at the binding molecule determining residue position. A storage means for storing information on the residue and the type of binding molecule or binding molecule of the binding molecule known protein corresponding to the binding molecule determination residue; and a binding molecule unknown of the same type as the binding molecule known protein. Sequence alignment input means for inputting information on the sequence alignment of the unknown binding molecule protein obtained by aligning the sequence of the unknown binding molecule protein with the sequence alignment between the binding molecule known proteins. Information and storage means for the entered sequence alignment A binding molecule determining means for determining the binding molecule or the type of the binding molecule of the unknown binding molecule protein from the stored information; and displaying the determined binding molecule or the type of the binding molecule binding to the determined unknown binding molecule protein. Display means, the information relating to the sequence alignment of the unknown protein of the binding molecule input by the sequence alignment input means, and the binding molecule determination residue and the binding molecule determination residue stored in the storage means. Based on the information on the binding molecule of the corresponding binding molecule known protein or the information on the type of the binding molecule, the binding molecule determination unit predicts the binding molecule or the type of the binding molecule of the unknown binding molecule protein, and is predicted by the binding molecule determination unit. Display the binding molecule or the type of the binding molecule of the unknown binding molecule by the display means. Computer for predicting the binding molecules of the child unknown protein - is even. According to such a computer, it is possible to obtain binding molecule-determined residue-binding molecule classification information based on the sequence alignment of the binding molecule known proteins, and thereby to obtain the sequence alignment of the binding molecule unknown protein. Only by obtaining, the binding molecule or the type of the binding molecule can be easily predicted.

Furthermore, the present invention

(19) A computer is connected to a sequence alignment input means for inputting information on the sequence alignment of the known binding molecule protein, and the amino acid sequence or the amino acid sequence of the binding molecule known protein input by the sequence alignment input means. Sequence alignment and binding molecules or types of binding molecules Sequence-binding molecule storage means for storing information on the amino acid sequence or sequence alignment of a known binding molecule protein stored by the sequence alignment-binding molecule storage means, and information on the type of the binding molecule or the binding molecule. A binding molecule-determining residue position determining means for determining the binding molecule-determining residue position using: a binding molecule determining residue position; an amino acid residue (binding molecule determining residue) at the binding molecule determining residue position; By associating the type with the binding molecule determining residue and the binding molecule or the type of the binding molecule, the binding molecule representing the correlation between the binding molecule determining residue and the type of the binding molecule is obtained. And the above step, for the unknown binding molecule protein of the same type as the known binding molecule protein. Sequence alignment input means for inputting information on sequence alignment of unknown binding molecules obtained by aligning sequences of unknown binding molecules with respect to sequence alignment between known proteins of combined molecules. program,

(20) The program according to the above (19), wherein the binding molecule determining residue position determining means uses at least one of the functions of the formulas 12 and 13 or both.

(21) The program according to (19) or (20), wherein the binding molecule determining residue position determining means uses a function represented by Formula 14.

(22) Using a computer to determine a molecule that binds to a protein with a known binding molecule in a sequence alignment of a protein with a known binding molecule that has the same type and binds to the protein with an unknown binding molecule. Binding molecule-determining residue position that is supposed to be involved in the binding molecule; binding molecule-determining residue that is an amino acid residue of a binding molecule known protein at the binding molecule-determining residue position; Storage means for storing information on the binding molecule or the type of binding molecule of the binding molecule known protein corresponding to the group; and a sequence alignment between the binding molecule known protein for the same type of binding molecule unknown protein as the binding molecule known protein. Unknown protein obtained by aligning the sequence of unknown protein to And Sequence § Line Instrument input means for inputting information about the Sequence alignment of Park protein, input Sequence § Line ment Information storage means the binding molecule from the information stored in the unknown protein A binding molecule determining means for determining the binding molecule or the type of the binding molecule, and a program functioning as a display means for displaying the determined binding molecule or the type of the binding molecule binding to the unknown binding molecule protein; and

(23) A recording medium storing the program according to any one of the above (19) to (22) is provided. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 shows a process chart from creation of ligand-determined residue-ligand classification information of the present invention.

FIG. 2 is a process chart showing one embodiment of the ligand-determining residue position specifying step of the present invention. FIG. 3 is a process chart showing another embodiment of the ligand-determining residue position specifying step of the present invention.

FIG. 4 shows the results of sequence alignment between silodopsin and TGR 23-1.

FIG. 5 shows the activity of increasing the intracellular Ca ion concentration of TGR23-1-expressing CHO cells with various concentrations of human TGR23.2 ligand (1-20) measured using FLIPR.

FIG. 6 shows the activity of increasing the intracellular Ca ion concentration of TGR23-2 expressing CHO cells by various concentrations of human TGR23-2 ligand (112) measured using FLIPR. BEST MODE FOR CARRYING OUT THE INVENTION

The present invention relates to a known binding molecule protein whose amino acid sequence and binding molecule are known, wherein the binding alignment is performed by associating at least two or more binding molecule known proteins with the binding molecule or the type of the binding molecule. A step of obtaining molecularly known protein classification information; and a binding molecule that is assumed to be involved in determining a binding molecule among sequence alignment positions of the known binding molecule protein using the binding molecule known protein classification information. One or more determined residue positions Determining the binding molecule-determining residue and the binding molecule or the binding molecule by associating the amino acid residue (binding molecule determination residue) at the binding molecule-determining residue position with the type of the binding molecule or the binding molecule. Obtaining the binding molecule-determining residue-binding molecule classification information indicating the correlation with the type of the binding molecule; and a sequence alignment between the binding molecule unknown proteins of the same type as the binding molecule unknown protein. Aligning the sequence of the unknown binding molecule protein to obtain a sequence alignment of the unknown binding molecule protein, and obtaining information on the binding molecule determining residue in the sequence alignment of the unknown binding molecule protein. By applying the binding molecule-determined residue-binding molecule classification information to the binding molecule unknown protein binding molecule or The present invention relates to a method for predicting a binding molecule of an unknown protein, which predicts the type of the binding molecule.

A binding molecule known protein means a protein for which a biomolecule that binds to the protein is known. For example, it refers to a protein to which a biological molecule specifically binds, such as a receptor whose ligand is known.

An unknown binding molecule protein refers to a protein of the same type as a known binding molecule protein, wherein the binding molecule is unknown. A sequence alignment of the unknown binding molecule protein obtained by aligning the sequence of the unknown binding molecule protein with respect to the sequence alignment between the known binding molecule proteins with respect to the same type of unknown binding molecule protein as the known binding molecule protein. In order to examine the similarity (homology) between the amino acid sequence of the unknown protein of the binding molecule and the amino acid sequence of the same type of protein as the unknown protein of the binding molecule, the insertion, taking into account the substitution, insertion, and deletion, A gap was inserted at the position corresponding to the deletion, and the entire sequence was juxtaposed.

Examples of unknown binding molecule proteins and known binding molecule proteins include G protein-coupled receptors (GPCRs), kinases, lipases, transporters, proteases, and ion channels. Of these, the present invention can be preferably applied to G protein-coupled receptors (GPCRs) and kinases. When the unknown binding molecule protein and the known binding molecule protein are G protein-coupled receptors (GPCRs), the unknown binding molecule protein is also called an orphan receptor. What As described above, the unknown binding molecule protein and the known binding molecule protein are of the same type. For example, if the protein whose binding molecule is unknown is a GPCR, the protein whose binding molecule is known is also a GPCR.

[Classification information of known binding molecules]

The binding molecule known protein classification information means a table in which sequence alignment of at least two or more binding molecule known proteins is associated with the binding molecule (and Z or the type of the binding molecule). This table is not particularly limited as long as it is in a form that can be stored electronically and visually recognized as a table, not only on paper. In addition, there is no particular limitation as long as the correspondence between the sequence alignment of the known binding molecule protein and the binding molecule or the type of the binding molecule can be recognized.

[Positioning residue for binding molecule]

The binding molecule determining residue position means a position of a sequence alignment of a protein having a known binding molecule, which is assumed to be involved in determining a binding molecule.

The number of ligand-determining residue positions is not particularly limited as long as it is 1 or more, preferably 1 or more and 10 or less, more preferably 2 or more and 6 or less, and particularly preferably 2 or more. There are only 20 types of amino acid residues that make up the living body. Therefore, up to 20 ligands can be assigned using only one ligand-determining residue. On the other hand, for example, in GPCR, more than 100 kinds of ligands are known, so that if there is only one ligand-determining residue position, all ligands cannot be associated with ligand-determining residues. Therefore, in a system having more than 20 kinds of ligands, such as a GPCR, in order to predict the ligand of a protein whose binding molecule is unknown, it is desirable that there are two or more ligand-determining residue positions. Also, the prediction accuracy increases as the number of ligand determination residue positions increases. When the ligands are classified into several types (X1 to Xp: p is the number of classifications), and the classification of the ligand of the unknown protein belongs to which classification, the number of ligands (value of ρ) is small. For example, it is possible to predict the type of ligand using only the amino acid residue at one ligand-determining residue position. However, even in such cases, amino acids at two or more ligand-determining residue positions are generally The prediction accuracy becomes higher when the type of ligand is predicted by combining acid residues.

[Binding molecule determining residue]

The binding molecule determining residue means an amino acid residue at the binding molecule determining residue position. In addition, specific residue positions (one or two or more) of a plurality of types of binding molecules and a plurality of types of amino acid residues may be combined with a known protein of the binding molecule to be used as a residue for determining a binding molecule. For example, the second and eighth amino acid residue positions in the sequence alignment are taken as examples of the binding molecule-determining residue positions, and the ninth and eleventh amino acid residues in the sequence alignment are taken as other examples. This is the case. Combining different binding molecule determining residue positions in this manner makes it possible to improve the accuracy of predicting the binding molecule or the type of the binding molecule. The information on the binding molecule determining residue means information on the amino acid residue at the binding molecule determining residue position. For example, if the second and eighth positions in the sequence alignment of the protein are binding molecule determinant residue positions, the second and eighth positions in the sequence alignment are information on the positions of the binding molecule determinant residues. Then, in the sequence alignment, information on the types of the second and eighth amino acid residues and information on the position of the binding molecule determining residue are combined to provide information on the binding molecule determining residue.

[Binding molecule determination residue-binding molecule classification information]

The binding molecule determining residue-binding molecule classification information is a table showing the correlation between the binding molecule determining residue and the binding molecule or the type of the binding molecule. This table is not particularly limited as long as it is in a form that can be stored electronically and visually recognized as a table as well as on paper. The binding molecule is not particularly limited as long as the correspondence between the binding molecule determining residue and the type of the binding molecule or the binding molecule can be recognized.

The binding molecule is not particularly limited as long as it can bind to known binding molecule proteins and unknown binding molecule proteins that are biopolymers.For example, ligands that bind to receptor proteins and GPCRs G α protein that binds. [Type of binding molecule]

The type of binding molecule is a type in which, when a plurality of binding molecules exist in the same type of known binding molecule protein, they are classified according to their functions and properties. For example, there are cases where GPCR ligands are classified into monoamines, lipids, and peptides.

[Computer]

The computer of the present invention is not particularly limited as long as it is an electronic device capable of performing a certain calculation or the like. For example, a known computer such as a personal computer, a super computer, and a mobile may be used. You may. A computer equipped with a browser is particularly preferable because it can be connected to the Internet and can access a well-known Web (Web) site. Hereinafter, the steps up to the generation of classification information for a ligand-determined residue-ligand will be described, taking the case where the binding molecule is a ligand as an example.

[Steps to create ligand-determined residue-ligand classification information]

FIG. 1 shows an example of a process from the generation of ligand-determined residue-ligand classification information, which comprises the following steps. That is, the step of obtaining information on sequence alignment and ligand (and / or ligand type) for at least two or more binding molecule known proteins whose amino acid sequence and ligand (and / or ligand type) are known ( S 101), sequence alignment and ligand (and type of Z or ligand) are associated, and sequence alignment ligand classification information obtaining step for obtaining binding protein known protein classification information (S 102), sequence alignment ligand classification information obtaining The ligand-determining residue positions, which are assumed to be involved in determining the ligand (and / or the type of ligand) in the sequence alignment of the known binding molecule proteins using the information on the classification of the known binding molecule proteins obtained by the process, are described. 1 Or the ligand-determining residue position specifying step (S103) for specifying two or more ligand-determining residue positions (S103); Is the ligand-determined residue that indicates the correlation between the ligand (and the type of Z or ligand) that is associated with the ligand-determined residue associated with the ligand type. This is a gand determining residue-ligand classification step (S104). Hereinafter, each step will be described. In the following, the term “ligand” may include the type of ligand.

[Step 1 0 1]

First, in order to create the ligand-determined residue-ligand classification information, the sequence alignment and ligand (and type of Z or ligand) for at least two or more binding molecule known proteins whose amino acid sequence and ligand are known are described. Information is acquired (S101). In this step, information on the amino acid sequence and ligand (and / or type of ligand) is obtained for a plurality of known binding molecule proteins, but sequence alignment may be obtained from the amino acid sequence, or a plurality of known binding molecule proteins may be obtained. If sequence alignment has already been requested for, information on the sequence alignment may be obtained directly. The method for obtaining sequence alignment and information on ligands (and Z or the type of ligand) for proteins with known binding molecules is not particularly limited. Even if the information is obtained from a database, the information is obtained by calculation. May be obtained. The database is preferably a database that contains information on sequence alignments and ligands (and / or types of ligands) of at least 100 or more types of known binding molecules, and more preferably 500 or more. It is particularly preferable to record 1000 or more. This is because the greater the number of proteins with known binding molecules, the higher the accuracy of the ligand-determined residue-ligand classification information.

The known database is not particularly limited as long as it describes the sequence alignment and ligand (and Z or the type of ligand) of the binding molecule known protein. For example, GPCRDB (ht tp: / /www.GPCR.org/7tm/). Sequence alignment can also be obtained by a known calculation method. Known calculation methods for sequence alignment include, for example, Clus tal W and BLAST, but may be calculated manually. In addition, a classification for the ligand is created in advance, and if the information on the ligand is input, the type of the ligand is automatically obtained. If the information on the ligand is obtained, the information on the type of the ligand is also obtained. You may make it available. [Step 1 0 2]

Next, the sequence alignment and the information on the ligand (and the type of Z or ligand) for the two or more binding molecule known proteins obtained in step 101, and the sequence alignment and the information on the ligand (and type of Z or ligand) are obtained. To obtain the classification information of the binding molecule known protein (sequence alignment ligand classification information obtaining step: S102). In step 101, when sequence alignment and ligand (and / or ligand type) information on one or more known binding molecules is obtained from a database or the like, the sequence alignment and the ligand (and / or ligand) have already been performed. If the type is associated with the sequence alignment, the information may be obtained with the sequence alignment and the ligand (and the type of Z or ligand) associated with each other.

[Step 1 0 3]

Next, the ligand (and the type of ligand or ligand) in the sequence alignment of the binding molecule known protein is determined using the binding molecule known protein classification information obtained in the sequence alignment ligand classification information acquisition step of step 102. One or more ligand-determining residue positions, which are assumed to be involved in the above, are specified (ligand-determining residue position specifying step: S103).

In this step, one or more ligand-determining residue positions are specified by combining preferred functions according to the type and properties of the target protein. A preferred embodiment of this step will be described later.

[Step 1 04]

Next, the ligand (and / or the type of ligand) is associated with the amino acid residue (ligand-determining residue) at the ligand-determining residue position specified in the ligand-determining residue position specifying step. To obtain ligand-determined residue-ligand classification information indicating the correlation between the ligand and the ligand (ligand determined residue-ligand classification step: S104).

In this step, information on the ligand-determined residue positions, ligand-determined residues, and ligands (and / or types of ligands) is obtained from the sequence alignment ligand classification information for each GPCR, and the ligand-determined residue-one ligand is obtained. Obtain classification information.

By using such ligand-determined residue-ligand classification information, it is possible to predict the ligand (and Z or the type of ligand) even for a GPCR whose ligand is unknown, simply by determining the sequence alignment.

[One Embodiment of Step 103]

With reference to FIG. 2, a preferred embodiment of the ligand-determining residue position identification step (S103) will be described. This step is for the case where the number of ligand-determining residue positions is one. Ligands are classified into!) Types and classified into XI to Xp, respectively. The information on all sequence alignments and ligands (and / or types of ligands) of known binding molecule proteins present in the classification information on binding molecules known in step 102 is input to the following equation 15 and the function fl Since the value of (n) is small, it is determined as a candidate for a ligand-determined residue position (a candidate ligand-determined residue position selection step: S201). At this time, if the sequence alignment of the obtained candidate ligand-determined residue positions is not sequence-alignable (indicated by-), but is present in 3% or more of all GPCRs, the ligand-determined residue positions are determined. It is preferable not to adopt as a ligand-determining residue position

f l (n) = ∑ (N (Res, XQ) XN (Res, Xr)) Equation 15

Res

[In Formula 15, n represents fl (n) as an evaluation function for the nth amino acid residue in the sequence alignment of the binding molecule known protein, and Res represents the amino acid residue. XQ and Xr represent ligands, Q represents an integer from 1 to p-1, r represents an integer greater than Q and less than or equal to p, p represents the number of ligands, and N ( Res, XQ) represents the number of proteins in which the nth amino acid residue in the sequence alignment is Res and the ligand is among the known binding molecule proteins present in the binding molecule known protein classification information, N (Res, Xr) is the number of proteins whose binding amino acid residue is Res and the ligand is Xr among known binding molecule proteins in the binding molecule known protein classification information. To Represent. The number of candidate ligand-determining residue positions in the ligand-determining residue position candidate selecting step is not particularly limited, but is preferably 1 or more and 100 or less, more preferably 1 or more and 10 or less. It is particularly preferable that the value is not less than 5 and not more than 5. This is because if there are too many candidate ligand-determined residue positions, the subsequent step of confirming the reliability of candidate ligand-determined residue positions becomes difficult.

The candidate ligand-determined residue position selected in the candidate ligand-determined residue position selecting step in step 201 may be used as it is as the ligand-determined residue position candidate. It is more preferable to use and study. (Step of examining reliability of candidate ligand-determined residue position: S202). In this case, the reliability of candidate ligand-determined residue positions is examined using the following equation (16). The smaller the obtained value of f 2 (n), the more suitable the position of the ligand-determining residue. f 2 (n) = ∑ (N (Res, XI) xN (Res, X2) X N (Res, Xp)) Equation 16

Res

[In Equation 16, n represents that f 2 (n) is an evaluation function for the nth amino acid residue in the sequence alignment of the binding molecule known protein, and Res represents the amino acid residue. Represents the type of group, XI to ¾ represent the ligand or the type of ligand, p represents the number of the ligand or the type of ligand, and N (Res, X) represents the known binding molecule obtained in step 102. Among the known binding molecule proteins present in the protein classification information, this indicates the number of proteins in which the n-th amino acid residue in the sequence alignment is Res and the ligand is X. Using the value of fl (n) obtained in step 201 and the value of Z or f 2 (n) obtained in step 202, the position of a ligand-determining residue is identified (ligand-determining residue position identification step: S 203). The position of the obtained f 2 (ii) having the smallest value may be determined as the ligand-determining residue position, and the product of Π (n) and the value of f 2 (n) that is the smallest may be determined as the ligand-determining residue position. As a position Alternatively, the smallest sum of the values of Π (n) and f 2 (n) may be used as the ligand-determining residue position. In addition, among the amino acid residue positions of the known binding molecule proteins present in the binding molecule known protein classification information obtained in step 102, the positions of fl (n) are ranked in ascending order, and further, f 2 (n) May be similarly ranked, and both ranks may be multiplied, and the lowest one may be used as the ligand-determining residue position. In addition, if the sequence alignment at the obtained ligand-determined residue position is not sequence-alignable (indicated by-), but is present in 3% or more of all GPCRs, the ligand-determined residue position is determined by the ligand. It is preferable not to adopt it as a residue position.

[Another Embodiment of Step 103]

Referring to FIG. 3, another preferred embodiment of the ligand-determining residue locating step (S103) will be described. This step is for the case of two ligand-determining residue positions. First, all sequence alignments of known binding molecule proteins present in the binding molecule known protein classification information obtained in step 102 and information on the ligand or the type of ligand are input to Equation 1, and the ligand-determined residue position is entered. (Step for selecting candidate residue position for determining ligand: S301).

The number of candidate ligand-determining residue positions in the ligand-determining residue position candidate selecting step is not particularly limited, but is preferably 1 or more and 100 or less, more preferably 1 or more and 20 or less, and 2 or more It is more preferably 10 or less, particularly preferably 2 or more and 6 or less. This is because if the number of candidate ligand-determined residue positions is large, the reliability confirmation process of the candidate ligand-determined residue positions becomes difficult, and if the number of candidate candidates is one, it cannot cope with various ligand types.

For the candidate ligand-determined residue positions listed in the step of selecting candidate ligand-determined residue positions in step 301, it is more preferable to include the consideration of the reliability of the candidate for the ligand-determined residue positions using Equation 2. (Step of examining reliability of candidate ligand determination residue position: S302). In addition, when the ligand determination residue position reliability examination step in step 302 is not performed, the process may proceed to step 303 after step 301. In the step of examining the reliability of the position of the ligand-determined residue, at least the amino acid residue position including the candidate position of the ligand-determined residue is input to equation 2 with the sequence alignment of the known binding molecule protein and the ligand. Small value of function f2 (n) The residue is suitable as a residue position for determining a ligand.

The position of the ligand-determining residue may be selected from those with the smallest value of the function f 2 (n), or the product of the smallest value of f 1 (n) and ί 2 (n) is determined as the position of the ligand-determining residue. Or the one with the smallest sum of the values of fl (n) and f2 (n) may be selected as the ligand-determining residue position. Further, among the amino acid residue positions of the known binding molecule proteins present in the information on the classification of known binding molecules obtained in step 102, the amino acid residue positions are ranked from the one with the lowest f 1 (n), and further the f 2 (n n) may be ranked in the same manner, and both ranks may be multiplied to select a ligand-determining residue position from the lower one. If the sequence alignment at the obtained ligand-determining residue position cannot be sequence-aligned (indicated by-), but is present in 3% or more of the GPCRs described in the information on the classification of known binding molecules, the ligand is determined. It is preferred that residue positions are not employed as residue determining residue positions.

Next, two candidates for the ligand-determined residue positions given in step 301 or step 302 are combined, and a candidate pair for the ligand-determined residue positions is given (step of selecting a pair of ligand-determined residue positions: S303). As a candidate pair of ligand-determined residue positions, a combination consisting of all candidate ligand-determined residue positions may be used, or a preferred ligand-determined residue position of Π (n) and other residue positions may be used. A candidate pair of ligand-determined residue positions may be given by the combination of

By inputting information about candidate pairs of ligand-determined residue positions into equation 17, a pair of ligand-determined residue positions is identified (pair identification step of ligand-determined residue positions: S304). f 3 (m, n) = {(number of amino acid residue pair types) / wX + wlx (2 cross residue pair types) + w2 X (3 cross residue pair types) tens… wp-1 (p cross Number of types of residue pairs) + wA X (number of non-alignable amino acid residues) + wB X (number of non-alignable amino acid residues)} Formula 17

[In the formula 17, (m, n) is f 3 (m, n) where the binding molecule is a known protein:

Evaluation function for the mth and nth amino acid residues in the statement The number of types of amino acid residue pairs indicates the number of types of combinations of the mth and nth amino acid residues in the sequence alignment of proteins with known binding molecules. And the number of 3 crossing residue pairs means the number of 2 and 3 ligands, respectively, of the combination of the mth and nth amino acid residues in the sequence alignment of the binding molecule known protein. The number of pairs of p-crossing residue pairs is determined by the sequence alignment of known binding molecules, the ligand of which is the combination of the mth and nth amino acid residues! ) Means the number of amino acid residues that cannot be sequence-aligned.One of the m-th and n-th amino acid residues in the sequence alignment of a protein with a known binding molecule has favorable homology. The number of amino acid residue pairs that cannot be sequence-aligned in order to obtain, and the number of amino acid residue pairs that cannot be sequence-aligned means that both the mth and nth amino acid residues in the sequence alignment of the binding molecule known protein are WX is a positive constant or a distribution function with the number of amino acid pair types as variables, and the number of amino acid pair types is 400 or less in order to obtain favorable homology. The distribution function that gives the maximum value when it is a positive number means wl "'wp_l, wA, and wB are weights Is a positive number.] WX may be a positive constant, a Gaussian function, or a distribution function. In the case of a positive constant, although not particularly limited, 1 is preferable. In the case of a function, wX is a positive constant or a distribution function with the number of amino acid pair types as a variable, and means a distribution function that gives the maximum value when the number of amino acid pair types is 400 or less. For example, a Gaussian function using the number of types of amino acid pairs as a variable, a mouth-to-Lenz function, etc. The maximum value is not particularly limited as long as it is a positive number. The value of the variable that gives the maximum value of is preferably 400 or less, more preferably 20 to 200, and even more preferably 40 to 140. This value indicates that the total number of ligand-determining amino acid pairs is 20 or less. X 20 = 400, based on experience It is determined because prediction with 40 to 140 amino acid pairs gives the best results.The full width at half maximum of the distribution function is preferably 10 or more and 100 or less, and 20 or more and 7 or less. It is more preferably 0 or less, more preferably 30 or more and 50 or less. The maximum value of the distribution function is preferably 1, although it depends on the relationship with other weights. The value of the weight (wl '"wp-l, wA, wB) is not particularly limited as long as it is a positive number. However, for example, when 3 pairs of residue pairs exist, 2 intersection residues It is preferable that the value of w2 is larger than the value of wl because it is more unfavorable in predicting the ligand than the base pair, for example, when the number of ligands is three, there may be more than three crossing residue pairs. There are no combinations of weights in this case: wl is 2, w2 is 5, wA and wB are 1.

Among the candidate pairs of ligand-determining residue positions, those giving a small value of f3 (m, n) are preferable pairs of ligand-determining residue positions.

Although not explicitly stated, it is another preferred embodiment of the present invention to obtain ligand-determined residue-ligand classification information in which pairs of ligand-determined residue positions are arbitrarily combined. The combination is not particularly limited, but the combination of the function i3 (n) having the smallest value and the second and / or Xth (where X is greater than 2 and less than 100) Represents an integer.) Is combined with a smaller one. The value of the function i3 (m, η) is the Xth smallest and the yth (where y is different from X and is greater than 2 and greater than 100) It represents a small integer. For example, the value of the f3 (π) function can be arbitrarily combined with the ligand-determining residues up to the 29th. Here, the X-th and y-th Xs and y's may be input to the computer in advance, or the computer may receive input information from the user and create a combination of ligand-determined residues. The combination of the smallest and the second smallest f 3 (m, n) function limits the number of ligand-locating residues that can predict ligand and / or ligand type. Becomes However, in this way, by combining a plurality of ligand positioning residues, it becomes possible to predict the ligands of many types of ligand unknown proteins and / or the types of ligands, and furthermore, the accuracy of the prediction is high. Can also increase. It is a preferred embodiment of the present invention that such a combination of the ligand-determining residues can be automatically combined by the ligand-determining residue combining means. By using such a means, it is possible to complement the prediction derived from the residue-determined residue position. Note that X and y above Is preferably larger than 2 and smaller than 100, more preferably smaller than 50, even more preferably smaller than 30 and particularly preferably smaller than 20.

In this embodiment, for example, among the amino acid residues of a certain ligand-known protein, the amino acid residue at the ligand-determining residue position where the value of the function ί 3 (ffl, n) is the smallest is extracted. Then, among the ligand-known proteins listed in the ligand-determined residue-ligand classification information, the number of those that match the extracted ligand-determined residues is determined. Further, among these, the number of the ligands or the types of the ligands corresponding to the ligand-known protein is determined. In this specification, the value obtained in this manner is expressed as N: ((Ligand-determined residues—the number of ligand-identified proteins listed in the ligand classification information that match the extracted ligand-determined residues. ) / (Number of ligands or ligands whose type matches the ligand-known protein))). Then, N is similarly obtained for the ligand-determining residue position where the value of the function f 3 (m, ii) is the second smallest. Then, the denominator of N and the numerator at the position of the first and second smallest ligand-determining residue of the function i3 (m, n) are added. In this manner, the ligand-determined residue-ligand classification information obtained by arbitrarily combining the pairs of the ligand-determined residue positions is obtained.

Each of the steps described above may be performed manually, but is particularly preferably performed by a computer having a predetermined medium or program installed. In addition, it is particularly preferable that this computer is connected to the Internet and can access an external device.

Such a computer includes at least a sequence alignment input means for inputting information on sequence alignment of the unknown binding molecule protein, an amino acid sequence or sequence alignment of the binding molecule known protein input by the sequence alignment input means. Sequence alignment binding molecule storage means for storing information on the binding molecule or the type of binding molecule; and the amino acid sequence or sequence alignment of the binding molecule known protein or the binding molecule stored by the sequence alignment binding molecule storage means. Alternatively, the position of the binding molecule-determining residue is predicted using information on the type of the binding molecule. And a binding molecule or a type of binding molecule by associating the amino acid residue (binding molecule determining residue) at the binding molecule determining residue position with the binding molecule or the type of the binding molecule. And a binding molecule-determining residue-binding molecule classification information obtaining means for obtaining binding molecule-determining residue-binding molecule classification information indicating a correlation between the binding molecule and the type of the binding molecule.

In the binding molecule determination residue locating means, the sequence alignment of known binding molecule proteins and information on the binding molecule or the type of the binding molecule are described by the Π (η), ί2 (η), and f3 (m, n) functions described above. Any or any combination thereof is determined to determine the position of the binding molecule determining residue.

The ligand-determined residue-ligand classification information may constitute a unique database, or may be configured as a relational database based on the binding molecule known protein classification information obtained in step 102. If the ligand-determined residue-ligand classification information is configured as a relational database based on the binding molecule known protein classification information, if the sequence alignment of the unknown protein and its ligand are confirmed, It is preferable because it can be easily re-evaluated at the position of the ligand-determining residue by incorporating it into the information on the classification of known binding molecules.

With such a computer, the binding molecule or the type of the binding molecule can be easily predicted by inputting the sequence alignment of the binding molecule unknown protein into the binding molecule determination residue-binding molecule classification information.

Note that the computer of the present invention has the binding molecule determining residue-binding molecule classification information preliminarily installed, and the sequence alignment input means for inputting the protein sequence alignment is used to execute the sequence alignment of the unknown protein. It may be a computer that predicts the type of the binding molecule and / or the binding molecule by inputting the data. Even with such a computer, it is easy to predict the binding molecule or the type of the binding molecule by inputting the sequence alignment of the unknown binding protein to the binding molecule determination residue-binding molecule classification information. Becomes possible.

[Example using a virtual known binding molecule protein] Hereinafter, the steps up to the creation of the ligand-determined residue-one ligand classification information and the step of predicting the binding molecule species of the unknown binding molecule protein according to the present invention will be described using a virtual binding molecule known protein.

Table 1 is an example of hypothetical binding molecule known protein classification information. In this system, six binding proteins with known ligands and known amino acid sequences (sequence alignment) are collected. Ligands are divided into three categories: P, A, and N. When determining the position of a ligand-determining residue based on the type of ligand, P, A, and N correspond to XI, X2, and X3 in Formulas 1 and 2, respectively. When determining the position of a ligand-determining residue based on a ligand, ΔX, XX, and Δ △ correspond to XI, 11, and X3 in Formulas 1 and 2, respectively. 〇〇〇 △ X X

Table 1. Example of ΔX X X X X example of hypothetical binding molecule known protein classification table No. Alignment Ligand Ligand

TLMRK P TMMQK A TCMTK N TLLRK P TMLQK A TLLRA P In Table 1, for example, the sequence alignment of the first known binding molecule protein is TLMRK, and the binding molecule (ligand) is 〇Χ. And the type of ligand of ligand I is Ρ.

The evaluation function Π (1) will be described. The first amino acid residue in the above sequence alignment is only Τ. Therefore, in the function f l (η), Res is Τ only.

Since the amino acid residue is T and the ligand is P, there are three types.

N (Res, Xl) = N (T, X1) = 3. Similarly,

N (Res, X2) = N (T, X2) = 2, N (Res, X3) = N (T, X3) = l.

From this, Π (1) is as follows. il (l) = ∑ (N (Res, XI) XN (Res, X2) + N (Res, XI) XN (Res, X3) IN (Res, X2) x N (

Res

Res, X3)) = ∑ (N (T, XI) XN (T, X2) + N (T, XI) XN (T, X3) + N (T, X2) X N (T, X3)) = 3X2 +

3X1 + 2X1 = 11 Next, f2 (l) will be described. f2 (l) = ∑ (N (Res, XI) XN (Res, X2) XN (Res, X3))

Res

= ∑ (N (T, XI) XN (T, X2) XN (T, X3)) = 3X2X1 = 6

Becomes The following describes fl (2) and f2 (2). The second amino acid residue in the sequence alignment is L, M, and C. Therefore, in the function Π (η), Res are M and C. N (L, X1) = N (L, P) = 3 because there are three kinds of amino acid residues where the second amino acid residue is L and the ligand is P. Since the second amino acid residue is [and the ligands do not have A and N, N (L, X2) = N (L, A) = N (L, X3) = N (L, N) = 0. Since the second amino acid residue is M and the ligand is A, there are two types, so N (M, X2) = N (M, A) = 2. N (M, X1) = N (M, P) = N (M, X3) = N (M, M N) = 0.

Since the second amino acid residue is C and the ligand is N, there is one kind, so N (C, X3) = N (C, N) = 1. Since the second amino acid residue is C and none of the ligands have P and A, N (C, X1) = N (C, P) = N (C, X2) = N (C, A) = 0.

From this, Π (2) is as follows. fl (2) = ∑ (N (Res, XI) xN (Res, X2) + N (Res, XI) XN (Res, X3) + N (Res, X2) x

Res

N (Res, X3)) = ∑ (N (L, XI) XN (L, X2) IN (L, XI) XN (L, X3) + N (L, X2) XN (L, X3)) + ∑ (N (M, XI) XN (M, X2) + N (M, X1) XN (M, X3) IN (M, X2) XN (M, X3)) + ∑ (N (C, XI) XN (C, X2) + N (C, X1) XN (C, X3) + N (C, X2) XN (C, X3)) = ∑ (N (L, P ) XN (L, A) + N (L, P) XN (L, N) + N (L, A) XN (L, N)) + ∑ (N (M, P) XN (M, A) + N (M, P) XN (M, N) + N (M, A) XN (M, N))

+ ∑ (N (C, P) XN (C, A) + N (C, P) XN (C, N) + N (C, A) XN (C, N)) = 0 and f2 (2) It is as follows. f2 (2) = ∑ (N (Res, XI) XN (Res, X2) XN (Res, X3))

Res

= (N (L, P) XN (L, A) XN (L, N)) + (N (M, P) XN (M, A) XN (M, N)) + (N (C, P) XN (C, A) XN (C, N)) = 0 Similarly, il (3) = 5, fl (4) = 0, Π (5) = 8, f2 (3) = l, f2 (4) = 0, f2 (5) = 4,

並べ When the values of (n) are arranged in ascending order, n is in the order of 2, 4, 3, 5, 1. Amino acid residue positions that give a small value of Π (η) are candidates for ligand-determined residue positions. Here, the second, fourth, and third amino acid residue positions are candidates for ligand-determined residue positions. How many amino acid residue positions are candidates for ligand-determining residue positions may be determined in advance. Since the value of f2 (n) at these amino acid residue positions is smaller than the value of ί2 (η) at the first and fifth amino acid residue positions, the second, fourth and third amino acid residue positions Is a desirable candidate as a candidate for a ligand-determining residue position.

[Pair determination process for ligand-determined residue positions]

Two cross-residue pairs when there are two ligands for a pair of ligand-determined residue positions, and three cross-residue pairs when there are three ligands for a pair of ligand-determined residue positions . Combine the 2nd, 4th and 3rd amino acid residue positions, respectively, and list candidate pairs of ligand-determined residue positions. In this example, three types (2, 3), (2,4), (3, 4) Pair candidates.

First, the candidate pairs (2, 3) for the ligand-determined residue positions are examined. The combination of the second and third amino acid residues in the sequence alignment is (L, M), (M, M), (C, M), (L, L), and (M, L). is there. Therefore, the “number of kinds of amino acid residue pairs” to which the present invention belongs is 5. The corresponding ligand for each of the combinations of these five types of amino acid residues is uniquely determined, so there are no two-crossing residue pair species and three-crossing residue pair species. From this, the value of f 3 (2, 3) is 5. Similarly, the value of f3 (2, 4) is 4, and the value of ί3 (3, 4) is 5. Therefore, a pair of residue positions (2, 4) is the most preferable, and is a pair of ligand-determining residue positions.

Note that f3 (l, 5) is determined using a combination of amino acid residue positions 1 and 5 to indicate that an unfavorable combination of amino acid residue positions increases the value of f3. There are two combinations of amino acid residues at the first and fifth positions in the sequence alignment: (Τ, K) and (T, A). Therefore, the “number of kinds of amino acid residue pairs” belongs to 2. Since there are three kinds of ligands corresponding to the combination of (Τ, Κ), Ρ, Α, and Τ, (Τ, ある) is a three-crossing residue pair kind. Therefore, the value of f3 (1, 5) is 7. This value is larger than the value of f 3 (2, 4), and it can be understood that the function f 3 is suitable for determining a pair of ligand-determining residue positions.

[Ligand-determined residue-ligand classification information creation process]

From the above, it was found that the pair of the ligand-determining residue positions is the (2, 4) -th sequence alignment. The amino acid residues corresponding to these and their ligands are extracted, and ligand-determined residue-ligand classification information is created. Table 2 Ligand-determined residue-ligand classification table

(2, 4) Ligand Type of ligand

L, R 〇 X P

M, Q X X A

C, ΤΔΔN From Table 2, for example, the second and fourth sequences: If R, the ligand is predicted to be ΔX and the ligand type is predicted to be P.

[Step of predicting binding molecule species of unknown binding protein]

If the information on the classification of the ligand-determined residue-ligand as described above can be obtained, the amino acid residue at the position of the ligand-determined residue of the binding molecule unknown protein (the binding molecule with unknown ligand) is determined, and the ligand determined residue-ligand classification is performed. By applying the information, it becomes possible to predict the ligand of the unknown protein of the binding molecule. For example, a sequence alignment of a certain unknown binding molecule protein is determined by a known method. If the second and fourth amino acid residues in the sequence alignment are M and Q, respectively, the unknown binding molecule protein is determined. The type of ligand is A, and the ligand is expected to be XX (Table 2). In this way, by predicting the ligand (or type of ligand) of a protein whose binding molecule is unknown, a pair of the protein and the ligand is determined, and the physical, chemical, and biological properties of the ligand are easily determined. This can lead to an estimation of the function of the protein. Therefore, the prediction method of the present invention is also useful for developing a novel medicine.

In the above description, a ligand was described. However, the present invention can predict not only a ligand but also a molecule that binds to a known binding molecule protein. For example, when the binding molecule known protein is GPCR, a G protein that binds to the GPCR can also be predicted.

To clarify the relationship between substances that regulate complex functions in various cells and organs and their specific receptor proteins, especially G protein-coupled receptor proteins, elucidation of complex functions in various cells and organs It provides a very important tool for drug development closely related to these functions. The use of the method for predicting binding molecules of unknown binding molecules of the present invention makes it possible to predict, for example, the type of ligand and Z or ligand of GPCR. If the ligand of the GPCR and / or the type of the ligand can be predicted, the function of the GPCR in vivo can be predicted. And, for example, using information on a GPCR whose function is predicted and whose ligand and / or ligand type is predicted, It is possible to easily produce a prophylactic or therapeutic drug for a disease or the like involving the GPCR.

Considering the function of GPCRs, this method is particularly suitable for the manufacture of either or both prophylactic agents and / or therapeutic agents for central diseases, inflammatory diseases, cardiovascular diseases, cancer, metabolic diseases, immune system diseases or digestive system diseases. The invention will be used effectively. The sequence numbers in the sequence listing in the present specification indicate the following sequences.

[SEQ ID NO: 1]

2 shows the amino acid sequence of rat TGR 23-2 ligand (1-18).

[SEQ ID NO: 2]

1 shows the amino acid sequence of rat TGR23-2 ligand (1-15).

[SEQ ID NO: 3]

2 shows the amino acid sequence of rat TGR2 3-2 ligand (1-14).

[SEQ ID NO: 4]

The base sequence of the primer used in the PCR reaction in Reference Example 6 below is shown. [SEQ ID NO: 5]

The base sequence of the primer used in the PCR reaction in Reference Example 6 below is shown. [SEQ ID NO: 6]

The base sequence of the primer used in the PCR reaction in Reference Example 6 below is shown. [SEQ ID NO: 7]

1 shows the nucleotide sequence of cDNA encoding human TGR23-2 ligand precursor. [SEQ ID NO: 8]

2 shows the amino acid sequence of human TGR 23-2 ligand precursor.

[SEQ ID NO: 9]

2 shows the amino acid sequence of human TGR 23-2 ligand (1-18).

[SEQ ID NO: 10]

2 shows the amino acid sequence of human TGR23-2 ligand (1-15).

[SEQ ID NO: 11]

2 shows the amino acid sequence of human TGR 23-2 ligand (1-14). [SEQ ID NO: 1 2]

The amino acid sequence of human TGR 23-2 ligand (1-120) is shown.

[SEQ ID NO: 13]

13 shows the nucleotide sequence of a primer used in the PCR reaction in Reference Example 7 below. [SEQ ID NO: 14]

13 shows the nucleotide sequence of a primer used in the PCR reaction in Reference Example 7 below. [SEQ ID NO: 15]

13 shows the nucleotide sequence of a primer used in the PCR reaction in Reference Example 7 below. [SEQ ID NO: 16]

This shows the base sequence of cDNA encoding mouse TGR 23-2 ligand precursor [SEQ ID NO: 17]

2 shows the amino acid sequence of mouse TGR 23-2 ligand precursor.

[SEQ ID NO: 18]

Fig. 3 shows the amino acid sequence of mouse TGR 23-2 ligand (1-18).

[SEQ ID NO: 19]

Figure 3 shows the amino acid sequence of mouse TGR 23-2 ligand (1-15).

[SEQ ID NO: 20]

Figure 3 shows the amino acid sequence of mouse TGR 23-2 ligand (1-14).

[SEQ ID NO: 21]

Figure 3 shows the amino acid sequence of mouse TGR 23-2 ligand (1-20).

[SEQ ID NO: 22]

The base sequence of the primer used in the PCR reaction in Reference Example 8 below is shown. [SEQ ID NO: 23]

2 shows the nucleotide sequence of cDNA encoding a part of rat TGR23-2 ligand precursor.

[SEQ ID NO: 24]

2 shows the amino acid sequence of a part of the rat TGR 23-2 ligand precursor.

[SEQ ID NO: 25] 2 shows the amino acid sequence of rat TGR 23-2 ligand (1-20).

[SEQ ID NO: 26]

2 shows the amino acid sequence of human TGR 23-2 ligand (1-16).

[SEQ ID NO: 27]

The base sequence of the primer used in the PCR reaction in Reference Example 9 below is shown. [SEQ ID NO: 28]

The base sequence of the primer used in the PCR reaction in Reference Example 9 below is shown. [SEQ ID NO: 29]

The base sequence of the primer used in the PCR reaction in Reference Example 9 below is shown. [SEQ ID NO: 30]

Rat TGR 23— Shows the nucleotide sequence of cDNA encoding the precursor of ligand 2

[SEQ ID NO: 31]

2 shows the amino acid sequence of rat TGR 23-2 ligand precursor.

[SEQ ID NO: 32]

The base sequence of primer 11 used in the PCR reaction in Reference Example 10 below is shown.

[SEQ ID NO: 33]

The base sequence of primer 2 used in the PCR reaction in Reference Example 10 below is shown.

[SEQ ID NO: 34]

Shown below is the nucleotide sequence of a primer used to measure the expression level of TGR23-1 gene in CHO cells expressing TGR23-1 in Reference Example 11 below.

[SEQ ID NO: 35]

The following shows the nucleotide sequences of primers used to measure the expression level of TGR 23-1 gene in CHO cells expressing TGR 23-1 in Reference Example 11 below.

[SEQ ID NO: 36]

Shown below is the nucleotide sequence of a probe used to measure the expression level of TGR23-1 gene in CHO cells expressing TGR23-1 in Reference Example 11 below. 5 'end is 6—power Rupoxyfluorescein (Fam), labeled at the 3 'end with 6-carboxy-tetramethylolurodamine (Tamra).

[SEQ ID NO: 37]

1 shows the amino acid sequence of human-derived G protein-coupled receptor protein TGR 23-1 (human TGR 23-1).

[SEQ ID NO: 38]

1 shows the nucleotide sequence of cDNA encoding human-derived G protein-coupled receptor protein TGR23-1. ⁼

[SEQ ID NO: 39]

1 shows the amino acid sequence of human-derived G protein-coupled receptor protein TGR 23-2 (human TGR 23-2).

[SEQ ID NO: 40]

1 shows the nucleotide sequence of cDNA encoding human-derived G protein-coupled receptor protein TGR23-2. Example

In the following examples, GPCR is mentioned as a protein with a known binding molecule and a protein with an unknown binding molecule, but the present invention is not limited to these without departing from the gist thereof.

[Example 1]

(Prediction of ligand type of GPCR)

Information on 1152 types of GPCR sequence alignments and ligands is obtained from the GP C RDB (database) in which GPCR sequence alignments and ligands are registered. 3) was obtained. After that, the obtained ligands were divided into three types. The three types are as follows.

P …… Peptides (peptide), C emokines (chemokine), Glycoproteins (glycoprotein) A ······· Monoamines, (adrenaline, acetylcholine, dopamine, serotonin, histamine)

N …… Lipids

Information on the sequence alignment and ligands of 1152 GPCRs was input to a computer. The information on the sequence alignment and ligands of the 1152 types of GPCRs that were input was used to determine the six types of ligands that were selected by the means for selecting candidate positions for determining residues using the Π (n) function and f 2 (n) function. Table 3 shows candidate base positions.

In the ligand-determining residue position candidate selecting means, a candidate having a small product of the value of the f l (n) function and the value of the f2 (n) function is selected as the ligand residue position candidate. Among them, if any of them could not be sequence aligned (-), 3% or more of them were excluded from candidate ligand-determined residue positions. In this manner, 20 ligand residue position candidates were selected. Table 3 shows the evaluation values and ranks of the function Π (η) and the function f2 (n) for the more preferable six candidate ligand residue positions among them. Table 3 Evaluation results and ranks of function i l (n) and function f 2 (n) for six preferred candidate ligand residue positions f l (n) f2 (n)

Residue position Rank Evaluation value Rank Evaluation value

82 1 9222 10 51554

86 5 13 317 3 22593

90 2 10 112 1 21 168

91 13 15016 12 62274

209 18 16 389 21 81 468

230 10 14618 25 93622 According to Table 3, for example, for residue number 86 in the sequence alignment, the function Π (η) ranks fifth and the function f2 (n) ranks third. It can be seen that it is. Combine the above 6 candidate ligand-determining residue positions to obtain 360 ligand-determining residue positions He gave a pair of candidates. The information on the candidate pairs of 360 ligand-determined residue positions was substituted into Equation 3 to determine the pair of ligand-determined residue positions. As a result, the f3 (n) value is the smallest, and the most preferable ligand-determining residue position (pair) is (86, 90), that is, the 86th and 90th positions in the GPCR sequence alignment. This was the position of the amino acid residue. In addition, using as an index whether or not the prediction derived from (86, 90) can be complemented, (209, 211) and (86, 236) were selected as the next preferred ligand-determining residue positions.

The 86th and 90th amino acid residue types and the number of ligand types were extracted from the sequence alignment of GPCR from the binding molecule known protein classification information to obtain ligand-determined residue-ligand classification information. Table 4 shows the excerpts. From Table 4, for example, of the 1152 types of GPCRs, there are 86 types in which the amino acids at amino acid residue positions 86 and 90 are A and G, respectively, and their ligands are all N (lipid). It can be seen that it is classified as Thus, it can be seen that most GPCRs can predict their ligands by the amino acids at the 86th and 90th amino acid positions. Table 4 Ligand-determined residue-ligand classification table (Binding molecule-determined residue-binding molecule classification table)

)

Amino acid residue Type of ligand Total

86 90 A N P

A G 0 86 0 86

D C 114 0 0 114

D M 26 0 0 26

D Q 13 0 0 13

D S 68 0 0 68

D V 26 0 0 26

F L 0 20 6 26

F M 0 2 13 15

G G 0 58 0 58

I L 0 4 23 27

I M 0 0 32 32

I V 0 1 7 8

K F 0 0 11 11

K M 0 0 10 10

LM 0 0 9 9 MG 0 26 0 26

M V 11 0 13 24

P M 0 .0 7 7

p V 0 0 8 8

Q I 0 0 15 15

Q M 0 0 25 25

Q V 0 0 48 48

T S 0 0 26 26

v F o 9 o 9

V G 0 30 0 30

V L 0 0 19 19

V T 17 0 0 17

Y F 0 0 42 42

YL 0 28 28 From Table 4, it can be seen that, for example, in the sequence alignment of GPCRs, the 86th and 90th amino acid residues are A and G, respectively.There are 86 types of GPCRs, and the ligands for them are all N It is understood that the ligand is classified as (lipid). Next, GPCRs in which ligands were recently discovered were randomly selected, and the accuracy of the above-mentioned ligand-determined residue-ligand classification information (ie, the accuracy of the present ligand prediction method) was analyzed. Table 5 shows the results. Table 5

G P C R Amino acid residue N Amino acid residue N Amino acid residue N

86 90 209 211 86 236

GPR2 I M 32/32 V F 4/26 I N 61/66

GPR5 A L 1/2 Q L 6/7 A H 0/0

GPR8 D I 1/1 A T 1/3 D N 40/234

GPR10 Q V 48/48 R L 3/20 Q S 34/34

GPR13 F F 4/4 E L 10/10 F H 5/9

GPR14 D M 26/26 A Y 0/0 D N 40/234

GP 16 C M 1/3 E S 0/0 C S 0/4

GPR24 D G 13/13 Q S 1/1 D N 40/234

GP 28 Y F 42/42 Q I 1/1 Y H 71/71

GPR29 Y L 28/28 S F 12/19 Y H 71/71

GPR54 Q V 48/48 Q L 5/6 Q N 52/52

GPR74 Q V 48/48 S Y 2/2 Q N 52/52

APJ I M 32/32 Y L 2/7 I N 61/66

Contract QV 48/48 IY 0/0 QH 3/3 Table 5 is explained. For example, in the sequence alignment of GPR2, the (86, 90) th amino acid residue (position of the residue determining the binding molecule) has a sequence of I and M, respectively. Among the binding molecule known GPC R1152 types in the binding molecule classification residue-binding molecule classification information, there are 32 types in which the (86, 90) th amino acid residues are I and M, respectively. All 32 types of ligands are peptides. Although this GPR 2 is not included in the 1152 known binding molecules of GPCR, the actual ligand type is a peptide, which is consistent with the predicted ligand type.

In addition, in the sequence alignment of GPR2, the amino acid residue at the position (209, 211) determined by the binding molecule is V and F, respectively. Of the above 1 1 52 types, there are 26 types in which the sequence alignment of 209 and 2 1 1 is V and F, and there are 4 types of those with a ligand type of peptide in 4 types. is there.

Table 5 shows that the binding molecule-determining residue positions (86, 90) are the most preferable binding molecule-determining residue positions among the three types. Furthermore, according to Table 5, it can be seen that the type of the ligand can be predicted with high accuracy.

Next, we analyzed the accuracy of the ligand-determined residue-ligand classification information combining the binding molecule-determined residue positions (86, 90) and (209, 211).

For example, from Table 5, the positions of the binding molecule determining residues of GPR2 (86, 90), and N (1112 types of GPCRs) at (209, 211), the binding molecule determining residues were identical to GPR2. And the number of ligands with the same type of ligand) are 3 2/3 2 and 4/26, respectively. Add these up to 36/58.

In this way, for all GPCRs listed in Table 5, the binding molecule-determining residue positions of GPR2 (86, 90) and N of (209, 211) were added. As a result, the number of evaluation targets increases, and the ligand type can be predicted with higher accuracy than when the ligand type is predicted solely at the binding molecule-determined residue position (86, 90) or (209, 211) alone. confirmed.

[Example 2] (Method of predicting the type of binding G a protein that binds to GPCR)

First, the binding Gα protein, which is a binding molecule of GPCR, was classified into three types, G i, Gq, and G s. This is a categorization of the binding G α protein into the 2000 Receptor & Ion channel Nomenclature Supplement of Trends in pharmacological sciences (TIPS). For simplicity, Gi / o is Gi and GQ / 11 is GQ in the 2000 Receptor & Ion channel Nomenclature Supplement of TIPS. In addition, of the binding Go; proteins listed in the 2000 Receptor & Ion channel Nomenclature Supplement of TIPS, those binding to two or more G proteins, and Gi / al, 3, and Gi / a2, 3 were excluded as exceptions. In this way, information on about 600 types of GPCR and its sequence alignment, and the binding Gα protein that binds to it, and its type were obtained.

The sequence alignment of GPCR and information on the type of binding Ga protein that binds to the G protein were input to a computer. The inputted sequence information of the GPCR and the information on the type of the binding Gcu protein were selected by means of a candidate selecting residue position for determining a binding molecule using the f1 (n) function and the f2 (n) function. As a result, two pairs of binding residue positions (177, 178) and (82, 230) were determined.

In the binding molecule determination residue position candidate selecting means, a candidate having a small product of the value of the Π (η) function and the value of the ί2 (η) function is selected as the candidate binding molecule residue position. Among them, if any of them could not be sequence-aligned (-), 3% or more of them were excluded from candidate ligand-determining residue positions. In this way, candidate ligand residue positions were selected.

Next, the accuracy of predicting the types of binding molecules of the two pairs of binding molecule-determining residue positions (177, 178) and (82, 230) was confirmed.

The type of Ga protein that binds to multiple GPCRs was obtained from the literature. Then, the binding Go; how the protein was obtained was determined by Ca influx, Arachidonic acid release: by arachidonic acid release, by PTX (pertussis toxin sensitive), and by cyclic adenosine. Phosphoric acid was used, and Ca, AA, PTX, and cAMP were used, respectively. In addition, GQ is determined only by Ca Were excluded because the bound Ga protein could be something else. In this way, a GPCR was selected.

The type of GPCR selected and the type of binding GPCR obtained from the literature, the method by which the type was obtained, the sequence alignment of the pair of binding molecule-determining residue positions (177, 178), and (82, 230); Table 6 shows the Go; protein types predicted from each pair. Table 6

GPCR Amino acid residue N Amino acid residue N Experiment Forecast

177 178 83 230

GPR5 R S 9/9 N R 0/0 Gi (PTX) Gi

GPR8 L G 4/4 L T 0/0 Gi (cAMP) GQ

GPR13 K N 3/3 T E 32/32 Gi (PTX) Gi

GPR14 A R 1/2 F T 1/1 Ga (AA) Ga

GPR16 F 3/3 H I 3/3 G (Ca) GQ

GPR24 I R 1/1 I I 9/9 Gi (cAMP) Gi

GPR39 S R 0/0 T E 32/32 GQ (AA) Gi

APJ G L 2/2 ST 0/0 Gi (cAMP) Gi For example, GPR5 in Table 6 will be described. The type of binding G a protein of GPR5 is Gi, which was obtained by PTX. And the 177th and 178th sequence alignments of GPR5 are R and S. Nine GPCRs have such a sequence alignment, and it can be seen that the type of conjugated Gcu protein is Gi. Among the above eight GPCRs, GPR5, GPR13, GPR14, GPR16, GPR24, and AP; [6] are predicting the binding Go; protein. From the above, it can be seen that according to the present invention, it is possible to predict the binding Ga protein with high accuracy.

[Example 3]

(Prediction of TGR23 ligand)

It became clear that the type of ligand can be predicted with high accuracy by evaluating the combination of two SDM pairs. Therefore, using the method of the present invention, Sequence alignment of the amino acid sequence information of TGR23, which is the G protein-coupled receptor protein represented by SEQ ID NO: 37 and SEQ ID NO: 39, was performed, and the TGR23 ligand was predicted from the results. Here, FIG. 4 shows the results of sequence alignment of TGR2311 and ゥ silodopsin.

As shown in FIG. 4, the amino acids at residues (86, 90), (209, 211) and (86, 236) in TGR 23-1 are (Q, L), (D , F) and (Q, N). These N (the number of 1152 types of GPCRs in which the binding molecule-determining residue coincided with TGR 23 and the type of ligand were the same) were 0/0 and 13/27, respectively. And 52/52, and the combined evaluation of (86, 90) and (86, 236) SDM pairs with a wide range of estimated GPCR numbers estimated that the ligand type was peptide. The following reference examples show that the ligand of TGR23 (TGR23_1 and TGR23-2) is actually a peptide. [Reference Example 1] ·

(Purification from rat whole brain extract of an active substance that specifically exhibits cAMP production promotion activity on CHO cells expressing TGR 23-12)

A substance exhibiting a ligand activity specific to TGR 23-2 was purified from rat whole brain using cGR production promoting activity on CGR cells expressing TGR 23-2 as an index.

High performance liquid chromatography (HPLC) fraction of rat whole brain extract was prepared by the method described below. Immediately after sequentially extracting 400 g (200 cats) of the whole brain of an 8-week-old Wistar rat purchased from Charles River Japan Co., Ltd., it was thrown into distilled water (300 ml) boiled in 25 pets and boiled for 10 min. did. Immediately after boiling, cool on ice, combine 200 heads (2.4 L), add 180 ml of acetic acid to a final concentration of 1.0 M, and use Polytron (10,000 rpm, 2 minutes) at low temperature. Crushed. The crushed liquid is centrifuged (8,000 rpm, 30 minutes), and the supernatant is collected. To the precipitate, 2.4 L of 1.0 M acetic acid is added, crushed again with a polytron, stirred overnight, and then centrifuged ( (8,000 rpm, 30 minutes) to obtain a supernatant. The supernatant obtained from each centrifugation is 2x After slowly adding dropwise (4.8 L) of cold acetone at 4 ° C, the supernatant obtained by the first centrifugation is stirred overnight, and the supernatant obtained by the second centrifugation is added by 4 Stirred for hours. The extract to which acetone was added was centrifuged (8,000 rpm, 30 minutes) to remove the precipitate, and the resulting supernatant was subjected to evaporation under reduced pressure to evaporate acetone. An equal amount of getyl ether was added to the extract from which acetone was distilled off, and the lipid-containing ether layer was separated using a separatory funnel, and the aqueous layer was collected. The extract degreased with ether was concentrated under reduced pressure at an evaporator to remove ether completely. The concentrated solution was filtered through a glass fiber filter paper (Advantech, DP 70 (9Οπιιη)), and the filtrate was packed in a glass column (30 X 240 mm) using an ODS column (Daisogel, Daisogel IR-120-0DS-A63). / 210 um). The column was washed with 400 ml of 1.0 M acetic acid and eluted with 500 ml of 60% acetonitrile containing 0.1% trifluoroacetic acid. The eluate was concentrated under reduced pressure to remove the solvent, and the concentrate was freeze-dried. 1.2 g of the obtained white powder was dissolved in 30 ml of 10% acetonitrile containing 0.1% trifluoroacetic acid, and 12.5 ml of each powder was dissolved in an ODS column (Tosoichi, TSKgel ODS-80TS (2. The sample was subjected to preparative HP LC using a gradient elution method of acetonitrile containing 10% to 60% of 0.1% trifluoroacetic acid using 5 φX 300 mm)). The HP LC was performed twice, and the eluate was divided into 60 fractions every 2 minutes, and the two eluates were combined. Each fraction was concentrated and dried under reduced pressure, and 0.4 ml of dimethyl sulfoxide (DMSO) was added to the residue, and then completely dissolved using a Portex mixer and an ultrasonic washer.

The DMSO solution of the HPLC fraction obtained above was administered to CHO cells expressing TGR23-2 according to the method described in Reference Example 3, and the amount of intracellular cAMP production was measured. And 22 to 23 showed remarkable cAMP production promoting activity. In addition, a similar sample was examined for arachidonic acid metabolite releasing activity according to a known method. As a result, remarkable activity was confirmed.

These activities were not observed in other receptor-expressing cells, indicating the presence of a ligand active substance specific for TGR23-2 in rat whole brain extracts. The three active fractions obtained were further purified by the following methods (a) to (c), respectively. In addition, for all active fractions, the first positive In the fractions showing the cAMP production promoting activity obtained in the purification process using the on-exchange column, the receptor-specific intracellular calcium release activity was also confirmed by FLIPR (Molecular Devices). Was done. Therefore, in confirming the activity in the subsequent purification steps, intracellular calcium release activity by FLIPR was used as an indicator, and it was appropriately confirmed that the fraction showing the activity exhibited cAMP production promoting activity.

(a) Fraction number 1 8

For fraction No. 18, dissolve in 10 ml of 1 OmM ammonium formate containing 10% acetonitrile, and use a cation exchange column (Tosoichi, TSKgel SP-5PW (20 mm φ X 15 Omm)) After elution, elution was carried out with a concentration gradient of 10 mM to 1.0 M ammonium formate containing 10% acetonitrile. The activity was recovered at around 0.4M ammonium formate. After freeze-drying the active fraction, dissolve it in 0.8 ml of 10% acetonitrile containing 0.1% trifluoroacetic acid, and use an ODS column (Tosoichi, TSKgel 0DS-80TS (4.6 φ X 25 Omm)) ) And eluted with a concentration gradient of 10% to 25% acetonitrile containing 0.1% trifluoroacetic acid. As a result, activity was observed around 13% of acetonitrile. The obtained active fraction was lyophilized, dissolved in 0.1 lm of DMSO, further added with 0.7 ml of 0.1% acetofluorobutyric acid in 10% acetonitrile, and added to an ODS column (Wako Pure Chemical Industries, Ltd.). Wakosii-II 3C18H G (2.Ο ΦΦ X 15 Omm)) and eluted with a concentration gradient of 10% to 37.5% acetonitrile containing 0.1% heptafluorobutyric acid. It was manually collected every time. The activity was observed around 26% of acetonitrile. The active fraction was further added with 0.7 ml of 0.1% trifluoroacetic acid containing 0.1% acetonitrile, applied to a QDS column (Wako Pure Chemical Industries, Wakosite II 3C18HG), and then purified with 0.1% trifluoroacetic acid. Elution was carried out with a gradient of 10% to 20% acetonitrile containing acetic acid, and the eluate was manually collected for each peak. The activity was obtained as a single peak around 11% of acetonitrile. The structure of the active substance contained in this fraction was determined as shown in Reference Example 5 below.

(b) Fraction number 20

For fraction number 20, 1 OmM ammonium formate containing 10% acetonitrile After dissolving in 10 ml of water and applying it to a cation exchange column (Tosoichi, TSKgel SP-5PW (20 mm x 150 mm)), a concentration gradient from 1 OmM to 1.0 M ammonium formate containing 10% acetonitrile was obtained. Eluted. The activity was recovered at around 0.6 M ammonium formate. After freeze-drying the active fraction, dissolve in 0.8 ml of 10% acetonitrile containing 0.1% trifluoroacetic acid, and use a CN column (Nomura Chemical, Develo sil CN-UG-5 (4.6 mm φ X 250 mm)) After elution with elution with a concentration gradient of 10% to 25% acetonitrile containing 0.1% trifluoroacetic acid, activity was observed around 12% of acetonitrile. The resulting active fraction was lyophilized, dissolved in 0.1 lm of DMS 0, and further added with 0.7 ml of 10% acetonitrile containing 0.1% trifluoroacetic acid, followed by ODS column (Wako Pure Chemical Industries, Wakosil- II 3C18HG (2.ΟπΐΓηφ X 150mm)) and eluted with a concentration gradient of 10% to 20% acetonitrile containing 0.1% trifluoroacetic acid, and the eluate was manually separated for each peak . The activity was obtained as a single peak around 15% of acetonitrile. The structure of the active substance contained in this fraction was determined as shown in Reference Example ^ = 3 below.

(c) Fraction number 22-23

Fraction Nos. 22-23 were dissolved in 10 ml of 1 OmM ammonium formate containing 10% acetonitrile and applied to a cation exchange column (TOSOKI, TSKgel SP-5PW (2ππηηι X 150 mm)). Elution was carried out with a concentration gradient of 1.0 M ammonium formate, 1 OmM force containing 10% acetonitrile. The activity was recovered at around 0.4M ammonium formate. After freeze-drying the active fraction, dissolve it in 0.8 ml of 10% acetonitrile containing 0.1% trifluoroacetic acid and attach to a CN column (Nomura Chemical, Develosil CN-UG-5 (4. βπιπι X 250 mm)). After that, elution was performed with a concentration gradient of 10% to 25% acetonitrile containing 0.1% trifluoroacetic acid, and as a result, activity was observed around 13% of acetonitrile. The resulting active fraction was freeze-dried, dissolved in 0.1 ml of DMS O, further added with 0.7 ml of 10% acetonitrile containing 0.1% trifluoroacetic acid, and added to an ODS column (Wako Pure Chemical Industries, Ltd.). The drug, Wako sil-II 3C18HG (2.0 mm φ X 150 mm)), was eluted with a concentration gradient of 10% to 20% acetonitrile containing 0.1% trifluoroacetic acid. The sample was manually collected for each work. Activity was observed around 16% of acetonitrile. To the active fraction, 0.7 ml of 10% acetonitrile containing 0.1% heptanofluorbutyric acid was further added, and the mixture was applied to an ODS column (Wako Pure Chemical Industries, Wakosil-II 3C18HG). Elution was performed with a gradient of 10% to 37.5% acetonitrile containing heptafluorobutyric acid, and the eluate was manually collected for each peak. The activity was obtained as a single peak around 28% of acetonitrile. The structure of the active substance contained in this fraction was determined as shown in Reference Example 4 below.

[Reference Example 2]

(Inactivation by pronase of an active substance that specifically promotes intracellular cAMP production in CHO cells expressing TGR23-12 in rat whole brain extract)

The HPLC fractions 18, 20, 22 and 23, which showed intracellular cAMP production promoting activity on CHO cells expressing TGR23-2 in Reference Example 1, were converted to pronase (pronase). Sigma, protease Type XIV (P5147)) was used to determine whether the active substance was proteinaceous.

Rat whole brain extract HP LC active fraction (fraction No. 18, 20, and 22 to 23) Add 4-1 each to 0.2 M ammonium acetate 1001, and add 3 ig pronase to this After incubation at 37 ° C for 2 hours, the added pronase was inactivated by heating in boiling water for 10 minutes. Distilled water (lm1) containing BSAO. 05mg and CHAPS 0.05mg was added thereto and freeze-dried. The freeze-dried sample was added to CHO cells expressing TGR23-2 according to a known method, and the activity of promoting intracellular cAMP production was measured.

As a result, the activity of each fraction completely disappeared by pronase treatment. Therefore, it was clarified that any of the active substances exhibiting an intracellular cAMP production promoting activity on CHO cells expressing TGR23-2 in a rat whole brain extract is a protein or a peptide.

[Reference Example 3]

(CHO cells expressing TGR 23-12 obtained from fraction number 20 of rat whole brain extract) Determination of the amino acid sequence of an active substance that specifically exhibits cAMP production promoting activity against TGR23-12 TGR23-expressing CHO cells contained in the three fractions of rat whole brain extract as shown in Reference Example 2. Since it was expected that any of the active substances that specifically show cAMP production promoting activity against, would be proteinaceous, amino acid sequence analysis was performed on each of them as follows.

Amino acid sequence of an active substance that specifically exhibits cAMP production promoting activity on TGR23_2-expressing CHO cells obtained from fraction number 20 of rat whole brain extract as shown in Reference Example 1. Analysis and mass spectrometry were performed. Amino-terminal amino acid sequence analysis using Procise 491cLC protein sequencer (Applied Biosystems) using the eluate containing the activity peak revealed that S FRNGVGS GVKKTS FRRA (SEQ ID NO: 1) from the N-terminal to 18 residues. The amino acid sequence of was obtained. The same eluate was used to perform mass spectrometry using a Thermo Fiimigan LCQ ion trap mass spectrometer (ThermoQuest) equipped with a nanospray ion source (Pro evening). The result was calculated from the amino acid sequence of SEQ ID NO: 1. (Measured value: 1954.9, calculated value: 1954.2).

From this, the active substance which specifically exhibits cAMP production promoting activity on TGR2-3-expressing CHO cells obtained from fraction number 20 of rat whole brain extract has the amino acid sequence shown in SEQ ID NO: 1. It was determined to have. [Reference Example 4]

(Determination of the amino acid sequence of an active substance that specifically exhibits cAMP production promoting activity on TGR23-2 expressing CH0 cells obtained from fraction numbers 22 to 23 of rat whole brain extract)

Amino acid sequence of an active substance that specifically exhibits cAMP production promoting activity on TGR23-2 expressing CHO cells obtained from fraction numbers 22 to 23 of rat whole brain extract as shown in Reference Example 1. Analysis and mass spectrometry were performed. Amino-terminal amino acid sequence analysis using Procise 491cLC protein sequencer (Applied Biosystems) using the eluate containing the activity peak revealed that S FRNGVGS GVKKTS F (SEQ ID NO: 2) from the N-terminal to 15 residues. Amino acid sequence is obtained Was. Using the same eluate, mass spectrometry using a Tiiermo Finnigan LCQ ion trap mass spectrometer (ThermoQuest) equipped with a nanospray ion source (Pro evening) was calculated from the amino acid sequence of SEQ ID NO: 2. (Measured value: 1570.8, calculated value: 1570.8).

Thus, the active substance that specifically exhibits cAMP production promoting activity on TGR23-2 expressing CHO cells obtained from fraction numbers 22 to 23 of rat whole brain extract is the amino acid shown in SEQ ID NO: 2. It was determined to have a sequence.

[Reference Example 5]

(Determination of the amino acid sequence of an active substance that specifically exhibits cAMP production promotion activity on TGR23-2 expressing CHO cells obtained from fraction number 18 of rat whole brain extract) As shown in Reference Example 1. Amino acid sequence analysis and mass spectrometry of an active substance that specifically exhibits cAMP production promoting activity on TGR23-12-expressing CHO cells obtained from fraction number 18 of rat whole brain extract were performed. Amino-terminal amino acid sequence analysis using Procise 491cLC protein sequencer (Applied Biosystems) using the eluate containing the activity peak revealed that the amino acid sequence of S FRNGVGSGVKKTS (SEQ ID NO: 3) was extended from the N-terminal to 14 residues. was gotten. The same eluate was used to perform mass spectrometry using a Thermo Finnigan LCQ ion trap mass spectrometer (SammoQuest) equipped with a nanospray ion source (protana). The mass was calculated from the amino acid sequence of SEQ ID NO: 3. The following mass values were obtained (actual value: 144.1, calculated value: 1423.6).

Thus, the active substance that specifically exhibits cAMP production promoting activity on TGR23-2 expressing CHO cells obtained from fractional number 18 of rat whole brain extract

: It was determined to have the amino acid sequence shown in 3:

[Reference Example 6]

(Cloning of cDNA encoding human TGR 23-2 ligand precursor) An active peptide that specifically exhibits cAMP production promoting activity on TGR23-12-expressing CHO cells obtained from rat whole brain extract In the description, rat TGR23— To clone the cDNA encoding the precursor of the human homolog (sometimes referred to as human TGR23-2 ligand in this specification) of the human hypolog, A PCR was performed using the cDNA as a type II. Using the following synthetic DNA primers, cDNA derived from the human hypothalamus was converted to type III and amplified by the PCR method. The composition of the reaction solution was human hypothalamus Marathon-Ready cDNA (CLONTECH) 0.8 \, SEQ ID NO: 4 and SEQ ID NO: 5, each of the synthetic DNA primers 1.0 M, 0.2 mM dNTPs, ET aq (Takara Shuzo) 0.1 II 1 and the ExTaq buffer attached to the enzyme, and the total reaction volume was 201. The amplification cycle was performed using a thermal cycler (PE Biosystems) at 94 ° C for 300 seconds, followed by 94 ° C for 10 seconds, 55 ° C for 30 seconds, 72 ° C for 30 seconds. A cycle of 35 seconds was repeated 35 times, and finally, the mixture was kept at 72 ° C for 5 minutes. Next, a PCR reaction solution 2 n 1, which was diluted 50-fold with DNase and RNase Free distilled water, synthetic DNA primers of SEQ ID NO: 4 and SEQ ID NO: 6, 1.0 M and 0.2 mM dNTPs, respectively ExT aq polymerase (Takara Shuzo) 0.11 and ExTaq buffer attached to the enzyme to make the total reaction volume 201, and heat it for 94 · 300 seconds using a thermocycler (PE Biosystems). Thereafter, a cycle of 94 ° C · 10 seconds, 55 ° C · 30 seconds, and 72 ° C · 30 seconds was repeated 35 times, and finally, the temperature was kept at 72 ° C for 5 minutes. After the amplified DNA was separated by 2.0% agarose gel electrophoresis, the band was cut out with a force razor, and the DNA was recovered using a QIAduick Gel Extraction Kit (Qiagen). This DNA was cloned into the pGEM-T Easy Vector 1 according to the protocol of the pGEM-T Easy Vector System (Promega). This was introduced into Escherichia coli JM109 competent cell (Takara Shuzo), transformed, and clones containing the cDNA fragment were selected on an LB agar medium containing ampicillin and X-ga1, and the color was white. Only clones exhibiting the above were separated using a sterilized toothpick to obtain a transformant. Individual clones were cultured overnight in LB medium containing ampicillin, and plasmid DNA was prepared using QIAwell 8 Plasmid Kit (Qiagen). The reaction for determining the nucleotide sequence was carried out using a BigDye Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems), and was decoded using a fluorescent automatic sequencer to obtain the DNA sequence shown in SEQ ID NO: 7. The nucleotide sequence of the DNA represented by SEQ ID NO: 7 includes the amino acid sequence of rat TGR23-2 ligand obtained from whole rat brain represented by SEQ ID NO: 1, SEQ ID NO: 2 and SEQ ID NO: 3. The presence of a frame encoding an amino acid sequence very similar to that of was predicted to be cDNA encoding the precursor of human TGR23-2 ligand or a part thereof.

An ATG, which is predicted to be the initiation codon for protein translation, is located 5 'upstream of the amino acid sequence translated from SEQ ID NO: 7 in a frame encoding an amino acid sequence considered to be a human TGR 23-2 ligand. However, when a hydrophobic input was performed, a highly hydrophobic region presumed to be the signal sequence appeared only when translated from the ATG 5 'upstream, and this ATG was the start codon. It was estimated that On the 3 'side, a stop codon was present downstream of the sequence thought to encode the human TGR 23-2 ligand. The amino acid sequence of the human TGR 23-2 ligand precursor deduced as described above is shown in SEQ ID NO: 8. In this sequence, at the N-terminal side of the amino acid sequence considered to correspond to human TGR 23-2 ligand, a Lys-Arg sequence (Seidah, NG et al., Ann. NY Acad. Sci., 839, 9-24, 1998). On the other hand, a termination codon was present at the C-terminal side, but two more residues were present between the terminal and the sequence corresponding to the rat TGR23-2 ligand having the amino acid sequence represented by SEQ ID NO: 1.

Thus, the amino acid sequence of human TGR23-2 ligand is the amino acid sequence of rat TGR23-2 ligand obtained from rat whole brain extract; SEQ ID NO: 1 [rat TGR23-2 ligand (1-18) )], SEQ ID NO: 2 [rat TGR23-2 ligand (1-15)] and SEQ ID NO: 3 [rat TGR23-2 ligand (1-114)], SEQ ID NO: 9 [ Human TGR23_2 ligand (111)], SEQ ID NO: 10 [human TGR23-2 ligand (1-15)] and SEQ ID NO: 11 [human TGR23-2 ligand ( 11-14)] and an amino acid sequence represented by SEQ ID NO: 9 with two residues extended to the C-terminal side of SEQ ID NO: 9 [SEQ ID NO: 12] [amino acid sequence represented by human TGR 23-2 ligand (1 20)]. Furthermore, the sequence of human TGR 23-2 ligand is TGR 23-2 ligand and rat Unlike the sequence of TGR 23-2 ligand, the sequence has a G 1 nA rg sequence instead of an Arg-A rg sequence, so the 16 residues shown in SEQ ID NO: 26 The amino acid sequence [human TGR 23-2 ligand (1-16)] was also deduced to be the ligand sequence.

[Reference Example 7]

(Cloning of cDNA encoding mouse TGR23-2 ligand precursor) Mouse homologue of rat TGR23-2 ligand obtained from rat whole brain extract (referred to as mouse TGR23-2 ligand in this specification) In order to clone the cDNA encoding the precursor of the mouse, PCR was performed using type I cDNA from the whole mouse brain.

Amplification by PCR was performed using the following synthetic DNA primers and cDNA of mouse whole brain as type II. The composition of the reaction solution was as follows: Mouse whole brain Marathon-Ready cDNA (CLONTECH) 0.8 Synthetic DNA primers of SEQ ID NO: 13 and SEQ ID NO: 14: 1.0 0, 0.2 mM dNTPs, ExT a α (Takara Shuzo) 0.1 H1 and ExTaQ buffer attached to the enzyme, and the total reaction volume was 201. The amplification cycle was performed using a thermal cycler (PE Biosystems) at 94 ° C for 5 minutes, followed by a cycle of 94 ° C for 10 seconds, 65 ° C for 30 seconds, and 72 and 30 seconds. The test was repeated 5 times, and finally kept at 72 ° C for 5 minutes. Next, the PCR reaction solution 21, which was diluted 100 times with DNase and RNase Free distilled water, the synthetic DNA primers of SEQ ID NO: 13 and SEQ ID NO: 15 were each 1.0 M and 0.2 mM dN. TP s, ExTaQ polymerase (Takara Shuzo) 0.1 1 and EXT aq buffer attached to the enzyme, the total reaction volume was 201, and the temperature was 94 ° C After heating for 5 minutes, a cycle of 94 ° C for 10 seconds, 60 hours, 30 seconds, and 72 ° C for 30 seconds was repeated 30 times, and finally, the temperature was kept at 72 ° C for 5 minutes. After separating the amplified DNA by 2.0% agarose gel electrophoresis, about 440 base-length DNA was cut out with a force razor, and the DNA was recovered using a QIAQuick Gel Extraction Kit (Qiagen). This DNA was cloned into the pGEM-T Easy Vector 1 according to the protocol of the pGEM-T Easy Vector System (Porta Mega). this Was transformed into Escherichia coli JM109 competent cell (Takara Shuzo), and clones having cDNA inserts were selected on LB agar medium containing ampicillin and X_ga1, and only white clones were sterilized. Separation was performed using a toothpick to obtain a transformant. Individual clones were cultured in LB medium containing ampicillin, and plasmid DNA was prepared using QIAwell 8 Plasmid Kit (Qiagen). The reaction for determining the nucleotide sequence was performed using a BigDye Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems), and the DNA sequence was decoded using a fluorescent automatic sequencer to obtain a DNA sequence represented by SEQ ID NO: 16.

The nucleotide sequence of DNA represented by SEQ ID NO: 16 includes the amino acid sequence of rat TGR23_2 ligand obtained from whole rat brain represented by SEQ ID NO: 1, SEQ ID NO: 2 and SEQ ID NO: 3. Since a frame encoding a very similar amino acid sequence was present, it was presumed to be a cDNA encoding the precursor of mouse TGR 23-2 ligand or a part thereof.

ATG predicted to be the initiation codon for protein translation 5 'upstream of the amino acid sequence translated from SEQ ID NO: 16 in a frame that encodes the amino acid sequence considered to be mouse TGR23-2 ligand However, when a hydrophobic plot was performed, a highly hydrophobic region presumed to be a signal sequence appeared only when translated from the 5 'upstream ATG. Was estimated. A stop codon appeared in the same frame 5 more upstream of this ATG codon. On the 3rd side, a stop codon was present downstream of the sequence thought to encode the mouse TGR 23-2 ligand. SEQ ID NO: 17 shows the amino acid sequence of the mouse TGR23-2 ligand precursor deduced as described above. In this sequence, the N-terminal of the amino acid sequence considered to correspond to the mouse TGR 23-2 ligand has a Lys-Arg sequence (Seidah, NG), which is usually considered to be a bioactive peptide cleaved from its precursor protein. et al., Ann. NY Acad. Sci., 839, 9-24, 1998). On the other hand, a termination codon was present at the C-terminal side, but two more residues were present between the sequence corresponding to the rat TGR23_2 ligand of SEQ ID NO: 1.

From this, the amino acid sequence of mouse TGR23-2 ligand was extracted from rat whole brain Amino acid sequence of rat TGR23-2 ligand obtained from the product; SEQ ID NO: 1 [rat TGR23-2 ligand (1-18)], SEQ ID NO: 2 [rat TGR23-2 ligand (1-15) ] And SEQ ID NO: 3 [rat TGR 23-2 ligand (1-114)], corresponding to each, SEQ ID NO: 18 [mouse TGR 23-2 ligand (1-1 8)], SEQ ID NO: 19 [Mouse TGR 23-2 ligand (1-15)] and the amino acid sequence represented by SEQ ID NO: 20 [mouse TGR 23-2 ligand (1-14)], and further, at the C-terminal side of SEQ ID NO: 18 It was presumed to be the amino acid sequence represented by SEQ ID NO: 21 extended by 2 residues [mouse TGR 23-2 ligand (1-20)].

[Reference Example 8]

(Cloning of cDNA encoding part of rat TGR 23-2 ligand precursor)

In order to clone the cDNA encoding the precursor of rat TGR 23-2 ligand, PCR was performed using cDNA from rat whole brain as type II.

Using the following synthetic DNA primer, cDNA derived from whole rat brain was used as type II and amplified by the PCR method. The composition of the reaction solution was rat whole brain Marathon-Ready cDNA (CLONTECH) 0.81, SEQ ID NO: 22 and SEQ ID NO: 14, each of the synthetic DNA primers 1.0 / iM, 0.2mM dNTPs, E xT aq (Takara Shuzo) 0.1 a 1 and the ExTaQ buffer attached to the enzyme, the total reaction volume was 20 n 1. The amplification cycle is performed using a thermal cycler (PE Biosystems) at 94 ° C for 5 minutes, followed by a cycle at 94 ° C for 10 seconds, 65 ° C for 30 seconds, and 72 ° C for 30 seconds. Was repeated 35 times, and finally kept at 72 ° C for 5 minutes. Next, PCR reaction solution 2n1, diluted 100 times with DNase and RNase Free distilled water, primer 1.0M of SEQ ID NO: 22, synthetic DNA primer of SEQ ID NO: 15 2 xM, 0.2 mM dNTPs, ExTaQ polymerase (Takara Shuzo) 0.1 x 1 and ExTaq buffer supplied with the enzyme, the total reaction volume is 201, and the total cycler is PE Biosystems. After heating for 94 'for 5 minutes, repeat the cycle of 94 ° C · 10 seconds, 60T · 30 seconds, 72 ° C · 30 seconds 30 times, and finally 5 times at 72 ° C. Incubated for a minute. After separating the amplified DNA by 2.0% agarose gel electrophoresis, about 200 bases long DNA was cut out with a force razor, and the DNA was recovered using QIAQUick Gel Extraction Kit (Qiagen). This DNA was cloned into a pGEM-T Easy vector according to the protocol of the pGEM-T Easy Vector System (Promega). After introducing this into Escherichia coli JM109 co-immediate etent cell (Takara Shuzo) and transforming, a clone having a cDNA insert is selected on an LB agar medium containing ampicillin and X-ga1, and the white clone is selected. Only one was isolated using a sterilized toothpick to obtain a transformant. Each clone was cultured overnight in LB medium containing ampicillin, and plasmid DNA was prepared using QIAwell 8 Plasmid Kit (Qiagen). The reaction for determining the nucleotide sequence was performed using the BigDye Terminator Cycle Seauencing Ready Reaction Kit (PE Biosystems), followed by decoding using a fluorescent automatic sequencer to obtain the DNA sequence represented by SEQ ID NO: 23. Was.

The nucleotide sequence of the DNA represented by SEQ ID NO: 23 includes the amino acid sequence of rat TGR23_2 ligand obtained from whole rat brain represented by SEQ ID NO: 1, SEQ ID NO: 2 and SEQ ID NO: 3. There was a frame to code. When the DNA sequence was translated using this frame as a reading frame, the amino acid sequence represented by SEQ ID NO: 24 was obtained. By comparing this sequence with the amino acid sequence of the mouse TGR23-2 ligand precursor obtained in Reference Example 7 (SEQ ID NO: 16), this sequence was found to be a part of the rat TGR23-2 ligand precursor. It was presumed to correspond to a sequence consisting of 54 amino acids at the C-terminal side. On the 3 'side, a termination codon was present downstream of the sequence encoding rat TGR 23-2 ligand. In this sequence, the N-terminal side of the amino acid sequence of rat TGR 23-2 ligand has a Lys-Arg sequence (Seidah, N.G. et al., Ann. NY Acad. Sci., 839, 9-24, 1998). On the other hand, a termination codon was present on the C-terminal side, but two more residues were present between the sequence of the rat TGR23-2 ligand of SEQ ID NO: 1.

From this, the amino acid sequence of rat TGR 23-2 ligand was obtained from SEQ ID NO: 1 [rat TGR 23-2 ligand (1-18)] obtained from rat whole brain extract, Amino acid sequence represented by SEQ ID NO: 2 [rat TGR 23-2 ligand (1-15)] and SEQ ID NO: 3 [rat TGR 23-2 ligand (1-14)], and further SEQ ID NO: 1 It was presumed to be the amino acid sequence [rat TGR 23-2 ligand (1-20)] represented by SEQ ID NO: 25, which was extended to the C-terminal side by 2 residues.

[Reference Example 9]

(Cloning of cDNA encoding rat TGR23-2 ligand precursor) To clone cDNA encoding rat TGR23_2 ligand precursor, cDNA derived from rat whole brain was cloned into type II. Was performed.

Using the following synthetic DNA primer, cDNA derived from rat whole brain was subjected to amplification by PCR using type III. The composition of the reaction solution was rat whole brain Marathon-Ready cDNA (CLONTECH) 0.8 / 1, SEQ ID NO: 27 and SEQ ID NO: 28, each of the synthetic DNA primers 1.0 M, 0.2 mM dNTPs, E xT aq (Takara Shuzo) 0.11 and ExT aq buffer attached to the enzyme, the total reaction volume was 201. The amplification cycle was performed using a thermocycler (PE Biosystems) at 94 ° C for 5 minutes, followed by a cycle at 94 ° C for 10 seconds, 65 ° C for 30 seconds, and a cycle of 72 ° C for 30 seconds. Was repeated 35 times, and finally, the mixture was kept at 72 ° C for 5 minutes. Next, PCR reaction solution 2 {i1, a primer of SEQ ID NO: 29, 1.0 M, and a synthetic DNA primer of SEQ ID NO: 28, diluted 50-fold with distilled water of DNase and RNA Free 0.2 M, 0.2 mM dNTPs, ExTaQ polymerase (Takara Shuzo) 0.1 x 1 and the ExTaq buffer attached to the enzyme to a total reaction volume of 20 n1. ), After heating for 94T for 5 minutes, repeat the cycle of 94 ° C for 10 seconds, 65 ° C for 30 seconds, 72 ° C for 30 seconds 30 times, and finally for 5 minutes at 72 ° C. Insulated. After the amplified DNA was separated by 2.0% agarose gel electrophoresis, DNA having a length of about 350 bases was cut out with a force razor, and the DNA was recovered using a QIAauick Gel Extraction Kit (Qiagen). This DNA was cloned into a pGEM-T Easy vector according to the protocol of the pGEM-T Easy Vector System (Promega). Escherichia. Coli JM109 competent cell (Takara Shuzo) After transfection, clones containing the cDNA insert were selected on LB agar medium containing ampicillin and Xga1, and only the white clones were isolated using a sterilized toothpick and transformed. I got a body. Each clone was cultured once in LB medium containing ampicillin, and plasmid DNA was prepared using QIAwell 8 Plasmid Kit (Qiagen). The reaction for determining the nucleotide sequence was carried out using a BigDye Terminator Cycle Sequencing Ready Reaction Kit (PE Biosystems), and was decoded using a fluorescent automatic sequencer to obtain a DNA sequence represented by SEQ ID NO: 30.

The nucleotide sequence of cDNA represented by SEQ ID NO: 30 is a further 5 'of the DNA sequence (SEQ ID NO: 23) encoding a part of the rat TGR23-2 ligand precursor obtained in Reference Example 8. The sequence was extended to the side. This sequence is constructed using a frame encoding an amino acid sequence corresponding to the amino acid sequence of rat TGR 23-2 ligand obtained from whole rat brain represented by SEQ ID NO: 1, SEQ ID NO: 2 or SEQ ID NO: 3. When translated as an open reading frame, 5, upstream, a cDNA (SEQ ID NO: 7 and SEQ ID NO: 16) presumed to encode human TGR23-2 ligand precursor and mouse TGR23-2 ligand precursor One ATG was present at a position corresponding to the ATG predicted to be the start codon of the existing protein translation. In addition, a stop codon appeared in the same frame further 5 ′ upstream of this ATG codon. On the 3 'side, a stop codon was present downstream of the sequence thought to encode mouse TGR 23-2 ligand. From this, it was estimated that the sequence represented by SEQ ID NO: 30 was a cDNA sequence encoding rat TGR23_2 ligand precursor. The amino acid sequence translated from the nucleotide sequence of the cDNA represented by SEQ ID NO: 30 is shown in SEQ ID NO: 31. [Reference Example 10]

(TGR23-1 (Human TGR23-1 is sometimes simply referred to as TGR23_1)) Expression of CHO cells

Plasmid pTB2173 containing a DNA fragment having the nucleotide sequence represented by SEQ ID NO: 38 encoding TGR 23-1 was designated as type I, and the Sa1I recognition sequence was A PCR reaction was performed using the added primer 1 (SEQ ID NO: 32) and the primer 2 (SEQ ID NO: 33) added with a SpeI recognition sequence. The composition of the reaction solution in the reaction was as follows: 10 ng of the above plasmid was used as type III, PiU Turbo DNA Polymerase (Stratagene) 2.5 U, primer 1 (SEQ ID NO: 32) and primer 1 (SEQ ID NO: 2). : 33) was added at 1.0 / M each, dNTPs at 200 M, and 2 × GC Buffer I (Takara Shuzo) to the reaction mixture at 25 × 1 to give a volume of 501. The PCR reaction was repeated 95 times at 95 ° C for 60 seconds, followed by a cycle of 95 ° C for 60 seconds, 55 times, 60 seconds, 72 V, -70 seconds, and finally 72 times. The extension reaction was performed at ° C for 10 minutes. The PCR reaction product was subcloned into a plasmid vector pCR-Bluntll-TOP0 (Invitrogen) according to the prescription of Zero Blunt TOPO PCR Cloning Kit (Invitrogen). This was introduced into E. coli TOP 10 (Invitrogen), and a clone containing the cDNA of TGR23-1 contained in pTB2173 was selected in an LB agar medium containing kanamycin. The E. coli transformed with the plasmid obtained by introducing TGR23_1 into which the sequences recognized by Sa1I and SpeI were added to the 5 'and 3' Plasmids were prepared from the clones using the Plasinid Miniprep Kit (Bio-Rad) and cut with restriction enzymes Sa1I and SpeI to cut out the insert. The insert DNA was excised from an agarose gel after electrophoresis, and then recovered using a Gel Extraction Kit (Qiagen). Vector plasmid pAKKO—111H for expression of animal cells obtained by cutting this insert DNA with Sa1I and SpeI (Hinuma, S. et al. Biochim. Biophys. Acta, Vol. 1219, pp. 251-259 (1994), the same vector plasmid as pAKKO l.11H), and ligated using DNA Ligation Kit ver.2 (Takara Shuzo) to obtain a plasmid pAKKO-TGR23 for protein expression. -1 was constructed. After culturing E. coli TOP10 transformed with this pAKKO-TGR23-1, plasmid DNA of pAKKO-TGR23-1 was prepared using Plasmid Miniprep Kit (Bio-Rad).

Hamster CHOZd hfr_ cells were placed on a falcon dish (3.5 cm in diameter) in an a-MEM medium (with ribonucleosides and deoxyribozyme leosides, GIBC0, Cat. No. 12571) containing 10% fetal serum. 1 0 and ^five seeding, 5% C0 ₂ The cells were cultured at 37 ° C overnight in an incubator. The expression plasmid pAKKO-TGR23-1GRA2Xg was transfected using Transiection Reagent FuGENE 6 (Roche) according to the method described in the attached instruction manual. After culturing for 18 hours, the medium was replaced with a fresh growth medium. After further culturing for another 10 hours, the transfected cells were collected by trypsin-EDTA treatment, and selected medium (Hi-MEM medium containing 10% dialyzed fetal calf serum (without ribonucleosides and deoxyribonucleosides, GIBC0, Cat. No. 12561) )) Was used to inoculate 10 flat bottom 96-well plates. Culture was continued while changing the selection medium every 3 to 4 days, and after 2 to 3 weeks, 81 DHF R + cell clones that had grown in a colony were obtained.

[Reference Example 11]

(Quantification of the expression level of TGR23-11 in the CHO cell line expressing the T23-23-1 expression using the TaqMan1 method)

81 clones of the CHO cell line expressing TGR23-11 obtained in Reference Example 10 were cultured in a 96-well plate, and all RNAs were prepared using an RNeasy 96 Kit (Qiagen). A reverse transcription reaction was performed on 50 to 200 ng of the obtained total RNA using a TadMan Gold RT-PCR Kit (PE Biosystems). A reverse transcript equivalent to 5 to 20 ng of the obtained total RNA, or a standard cDNA prepared as described later, lxUniversal PC R Master Mix (PE Biosystems), a primer represented by SEQ ID NO: 34 and a sequence ABI PRISM 7700 Seauence Detector (PE Biosystems) for a reaction mixture 251 containing 500 nM of each of the primers represented by 35 and 500 nM of the TaqMan probe represented by SEQ ID NO: 36 PCR was performed using. PCR was performed by treating the cells at 50 ° C for 2 minutes and 95 ° C for 10 minutes, and then repeating the cycle of 95 ° C for 15 seconds and 60 ° C for 60 seconds 40 times.

The standard cDNA is prepared by measuring the absorbance at 260 nm of the plasmid PTB2174 containing the DNA fragment having the nucleotide sequence represented by SEQ ID NO: 40, calculating the concentration, calculating the exact copy number, and then including ImM EDTA. It was diluted with a 10 mM Tris-HCI (pH 8.0) solution to prepare 2 × 10 ⁶ copies of a standard cDNA solution from 2 copies. The probe and primer for TaciMan PCR are P Designed by rimer Express (Version 1.0) (PE Biosystems). The expression level was calculated using ABI PRISM 7700 SDS software. The number of cycles at the moment when the fluorescence intensity of the reporter reached the set value was plotted on the vertical axis, and the logarithmic value of the initial concentration of the standard cDNA was plotted on the horizontal axis, to create a standard curve. The initial concentration of each reverse transcript product was calculated from the standard curve, and the amount of TGR23-1 gene expression per total RNA of each clone was determined. As a result, one CH〇 cell line having high expression of TGR23-1 was selected and cultured in a 24-well plate. For these cells, the expression level of TGR 23-1 was re-examined. Total RNA was prepared using RNeasy Mini Kits (Qiagen), and then treated with RNase-free DNase Set (Qiagen). A reverse transcription reaction was performed from the obtained total RNA in the same manner as described above, and the expression level of the TGR23-1 gene per total RNA of each clone was determined by TaQMan PCR. As a result, it was found that CHO cell lines clones 49 and 52 expressing TGR23-1 showed high expression levels.

In the following reference examples, cells expressing these two clones were used.

[Reference Example 1 2]

(Human TGR 23-2 ligand (1-20): Ser-Phe-Arg-Asn-Gly-Val-Gly-Thr-Gly-Met-Lys-Lys-Thr-Ser-P e-Gln-Arg-Ala Of -Lys-Ser (SEQ ID NO: 1 2)

)

Commercially available Boc- Ser (Bzl) - a 0CH ₂ -PAM resin, is input to the reaction vessel of the peptide synthesizer ACT 90, the B oc swelling after TFA was removed in DCM, and neutralized with DI EA. This resin was suspended in NMP, and Boc-Lys (Cl-Z) was condensed with HOBt-DIPCI. After the reaction, the presence of free amino groups was examined by a ninhydrin test. When the ninhydrin test was positive, the same amino acid was condensed again. Even after the recondensation, when the ninhydrin test was positive, acetylation was performed with acetic anhydride. Repeat this cycle Boc_Ala, Boc-Arg (Tos), Boc-Gln, Boc-Phe, Boc-Ser (Bzl), Boc-Thr (Bzl) Boc-Lys (Cl-Z), Boc-Lys (CL-Z ), Boc-Met, Boc-Gly, Boc-Thr (Bzl), Boc-Gly, Boc-VaL Boc-GIy, Boc-Asn, Boc-Arg (Tos), Boc-Phe, Boc-Ser (Bzl) Condensation was performed in the order of arrangement to obtain 0.24 g of the desired protected peptide resin. Fluoride this resin with 1.5 ml of p-cresol. After stirring at 0 ° C. for 60 minutes in about 15 ml of hydrogen fluoride, hydrogen fluoride was distilled off under reduced pressure, and getyl ether was added to the residue, followed by filtration. Water and acetic acid were added to the filtrate to extract the peptide, which was separated from the resin. The extract was concentrated, applied to a Sephadex (trademark) G-25 column (2.0 x 80 cm) filled with 50% acetic acid, developed with the same solvent, and the main fractions were collected and lyophilized. An aliquot (45 mg) was applied to a reversed-phase chromatography column (2.6 x 60 cm) packed with LiChroprep ™ RP-18, washed with 200 ml of 0.1% TFA water, and 300 ml of 0.1% TFA water. A linear gradient elution was performed using 300 ml of 25% acetonitrile water containing 1 and 0.1% TFA, and the main fraction was collected and lyophilized to obtain 12.7 mg of the target peptide.

ESI-MS: molecular weight MW 21 88.0 (theoretical 2 187.5) HPLC elution time 10.6 min

Column conditions: Column: Wakosil 5C18T 4.6 X 100mm

Eluent: Solution A—0.1% TFA water, Solution B—0.1% TF A in acetonitrile A / B: Linear gradient elution to 95Z5 to 45/55 (25 minutes) Flow rate: 1.0m 1 minute

[Reference Example 13]

(Measurement of intracellular Ca ion concentration increasing activity of TGR23-1 expressing CHO cells and TGR23-2 expressing CH0 cells by human TGR23-2 ligand (1-20) using FLIPR)

The human TGR 23-2 ligand (1-20) obtained in Reference Example 12 was administered at various concentrations to TGR 23-1 -expressing CHO cells and TGR 23-2 -expressing CHO cells at various concentrations according to a known method. When the Ca ion concentration increasing activity was measured using FLIPR, human TGR23-2 ligand (1-20) was found to be dependent on the concentration of TGR23-1 expressing CHO cells and TGR23-2 expressing CH0. Increased intracellular Ca ion concentration in cells. The results are shown in FIGS.

Thus, the polypeptide having the amino sequence represented by SEQ ID NO: 12 [human TGR 23-2 ligand (1-20)] increased the intracellular Ca ion concentration of TGR 23-1 and TGR 23-2. It is clear that it has activity. Industrial applicability

According to the present invention, it is possible to predict a binding molecule or a type of a binding molecule only by obtaining information on an amino acid sequence of a protein whose binding molecule is unknown (and / or a sequence alignment obtained using the amino acid sequence). Becomes possible. This makes it possible to predict binding molecules (ligands, etc.) much faster than conventional molecular modeling methods that predict even three-dimensional structures.

Further, according to the present invention, it is possible to predict the binding molecule or the type of the binding molecule for a protein for which various types of binding molecules are unknown. Further, it is possible to predict the binding molecule or the kind of the binding molecule easily and quickly rather than experimenting whether or not any binding molecule actually binds to the protein whose binding molecule is unknown.

By using a known technique, a function common to a group of proteins having homology can be estimated by calculating a sequence alignment or creating a three-dimensional structure model. On the other hand, identification of ligands (and their types) from similarity assessments by sequence alignment and molecular docking calculations is not possible with current technology. However, in order to estimate the individual functions of individual GPCRs useful for drug development, it is necessary to predict the binding ligand and conjugated G protein. Detailed functions such as investigating intracellular responses through conjugated G-proteins, measuring biodistribution and changes in expression levels, gene transfer, and creating deficient animals only after such estimation and determination of a set of GPCRs and ligands Research will progress. The prediction method using ligand-determined residues and the computer for it are directly useful for molecular identification and function elucidation essential for such drug development.

According to the present invention, it is possible to very easily predict the binding molecule or the type of the binding molecule of a binding molecule unknown protein (such as an orphan G protein-coupled receptor). It is possible to easily produce a prophylactic or therapeutic drug for a disease or the like involving a binding molecule unknown protein.

Claims

The scope of the claims

1. A binding molecule prediction method for a binding molecule unknown protein, which predicts a binding molecule that binds to the binding molecule unknown protein,

For known binding molecule known proteins whose amino acid sequence and binding molecule are known, sequence alignment of at least two or more known binding molecule proteins and binding molecule known protein classification information in which the binding molecules or types of binding molecules are associated with each other. The process of obtaining

Using the binding molecule known protein classification information, one or more binding molecule determining residue positions that are assumed to be involved in determining the binding molecule among the sequence alignment positions of the binding molecule known protein By specifying the identifying step, associating the amino acid residue (binding molecule determining residue) at the binding molecule determining residue position with the type of the binding molecule or the binding molecule, the binding molecule determining residue and the binding molecule or the binding molecule Obtaining binding molecule-determined residue-binding molecule classification information indicating a correlation with the type of

Aligning the sequence of the unknown binding molecule protein with respect to the sequence alignment between the known binding molecule proteins for the same type of unknown binding molecule protein, and obtaining a sequence alignment of the unknown binding molecule protein;

Applying information on at least one binding molecule determining residue in the sequence alignment of the binding molecule unknown protein to the binding molecule determining residue-binding molecule classification information, Predicting the type of

Binding molecule prediction method for unknown protein.

2. The method according to claim 1, wherein the binding molecule is one of a ligand, a regulator, an effector, and a coenzyme.

3. The method according to claim 1 or 2, wherein the binding molecule is classified into two or more types, and the type of the classified binding molecule is predicted.

4. Binding protein unknown protein is G protein-coupled receptor, kinase, lipase The method for predicting a binding molecule of an unknown protein for a binding molecule according to any one of claims 1 to 3, wherein the method is any one of enzyme, transporter, protease, and ion channel.

5. The step of identifying one or more binding molecule-determining residue positions, wherein one or more binding molecule-determining residue positions are identified from the amino acid residues constituting the sequence alignment and the type of binding molecule. 5. The method for predicting a binding molecule of an unknown protein according to any one of claims 1 to 4.

6. The method for predicting a binding molecule of an unknown protein for a binding molecule according to any one of claims 1 to 4, wherein the position of the residue for determining the binding molecule is determined by using one or both of the following formulas 1 and / or 2. . f l (n) = ∑ (N (Res, Χ) XN (Res, Xr)) Equation 1

Res

[In Equation 1, n represents fl (n) as an evaluation function for the nth amino acid residue in the sequence alignment of the binding molecule known protein, and Res represents the type of amino acid residue. Χα and Xr represent a binding molecule or a type of a binding molecule, Q represents an integer from 1 to P-1, r represents an integer greater than Q and less than or equal to p, and p represents a binding molecule or a bond. N (Res, XQ) represents the number of types of molecules, and among the known binding molecule proteins in the binding molecule known protein classification information, the nth amino acid residue in the sequence alignment is Res, and the binding molecule is Represents the number of XQs, and N (Res, Xr) is the nth amino acid residue of the sequence alignment among the proteins known in the binding molecule known protein classification information that is Res, and Join Represents the number of molecules whose molecule is Xr. ] f 2 (n) = ∑ (N (Res, XI) xN (Res, X2)-X N (Res, Xp)) Equation 2

Res

[In Formula 2, n represents that f 2 (n) is an evaluation function for the nth amino acid residue in the sequence alignment of the binding molecule known protein, and Res represents the amino acid residue. Represents the type, XI to Xp represent the binding molecule or the type of the binding molecule, p Represents the number of binding molecules or types of binding molecules, and N (Res, ¾ represents the nth amino acid residue in the sequence alignment among the known binding molecule proteins in the binding molecule known protein classification information. Is Res and the number of binding molecules is X.]

7. The unknown molecule of the binding molecule according to any one of claims 1 to 4, wherein the residue position is determined using at least one of the following expressions 3, 4, and 5: Binding molecule prediction method. fl (n) = ∑ (N (Res, XQ) XN (es, Xr)) Equation 3

Res

[In Equation 3, n represents fl (n) as an evaluation function for the nth amino acid residue in the sequence alignment of the binding molecule known protein, and Res represents the type of amino acid residue. XQ and Xr represent a binding molecule or a type of a binding molecule, _Q represents an integer from 1 to p-1, r represents (an integer greater than or equal to and less than or equal to p, and p represents a binding molecule or N (Res, XQ) represents the number of types of binding molecules. Among the known binding molecule proteins in the binding molecule known protein classification information, the n-th amino acid residue in the sequence alignment is Res. N (Res, Xr) is the nth amino acid residue in the sequence alignment among the known binding molecule proteins in the binding molecule known protein classification information. Is Res, and the bond is There a number from what is Xr.] F2 (n) = Σ (N (Res, XI) XN (Res, X2) ~xN (Res, Xp)) Equation 4

Res

[In the formula 4, n is ί2 (η).

Represents the evaluation function for the ηth amino acid residue of the group, Res represents the type of the amino acid residue, XI to Xp represents the binding molecule or the type of the binding molecule, and p represents the binding function. N (Res, X) represents the number of types of molecules or binding molecules. The nth amino acid residue of the comment is Res and the number of binding molecules is X. 1 f 3 (m, n) = {(number of amino acid residue pair types) / wX + wl x (number of 2 cross residue pair types) + w2 X (3 cross residue residue pair types) tens… wp-1 ( p crossing residue pair type) + wAX (number of non-alignable amino acid residues) + wB X (number of non-alignable amino acid residues)} Formula 5

[In the formula 5, (in, n) is f 3 (m, n), a sequence of proteins with known binding molecules:

This is an evaluation function for the m-th and n-th amino acid residues in the sequence, and the number of amino acid residue pair types is the m-th and n-th amino acid residues in the sequence alignment of the binding molecule known protein. Indicates the number of combinations of amino acid residues.The number of 2 cross residue pairs and the number of 3 cross residue pairs are the m-th and n-th amino acids in the sequence alignment of proteins with known binding molecules, respectively. The number of residue combinations means the number of two or three ligands.The number of p-crossing residue types is the number of the mth and nth amino acid residues in the sequence alignment of binding protein known proteins. Means the number of p-type ligands out of the combinations, and the number of amino acid residues that cannot be sequence-aligned refers to the number of amino acid residues in the protein with a known binding molecule. One of the m-th and n-th amino acid residues in the sequence alignment means the number of sequences that cannot be sequence-aligned to obtain favorable homology, and the number of amino acid residue pairs that cannot be sequence-aligned. The number means that both the m-th and the n-th amino acid residues in the sequence alignment of the binding molecule known protein were determined to be incapable of sequence alignment in order to obtain favorable homology, and wX represents A distribution function that takes a positive constant or the number of amino acid pair types as a variable, and means a distribution function that gives the maximum value when the number of amino acid pair types is a positive number of 400 or less, and wl to wp-1 , WA and wB are weights and are positive numbers. ]

8. The step of obtaining ligand-determined residue-ligand classification information comprises:

Among the amino acid residues of the ligand-known protein, Riga with the smallest value of the function ί3 (η) Extracting those at the residue positions determined by the

A step of determining the number (A) of extracted ligand-determined residues that match the extracted ligand-determined proteins among the ligand-determined residue-ligand classification information; Determining the number (B) of the ligand-identified proteins that correspond to the extracted ligand-determining residues among the ligand-identified proteins, and the type of the ligand or the ligand matches the ligand-identified protein. The groups at the ligand-determining residue positions where the value of the function f 3 (n) is the second smallest or the X-th (where X represents an integer greater than 2 and less than 100) The extraction step, and the ligand-determined residue matches the extracted ligand-determined residue among the known ligand proteins listed in the ligand classification information. Determining the number (C) of the ligands, and determining whether the ligand or the type of ligand is the ligand or the ligand among the ligand-determined residues extracted from the known ligand proteins listed in the ligand-determined residue-ligand classification information. Determining a number (D) that matches that of the known protein;

Obtaining a sum (E) of (A) and (C);

Obtaining the sum (F) of (B) and (D);

Including

8. The method for predicting a binding molecule of an unknown protein for a binding molecule according to claim 7, wherein the information on (E) and (F) is further displayed.

9. For at least two or more proteins with known binding molecules whose amino acid sequence and binding molecule are known, a binding molecule that associates the sequence alignment of the protein with known binding molecule with the type of binding molecule or binding molecule. Obtaining known protein classification information;

A step of identifying one or more binding molecule determining residue positions that are assumed to be involved in determining the binding molecule in the sequence alignment of the binding molecule known protein using the binding molecule known protein classification information. When,

By associating the amino acid residue at the position of the binding molecule determining residue (binding molecule determining residue) with the type of the binding molecule or the type of the binding molecule, binding to the binding molecule determining residue Obtaining a binding molecule-determined residue-binding molecule classification information indicating a correlation with a molecule; and a method for predicting a binding molecule of an unknown protein.

10. The binding molecule-determined residue indicating the correlation between the binding molecule-determined residue and the binding molecule. The binding molecule classification information includes the binding molecule unknown protein of the same type as the binding molecule unknown protein. Input information on the sequence alignment of the unknown binding molecule protein obtained by aligning the sequence of the unknown binding molecule protein with the sequence alignment between the two, and the binding molecule or the binding molecule that binds to the unknown binding molecule protein A method for predicting the binding molecule of an unknown protein, which predicts the type of molecule.

11. The step of predicting a binding molecule that binds to a binding molecule unknown protein or a type of the binding molecule using the binding molecule unknown protein prediction method according to any one of claims 1 to 10. A method for producing a medicament comprising:

12. The claim wherein the medicament is one or both of a preventive agent and / or a therapeutic agent for a central disease, an inflammatory disease, a circulatory disease, a cancer, a metabolic disease, an immune system disease or a digestive system disease. The method for producing a medicament according to the above.

13. A method for determining the position of a residue for determining a binding molecule using one or both of the following formulas 6 and 7. f l (n) = ∑ (N (Res, Χ) XN (Res, Xr)) Equation 6

Res

[In Equation 6, n indicates that fl (n) is an evaluation function for the nth amino acid residue in the sequence alignment of the binding molecule known protein, and Res indicates the type of amino acid residue. XQ and Xr represent binding molecules or types of binding molecules, Q represents an integer from 1 to P-1, r represents an integer greater than Q and less than or equal to p, and p represents a binding molecule or a bond. N (Res, XQ) represents the number of types of molecules, and among the known binding molecule proteins in the binding molecule known protein classification information, the nth amino acid residue in the sequence alignment is Res, and the binding molecule is ¾ represents the number of ある, and N (es, Xr) is the known binding molecule protein in the binding molecule known protein classification information. Among the proteins, S: represents the number of amino acids in which the n-th amino acid residue is Res and the binding molecule is Xr. ] f2 (n) = ∑ (N (Res, XI) XN (Res, X2) -XN (Res, Xp)) Equation 7

Res

[In equation 7, n represents that ί2 (η) is an evaluation function for the ηth amino acid residue in the sequence alignment of the known binding molecule protein, and Res represents the type of amino acid residue. XI to Xp represent binding molecules or types of binding molecules, p represents the number of binding molecules or types of binding molecules, and N (Res, X) represents the binding molecule known protein classification information. Among the known binding molecule proteins, this indicates the number of proteins in which the n-th amino acid residue in the sequence alignment is Res and the binding molecule is X. ]

14. A method for determining a binding molecule determining residue position using the following formula 8. f 3 (m, n) = {(number of amino acid residue pair types) / wX + wlx (number of 2 cross residue pairs) + w2X (3 cross residue pair types) + wp-1 (p cross residues + WAX (number of non-alignable amino acid residues) + wBX (number of non-alignable amino acid residues)} Formula 8

[In Equation 8, (m, n) indicates that f3 (m, n) is an evaluation function for the mth and nth amino acid residues in the sequence of the protein with a known binding molecule. The number of amino acid residue pair types represents the number of combinations of the mth and nth amino acid residues in the sequence alignment of the binding molecule known protein. The number of pairs means the number of the two and three ligands in the m-th and n-th amino acid residue combinations in the sequence alignment of the binding molecule known protein, respectively. The number of baie types means the number of p-type ligands out of the m-th and n-th amino acid residue combinations in the sequence alignment of binding protein known proteins. The number of amino acid residues that cannot be sequence-aligned means that one of the m-th and n-th amino acid residues in the sequence alignment of a protein with a known binding molecule has a sequence alignment in order to obtain favorable homology. The number of amino acid residue pairs that cannot be sequence-aligned means that both the m-th and n-th amino acid residues in the sequence alignment of a protein with a known binding molecule have favorable homology. WX is a positive constant or a distribution function with the number of amino acid pair types as a variable, where wX is a positive number with 400 or less amino acid pair types. Wl '"wp-1, wA, wB are weights and mean a distribution function that gives the maximum value at a time. That.]

15. Determining the binding molecule among the positions of the sequence alignment of the known binding molecule protein using the amino acid sequence or sequence alignment of the binding molecule known protein and the information on the binding molecule or the type of the binding molecule. Binding molecule, which is the amino acid residue at the position supposed to be involved in the binding molecule (binding molecule determining residue position), and a binding molecule indicating the correlation between the binding molecule or the type of the binding molecule. A computer for predicting a binding molecule of an unknown protein, which obtains classification information,

The computer is

A sequence alignment input means for inputting information regarding the sequence alignment of the binding molecule known protein;

Sequence alignment binding molecule storage means for storing the amino acid sequence or sequence alignment of the binding molecule known protein input by the sequence alignment input means, and information on the type of binding molecule or binding molecule;

Using the amino acid sequence or sequence alignment of the known binding protein stored by the sequence alignment binding molecule storage means and the information on the type of binding molecule or binding molecule to determine the position of the binding molecule determining residue A binding molecule determining residue position determining means;

An amino acid residue (binding molecule determining residue) at the binding molecule determining residue position; By associating a molecule or a binding molecule type with a binding molecule determining residue, a binding molecule determining residue indicating the correlation between the binding molecule or the binding molecule or the type of the binding molecule, and a binding molecule determining residue for obtaining binding molecule classification information One binding molecule classification information obtaining means,

For the unknown binding molecule protein of the same type as the binding molecule unknown protein, the binding molecule unknown protein obtained by aligning the sequence of the binding molecule unknown protein with the sequence alignment between the binding molecule known protein Sequence alignment input means for inputting information regarding sequence alignment,

Using the information on the sequence alignment of the unknown protein of the binding molecule input by the sequence alignment inputting means to the binding molecule determination residue-binding molecule classification information, the binding molecule or the binding of the unknown protein of the binding molecule. Predict the type of molecule,

Computer for predicting binding molecules of unknown proteins.

16. The binding molecule-determining residue-determining means uses at least one of the following formulas 9 and / or 10 or a function of both to predict the binding molecule of the unknown binding molecule protein according to claim 15. Computer. f l (n) = ∑ (N (Re s, XQ) XN (Res, Xr)) Equation 9

Res

[In Formula 9, n is fl (n), and is a known binding molecule protein:

Represents the evaluation function for the nth amino acid residue in the sequence, Res represents the type of amino acid residue, XQ and Xr represent the binding molecule or the type of binding molecule, and Q represents represents an integer up to p-1, r represents an integer greater than Q and less than or equal to p, p represents the number of binding molecules or types of binding molecules, and N (Res, XQ) represents a class of proteins known to have binding molecules. Among the known binding molecule proteins in the information, the number of the nth amino acid residue in the sequence alignment is Res and the binding molecule is XQ, and N (Res, Xr) is the classification of known binding molecule proteins. Binding molecule known tamper existing in information Among the proteins, the number of amino acids in which the n-th amino acid residue is Res and the binding molecule is Xr is shown. f2 (n) = ∑ (N (Res, XI) XN (Res, X2) -XN (Res, Xp)) Equation 10

Res

[In Formula 1O, n represents that ί2 (η) is an evaluation function for the η-th amino acid residue among the sheets of the binding molecule known protein, and Res represents the type of the amino acid residue. And XI to Xp represent the binding molecule or the type of the binding molecule, p represents the number of the binding molecule or the type of the binding molecule, and N (Res, X) represents the binding molecule known in the protein classification information. It indicates the number of proteins whose binding molecule is X and the nth amino acid residue in the sequence alignment is Res among known binding molecule proteins. ]

17. The computer for predicting a binding molecule of a binding molecule unknown protein according to claim 15 or 16, wherein the binding molecule determination residue position determining means uses a function represented by the following formula 11: . f 3 (m, n) = {(the number of amino acid residue pair types) / wX + wlx (the number of 2 cross residue pairs) + w2X (the number of 3 cross residue pairs) tens… wp-1 (p cross residue + WAX (number of non-alignable amino acid residues) + wBX (number of non-alignable amino acid residues)} Formula 11

[In Equation 11, (πι, η) indicates that f 3 (m, n) is an evaluation function for the mth and nth amino acid residues in the sequence alignment of the binding molecule known protein. And the number of amino acid residue pair types represents the number of types of combinations of the mth and nth amino acid residues in the sequence alignment of the binding molecule known protein. The number of types of cross-residue pairs means the number of the two and three ligands, respectively, of the combination of the m-th and n-th amino acid residues in the sequence alignment of the known binding molecule proteins. The P cross residue The number of pairs means the number of combinations of the m-th and n-th amino acid residues in the sequence alignment of the binding molecule known protein with ligands of p types, and the number of amino acid residues that cannot be sequence-aligned means Sequence alignment means that one of the m-th and n-th amino acid residues in the sequence alignment of the protein whose binding molecule is known is considered to be incapable of sequence alignment in order to obtain favorable homology. The number of unsuccessful amino acid residue pairs means that both m-th and n-th amino acid residues in the sequence alignment of a known binding molecule were not sequence-aligned due to favorable homology. WX is a positive constant or the number of amino acid pairs Wl '"wp_l, wA, and wB are weights, and are distribution functions that give the maximum value when the number of amino acid pairs is a positive number of 400 or less. is there. ]

1 8. Binding molecule A computer for predicting the binding molecule of an unknown protein,

It is assumed to be involved in determining a molecule that binds to the protein with a known binding molecule in the sequence alignment of the protein with a known binding molecule, which is of the same type as the unknown protein and has a known binding molecule. And the binding molecule determinant residue corresponding to the binding molecule determinant residue, and the binding molecule determinant residue corresponding to the amino acid residue of the protein known at the binding molecule determinant residue position. Storage means for storing information relating to the binding molecule of the protein or the type of the binding molecule;

Information on the sequence alignment of the unknown binding molecule protein obtained by aligning the sequence of the unknown binding molecule protein with the sequence alignment between the known binding molecule proteins for the same type of unknown binding molecule protein as the known binding molecule protein. And a binding molecule determining means for determining a binding molecule or a type of a binding molecule of the unknown protein of the binding molecule from the input information on the sequence alignment and the information stored in the storage means. When, Display means for displaying the determined binding molecule or the type of the binding molecule binding to the unknown binding molecule unknown protein,

The information on the sequence alignment of the unknown binding molecule protein input by the sequence alignment input means, the binding molecule determinant residue stored in the storage means, and the binding molecule of the binding molecule known protein corresponding to the binding molecule determinant, or The binding molecule determining unit predicts the binding molecule of the binding molecule unknown protein or the type of the binding molecule based on the information on the type of the binding molecule and the binding molecule of the binding molecule unknown protein predicted by the binding molecule determination unit. Display the type of binding molecule by display means

Computer for predicting binding molecules of unknown proteins.

1 9.

Inputting information on the sequence alignment of the binding molecule known protein Stores the amino acid sequence or sequence alignment of the binding molecule known protein input by the sequence alignment input means, and information on the type of the binding molecule or the binding molecule. Sequence alignment binding molecule storage means,

The binding molecule determination residue for determining the binding molecule determination residue position using the amino acid sequence or sequence alignment of the binding molecule known protein stored by the sequence alignment binding molecule storage means and information on the type of binding molecule or binding molecule. Base position determining means,

By associating the amino acid residue (binding molecule determining residue) at the binding molecule determining residue position with the type of the binding molecule or the binding molecule, the correlation between the binding molecule determining residue and the type of the binding molecule or the binding molecule is determined. Means for obtaining binding molecule-determined residue-binding molecule classification information for obtaining binding molecule-determined residue-binding molecule classification information representing the relationship;

For a binding molecule unknown protein of the same type as the binding molecule unknown protein, input information on the sequence alignment of the binding molecule unknown protein obtained by aligning the sequence of the binding molecule unknown protein with respect to the sequence alignment between the binding molecule unknown proteins. Sequence alignment input means A program to function as

20. The program according to claim 19, wherein the binding molecule determining residue position determining means uses at least one of the following formulas 12 and 13 or a function of both. fl (n) = ∑ (N (Res, Χα) XN (Res, Xr)) Equation 1 2

Res

[In Equation 12, n represents that Π (η) is an evaluation function for the ηth amino acid residue in the sequence alignment of the binding molecule known protein, and Res represents the amino acid residue XQ and Xr represent binding molecules or binding molecule types, Q represents an integer from 1 to p-1, r represents an integer greater than Q and less than or equal to p, and p represents a bond N (Res, XQ) represents the number of types of molecules or binding molecules, and among the known binding molecule proteins present in the binding molecule known protein classification information, the nth amino acid residue in the sequence alignment is Res And N (Res, Xr) is the nth amino acid residue in the sequence alignment among the known binding molecule proteins in the binding molecule known protein classification information. Is Res and AND Represents the number of molecules whose molecule is Xr. ] f2 (n) = ∑ (N (Res, XI) xN (Res, X2) —XN (Res, Xp)) Equation 13

Res

[In Formula 13, n is 、 2 (η), and is a known binding molecule protein:

Represents the evaluation function for the ηth amino acid residue of the amino acid residue, Res represents the type of the amino acid residue, XI to Xp represents the binding molecule or the type of the binding molecule, and p represents N (Res, X) indicates the number of binding molecules or types of binding molecules, and N (Res, X) indicates that the nth amino acid residue in the sequence element is the known binding molecule among the proteins known in the protein classification information. Res and represents the number of those in which the binding molecule is X. ]

2 1. The binding molecule determining residue position determining means uses a function represented by the following equation 14. 21. The program according to claim 19 or 20. f 3 (m, n) = {(number of types of amino acid residue pairs) / wX + wl x (number of types of 2 crossing residue pairs) + w2 X (number of types of 3 crossing residue pairs) tens… wp-1 (p crossing Number of types of residue pairs) + wAX (number of non-alignable amino acid residues) + wB X (number of non-alignable amino acid residues)} Formula 14

[In Equation 14, (m, n) indicates that f 3 (m, n) is an evaluation function for the m-th and n-th amino acid residues in the sequence alignment of proteins with known binding molecules. The number of amino acid residue pair types represents the number of types of combinations of the mth and nth amino acid residues in the sequence alignment of the binding molecule known protein. 3 The number of crossing residue pair types means the number of two and three ligands out of the combination of the mth and nth amino acid residues in the sequence alignment of the binding molecule known protein, respectively. The number of types of p-crossing residue pairs means the number of combinations of the ligand power of the m-th and n-th amino acid residues in the sequence alignment of the binding molecule known protein. The number of amino acid residues that cannot be aligned means that one of the m-th and n-th amino acid residues in the sequence alignment of a protein with a known binding molecule is sequence alignment in order to obtain favorable homology. The number of amino acid residue pairs that cannot be sequence-aligned refers to the sequence alignment of a protein with a known binding molecule; both the m-th and n-th amino acid residues in the sequence are preferred. WX is a positive constant or a distribution function with the number of amino acid pair types as a variable, and wX is a positive function with 400 or less amino acid pair types in order to obtain homology. Wl '"wp-1, wA, and wB are weights and positive numbers."

2 2.

A binding molecule whose binding molecule is of the same type as the unknown protein and whose binding molecule is known is included in the sequence alignment of the protein whose binding molecule is known. Binding molecule determined residue positions that are assumed to be involved in determining molecules that bind to proteins, and binding molecule determination residues that are amino acid residues of known binding molecule proteins at the determined binding molecule determined residue positions. A storage means for storing information about the group and the type of the binding molecule or the type of the binding molecule of the known binding molecule corresponding to the binding molecule determining residue;

Information on the sequence alignment of the unknown binding molecule protein obtained by aligning the sequence of the unknown binding molecule protein with the sequence alignment between the known binding molecule proteins for the same type of unknown binding molecule protein as the known binding molecule protein. A sequence alignment inputting means for inputting the information of the sequence alignment, and information stored in the storage means to determine a binding molecule or a type of the binding molecule of the unknown protein of the binding molecule;

A program that functions as display means for displaying the determined binding molecule or the type of binding molecule that binds to the unknown binding molecule protein.

23. A recording medium storing the program according to any one of claims 19 to 22.