WO2003056461A1 - Dispositif de prediction de structure proteique, procede de prediction de structure proteique, programme et support d'enregistrement associes - Google Patents
Dispositif de prediction de structure proteique, procede de prediction de structure proteique, programme et support d'enregistrement associes Download PDFInfo
- Publication number
- WO2003056461A1 WO2003056461A1 PCT/JP2002/013832 JP0213832W WO03056461A1 WO 2003056461 A1 WO2003056461 A1 WO 2003056461A1 JP 0213832 W JP0213832 W JP 0213832W WO 03056461 A1 WO03056461 A1 WO 03056461A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- fragment
- sequence
- matrix
- cluster
- query
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B30/00—ICT specially adapted for sequence analysis involving nucleotides or amino acids
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B15/00—ICT specially adapted for analysing two-dimensional or three-dimensional molecular structures, e.g. structural or functional relations or structure alignment
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B15/00—ICT specially adapted for analysing two-dimensional or three-dimensional molecular structures, e.g. structural or functional relations or structure alignment
- G16B15/20—Protein or domain folding
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B30/00—ICT specially adapted for sequence analysis involving nucleotides or amino acids
- G16B30/10—Sequence alignment; Homology search
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
- G16B40/30—Unsupervised data analysis
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
Definitions
- the present invention relates to a protein structure prediction device, a protein structure prediction method, a program, and a recording medium, and in particular, to a protein structure prediction device, a protein structure prediction method, a program, which predicts a three-dimensional structure of a protein based on a sequence-structure correlation. And a recording medium.
- the three-dimensional structure of a protein is uniquely determined from sequence information.
- comparing the size of the array space and the structure space (native structure space) it can be said that the array space is larger. This is because, from an evolutionary perspective, even if the sequence changes slightly, the structure does not seem to change much. In other words, the structure is evolutionarily more conservative than the sequence.
- Literature 1 states that the local structure is restricted to a specific biased structure by the local sequence, so that the structural space is reduced, that the structure is similar to the structure of a protein close in sequence, It discloses that a profile of a sequence is obtained by multiple alignment, and a short distance from a query sequence is obtained.
- Reference 2 also shows that if there is a correlation between the fragment structure and sequence, a limited number of structure candidates can be extracted from the fragment sequence tendency, and the structure is clustered using two structural indices.
- the sequence is calculated using the distance of frequencyprofi 1 e.Cluster creation is actually performed by searching for similar structures and searching for those with similar structures. It is disclosed to create a cluster of one structural fragment.
- FIG. 1 is a diagram showing an example of a case where an array is represented by a profile according to the conventional technology
- FIG. 2 is a diagram showing an image of creating a structural cluster according to the conventional technology.
- the array is represented by a profile.
- a profile is created by setting “1” to the amino acid corresponding to the sequence (AGGED).
- AGGED amino acid corresponding to the sequence
- Fig. 1 (b) a profile of this cluster is created as shown in Fig. 1 (b). That is, a profile is created by setting the frequency of amino acids present at a certain position with respect to a sequence belonging to a cluster.
- clustering is performed in the sequence space (FIG. 2 (a)) so that similar sequences have the same cluster (1 to 5 in FIG. 2 (a)). That is, by calculating the similarity of the sequence profile and calculating the similarity of the entire sequence, an isotropic cluster is created.
- the present invention provides a protein structure predicting apparatus, a protein structure predicting method, and a method for calculating a partial structure correlation from a partial sequence so as to be able to represent a complex variety of correlation and the confidence of the correlation. It aims to provide programs and recording media. Disclosure of the invention
- the protein structure prediction apparatus creates a fragment sequence obtained by dividing sequence information into a predetermined length and a fragment structure corresponding to the fragment sequence based on the sequence information and three-dimensional structure information of the protein.
- a fragment structure cluster creating means for creating a fragment structure cluster based on the degree of similarity; a fragment sequence similarity search means for performing a sequence similarity search on the fragment sequence with the surrounding fragment sequences in the sequence space; and the fragment sequence
- a confidence matrix creating means for creating a confidence matrix that displays a confidence that is a probability that the similar sequence of the above belongs to the fragment structure cluster as a matrix of the fragment sequence and the structure cluster;
- a query sequence input unit to be input, and the query sequence input by the query sequence input unit is divided into predetermined lengths.
- Query fragment sequence creating means for creating a combined fragment sequence
- query fragment sequence similarity searching means for performing a sequence similarity search with the fragment sequence for the query fragment sequence created by the query fragment sequence creating means
- Fragment structure probability for calculating the probability that the query fragment sequence belongs to the fragment structure cluster based on the certainty matrix created by the certainty matrix creation means and the search result of the query fragment sequence similarity search means
- a fragment sequence structure predicting means for predicting the fragment structure of the query sequence based on the probability calculated by the fragment structure probability calculating means.
- a fragment sequence obtained by dividing sequence information into a predetermined length and a fragment structure corresponding thereto are created based on the sequence information and three-dimensional structure information of the protein, and based on the similarity between the fragment structures.
- a fragment structure cluster is created by performing a sequence similarity search on the fragment sequence with the surrounding fragment sequences in the sequence space. Create a confidence matrix that displays the confidence that is the probability that belongs to the fragment structure cluster as a matrix of the fragment sequence and the structure cluster. Then, the user inputs a query sequence, divides the input query sequence into predetermined lengths to create a query fragment sequence, and performs a sequence similarity search with the fragment sequence for the created query fragment sequence.
- the probability that the query fragment sequence belongs to the fragment structure cluster is calculated based on the created confidence matrix and the search results, and the fragment structure of the query sequence is predicted based on the calculated probability. It is possible to calculate the correlation of the substructure from the subsequence and predict the substructure so that various manifolds and the certainty of the correlation can be expressed. That is, according to the present invention, when calculating a structure, the probabilities (confidence) of a plurality of structure candidates are given and given according to the degree of correlation (a function of the confidence is used as the probability of structural change). Can be.
- the present apparatus first forms a cluster of partial structures and takes into account the complex shape of the structure sequence correlation polymorphism and queries ( query)
- a sequence correlation cluster can be created dynamically after an array is given.
- the system creates structural clusters from different viewpoints (eg, fragment sequence length, fragment structure resolution, structural cluster size, degree of correlation, etc.) It becomes possible to calculate the structure by integrating the structure prediction results.
- viewpoints eg, fragment sequence length, fragment structure resolution, structural cluster size, degree of correlation, etc.
- the protein structure predicting apparatus is the protein structure predicting apparatus according to the above, wherein the similarity search result is obtained by performing a similarity search on the fragment sequence by the fragment sequence similarity search means in the form of a fragment sequence matrix.
- a similarity matrix creating means, and a structural cluster information matrix displayed by the fragment sequence and the matrix of the structural clusters showing structural cluster information indicating which fragment structural cluster the fragment sequence belongs to
- the information processing apparatus further comprises: a structure cluster to be created and an information matrix creating means, wherein the certainty matrix creating means comprises: the similarity matrix created by the similarity matrix creating means;
- the method is characterized in that the certainty matrix is created based on the structural cluster information matrix created by the raster information matrix creating means. This more specifically shows an example of creating the certainty matrix.
- a similarity matrix creating means for creating a similarity matrix represented by a matrix of fragment sequences based on the similarity search result of the fragment sequences, and a fragment structure cluster to which the fragment sequences belong Since the structural cluster information shown is represented by a fragment array and a matrix of structural clusters, a structural cluster one information matrix is created, and a confidence matrix is created based on the created similarity matrix and structural cluster information matrix. Using the calculation method, the certainty factor can be easily and precisely calculated based on the similarity search result.
- the protein structure predicting apparatus is the protein structure predicting apparatus according to the above, wherein the whole structure optimization that performs a predetermined optimization on the initial whole structure determined by the fragment structure having the maximum certainty factor. Characterized by further comprising a conversion means. According to this device, a predetermined optimization is performed on the initial overall structure determined by the fragment structure having the maximum certainty factor. First, when the initial structure is created, it is divided into various possible fragment sequences and Will be able to integrate the best prediction results. Further, by further optimizing the initial structure, the accuracy of the overall structure prediction can be further improved.
- the present invention relates to a method for predicting protein structure, and a method for predicting protein structure according to the present invention relates to a method for predicting a protein structure, comprising the steps of: A fragment structure corresponding to the fragment structure, and a fragment structure cluster creating step of creating a fragment structure cluster based on the similarity of the fragment structure; and the fragment sequence and sequence around the fragment sequence in the sequence space.
- Confidence matrix creation step for creating a query sequence, and query sequence input step for the user to enter a query sequence A query fragment sequence creating step of dividing the query sequence input in the query sequence input step into a predetermined length to create a query fragment sequence; and the query created in the query fragment sequence creating step.
- a query fragment sequence similarity search step for performing a sequence similarity search with the fragment sequence, the certainty matrix created in the certainty matrix creation step, and a query fragment sequence similarity search step
- a fragment sequence structure predicting step of predicting the fragment structure based on the sequence information and three-dimensional structure information of the protein, a fragment sequence obtained by dividing the sequence information into a predetermined length and a fragment structure corresponding to the fragment sequence are created, and based on the similarity between the fragment structures.
- a fragment structure cluster is created by performing a sequence similarity search on the fragment sequence with the surrounding fragment sequences in the sequence space, and the confidence that the probability that the similar sequence of the fragment sequence belongs to the fragment structure cluster is calculated using the fragment sequence and the structure.
- the user inputs a query sequence, divides the input query sequence into predetermined lengths to create a query fragment sequence, and performs a sequence similarity search with the fragment sequence for the created query fragment sequence.
- the probability that the query fragment sequence belongs to the fragment structure cluster is calculated based on the created confidence matrix and the search result, and the fragment structure of the query sequence is predicted based on the calculated probability.
- the correlation of the substructure from the subsequence and predict the substructure so that various manifolds and the certainty of the correlation can be expressed. That is, according to the present invention, when calculating the structure, the probabilities (confidence) of a plurality of structure candidates are given and given according to the degree of correlation (function of the confidence is used as the probability of structural change). it can.
- this method first creates a cluster of partial structures, takes into account the complex shape of the structural sequence correlation polymorphism, and queries ( query) dynamically after an array is given Sequence correlation clusters can be created.
- the method also creates structural clusters from different perspectives (eg, fragment sequence length, fragment structure resolution, structural cluster size, degree of correlation, etc.) and shoes. It will be possible to calculate the structure by integrating the prediction results.
- the protein structure predicting method according to the next invention is the protein structure predicting method according to the above, wherein the similarity search is performed on the fragment sequence in the fragment sequence similarity search step, and the similarity is represented by a matrix of fragment sequences.
- the method further includes a structure cluster information matrix creation step of creating a matrix, wherein the confidence matrix creation step includes the similarity matrix created in the similarity matrix creation step and the structure cluster information matrix creation step.
- the created structure class Based on the terpolymer information matrix, characterized in that to create the confidence the matrix.
- the similarity search is performed on the fragment sequences, and a similarity matrix creating step of creating a similarity matrix represented by a matrix of fragment sequences, and determining which fragment structure cluster the fragment sequence belongs to,
- a structural cluster-information matrix is created by displaying the structural cluster information shown in the form of a fragment array and a matrix of structural clusters.
- the confidence matrix is calculated. Since it is created, it is possible to easily and precisely calculate the certainty factor based on the similarity search result using a matrix operation method.
- the protein structure prediction method according to the next invention is the protein structure prediction method according to the above, wherein the entire structure optimization is performed by performing a predetermined optimization on the initial overall structure determined by the fragment structure having the maximum confidence. Characterized in that the method further comprises a You.
- a predetermined optimization is performed on the initial overall structure determined by the fragment structure having the maximum certainty.
- the initial structure is divided into various possible fragment sequences. Will be able to integrate the best prediction results. Further, by further optimizing the initial structure, the accuracy of the overall structure prediction can be further improved.
- the present invention also relates to a program.
- the program according to the present invention comprises a fragment sequence obtained by dividing sequence information into a predetermined length based on protein sequence information and three-dimensional structure information, and a fragment structure corresponding thereto.
- Gender search step and a certainty matrix that creates a certainty matrix that displays the certainty factor indicating the probability that the similar sequence of the fragment sequence belongs to the fragment structure cluster in a matrix of the fragment sequence and the first structural cluster
- a creating step a query sequence input step for allowing a user to input a query sequence, and the above query sequence input step.
- a query fragment array creating step for creating the query fragment array by dividing the query sequence input in the step into a predetermined length; and for the query fragment sequence created in the query fragment array creating step, A query fragment sequence similarity search step for performing a sequence similarity search with the fragment sequence, the certainty matrix created in the certainty matrix creation step, and a search result of the query fragment sequence similarity search step
- a fragment structure probability calculating step of calculating a probability that the query fragment sequence belongs to the fragment structure cluster based on the above, based on the probability calculated in the fragment structure probability calculating step, Causing the computer to execute a protein structure prediction program including a fragment sequence structure prediction step for predicting a fragment structure. It is characterized in.
- a fragment sequence obtained by dividing the sequence information into a predetermined length and a fragment structure corresponding thereto are created, A fragment structure cluster is created based on the similarity of the fragment structure, a sequence similarity search is performed with respect to the fragment sequence in the sequence space with respect to the fragment sequence, and the similarity of the fragment sequence belongs to the fragment structure cluster with a probability.
- the probability that the query fragment sequence belongs to the fragment structure cluster is calculated based on the created confidence matrix and the search result, and the fragment structure of the query sequence is predicted based on the calculated probability. It is possible to calculate the correlation of substructures from subsequences and predict the substructures so that complex manifolds and the certainty of correlations can be expressed. That is, according to the present invention, when calculating a structure, the probabilities (confidence) of a plurality of structure candidates are given and given according to the degree of correlation (a function of the confidence is used as the probability of structural change). it can.
- this program first creates a cluster of partial structures and takes into account the complex shape of the structural sequence correlated manifold. ) After the sequence is given, a sequence correlation cluster can be created dynamically.
- the program also creates a number of structural clusters from different perspectives (eg, fragment sequence length, fragment structure resolution, structural cluster size, degree of correlation, etc.), and the structure prediction results from each data set Can be integrated to calculate the structure.
- perspectives eg, fragment sequence length, fragment structure resolution, structural cluster size, degree of correlation, etc.
- a program according to the next invention is the program according to the above, wherein a similarity matrix is created by displaying a result of similarity search on the fragment sequence in the fragment sequence similarity search step as a matrix of fragment sequences.
- Structural cluster information for creating a structural cluster information matrix in which the property matrix creating step and structural cluster information indicating which fragment structural cluster the fragment sequence belongs to are displayed in a matrix of the fragment sequence and the structural cluster.
- Matrix work And the similarity matrix created in the similarity matrix creation step; and the structure cluster one information matrix created in the structure cluster one information matrix creation step. It is characterized in that the certainty matrix is created based on
- the program according to the next invention is the program according to the above, further comprising an overall structure optimization step of performing a predetermined optimization on the initial overall structure determined by the fragment structure having the maximum confidence.
- a predetermined optimization is performed on the initial overall structure determined by the fragment structure having the maximum certainty.
- the initial structure is created, it is divided into various possible fragment sequences. It will be possible to integrate those optimal prediction results. Further, by further optimizing the initial structure, the accuracy of the overall structure prediction can be further improved.
- the present invention relates to a recording medium, and the recording medium according to the present invention is characterized by recording the program described above.
- the program described above can be realized using a computer by causing a computer to read and execute the program recorded on the recording medium. Similar effects can be obtained.
- FIG. 1 is a diagram showing an example of a case where an array is represented by a profile according to the prior art
- FIG. 2 is a diagram showing a structural cluster one creation image according to the prior art
- FIG. FIG. 4 is a conceptual diagram showing a basic principle
- FIG. 4 is a block diagram showing an example of a configuration of the present system to which the present invention is applied
- FIG. 5 is a fragment structure prediction of the present system in the present embodiment.
- FIG. 6 is a flowchart showing an example of the processing.
- FIG. 6 shows an example in which the fragment structure cluster creating unit 102a acquires a fragment sequence and its corresponding fragment structure from the protein structure database 106a.
- FIG. 7 is a diagram showing an example of a fragment structure cluster of the fragment sequence created by the fragment structure cluster creating section 102a
- FIG. 8 is a diagram showing a hierarchical cluster method using the hierarchical cluster method.
- Fragment structural cluster FIG. 9 is a diagram showing an example of the case of creating a fragment sequence.
- FIG. 9 shows fragment sequence A, similar fragment sequences (D, F, G, S, I, etc.) and similarity score (50, 30). , 28, 25, 20 etc.) and a fragment structure cluster ( ⁇ , HI, J3, ⁇ , ⁇ , etc.) to which the fragment sequence belongs are searched.
- FIG. 11 is a diagram illustrating an example of information stored in a similarity matrix 106 b.
- FIG. 11 is a diagram illustrating an example of information stored in a structure cluster information matrix 106 c.
- Figure 2 is a conceptual diagram showing that the confidence matrix creation unit 102 e creates a confidence matrix 106 d based on the similarity matrix 106 b and the structural cluster-information matrix 106 c.
- Fig. 13 shows a similarity search for the query sequence (query fragment sequence) X and the search
- FIG. 14 is a conceptual diagram showing an example of calculating the probability of a fragment structure belonging to the result by multiplying the confidence matrix 106 d by the result.
- FIG. 14 shows the fragment structure prediction by the fragment sequence structure prediction unit 102 j.
- FIG. 15 is a conceptual diagram showing an example of the present invention.
- FIG. 15 is a conceptual diagram showing an example of the present invention.
- FIG. 15 is a flowchart showing an example of an overall structure prediction process of the present system in the present embodiment.
- BEST MODE FOR CARRYING OUT THE INVENTION embodiments of a protein structure prediction device, a protein structure prediction method, a program, and a recording medium according to the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited by the embodiment.
- FIG. 3 is a conceptual diagram showing the basic principle of the present invention.
- the present invention generally has the following basic features. That is, the present invention proposes a new calculation method of the correlation from the partial sequence to the partial structure, which can express a complex variety of the correlation and calculate the degree of the correlation (certainty).
- the present invention first creates structural clusters of various sizes from various data sets, and extracts sequence similarity data from among them. Then, after a query sequence is given by the user, a pseudo-dynamic array-to-structure correlation cluster is used by using structural clusters of various sizes for various subsequence divisions. Is calculated, and the magnitude of the correlation with the subsequence is calculated. The partial structure is predicted from the correlation cluster.
- the present invention classifies the structure of sequence fragments. That is, typical fragment structures are extracted based on sequence information and structure information stored in a known protein structure database or the like, and those fragment structures are classified.
- Fig. 3 (a) the structure of the periphery of a fragment sequence in the sequence space is examined.
- Fig. 3 (b) creating a virtual cluster between the sequence and the structure by examining what typical structure is obtained around each fragment sequence Can be. That is, according to the present invention, the sequence similar to this sequence existing around a certain sequence A belongs to which structural cluster in the structural space. ?) And create a virtual cluster around this array.
- this fragment when a certain unknown structural sequence fragment X is given, this fragment is similar to the sequence A. It obtains information such as the sequence, similarity to sequence C, etc., and combines virtual clusters based on this information to predict which structural cluster the sequence will eventually belong to.
- the prediction of the overall structure of the present invention is performed in the following procedure.
- the degree of correlation strength (confidence) is compared from the obtained partial structure candidates, and a partial structure with a strong correlation and a long partial sequence is used.
- Partial structures with weak correlations are also stored as data together with probabilities. Then, using the stored data, this is used as a candidate for the next structure, and the structure is changed to perform a folding 'simulation (fol d in g s im u l a t i o n).
- the structure is refined (optimized) for all atomic systems.
- FIG. 4 is a block diagram showing an example of a configuration of the present system to which the present invention is applied, and conceptually shows only a part related to the present invention in the configuration.
- This system is composed of a network 300, which comprises a protein structure prediction device 100 and an external system 200 that provides an external database for protein structure information, etc. It is configured to be communicably connected via a PC.
- the network 300 has a function of interconnecting the protein structure prediction device 100 and the external system 200, and is, for example, the Internet.
- the external system 200 is interconnected with the protein structure predicting device 100 via the network 300, and provides the user with an external database for protein structure information and the like. It has a function to provide a website to execute an external analysis program such as.
- the external system 200 may be configured as a WEB server, an ASP server, or the like, and its hardware configuration is generally a commercially available workstation, NO. It may be constituted by an information processing device such as a sonar computer and its attached devices.
- each function of the external system 200 It is realized by a CPU, a disk device, a memory device, an input device, an output device, a communication control device and the like in the configuration and a program for controlling them.
- the protein structure predicting device 100 is generally composed of a control unit 102 such as a CPU that controls the entire protein structure predicting device 100 as a whole, a router connected to a communication line, and the like.
- a communication control interface 104 connected to a communication device (not shown), an input / output control interface 108 connected to an input device 112 and an output device 114, and
- the system is provided with a storage unit 106 for storing various databases and tables (protein structure database 106a to certainty matrix 106d), and these units are connected to arbitrary communication channels.
- the protein structure prediction device 100 is communicably connected to a network 300 via a communication device such as a router and a wired or wireless communication line such as a dedicated line.
- protein structure database 106a to certainty matrix 106d stored in the storage unit 106 are storage means such as a fixed disk device, and are used for various processes. It stores programs, tables, files, and files for database web pages.
- the protein structure database 106a is a database storing protein structure information recorded in association with amino acid sequence information (primary structure) and three-dimensional structure information. It is. It is preferable that the protein structure database 106a excludes sequence redundancy.
- the protein structure database 106a may be an external protein structure database (for example, PDB-SELECT, etc.) accessed via the Internet, or may be a copy of these databases or an original database. It may be an in-house database created by storing the protein structure of the original or by adding unique annotation information and the like.
- the similarity matrix 106 b is a matrix table that stores information on similarity search results regarding fragment sequences and the like.
- the structure cluster information matrix 106 c is a matrix table for storing information indicating which fragment structure cluster the fragment sequence belongs to.
- the confidence matrix 106 d stores information indicating the confidence (probability) that a fragment sequence belongs to the fragment structure when information that a fragment sequence is similar to another fragment sequence is obtained. It is a matrix table to perform.
- control unit 102 has a control program such as an OS (Operating System), a program defining various processing procedures, and the like, and an internal memory for storing required data. With these programs and the like, information processing for executing various processes is performed.
- OS Operating System
- program defining various processing procedures and the like
- internal memory for storing required data.
- the control unit 102 is functionally conceptually composed of a fragment structure cluster creation unit 102 a, a fragment sequence similarity search unit 102 b, a similarity matrix creation unit 102 c, and a structure cluster information matrix creation Part 102 d, confidence matrix creation part 102 e, query sequence input part 102 f, query fragment sequence creation part 102 g, query fragment sequence similarity search part 102 h, fragment structure It comprises a probability calculation unit 102 i, a fragment sequence structure prediction unit 102 j, and an overall structure optimization unit 102 k.
- the fragment structure cluster creating unit 102a calculates the fragment sequence obtained by dividing the sequence information into a predetermined length and the fragment structure corresponding thereto.
- This is a fragment structure cluster creating means for creating a fragment structure cluster based on the similarity of the fragment structures.
- the fragment sequence similarity search unit 102b is a fragment sequence similarity search unit that performs a sequence similarity search on a fragment sequence with surrounding fragment sequences in the sequence space.
- the similarity matrix creation unit 102 c creates a similarity matrix that creates a similarity matrix that displays the results of similarity search for fragment sequences by a fragment sequence similarity search unit as a matrix of fragment sequences. Means.
- the structural cluster one information matrix creating unit 102 d generates a structural cluster one information matrix in which the structural cluster information indicating which fragment structural cluster the fragment sequence belongs to is displayed by the fragment array and the matrix of the structural cluster. To create This is a means for creating a cluster information matrix.
- the confidence matrix creation unit 102 e creates a confidence matrix that displays the confidence, which is the probability that a similar sequence of the fragment sequence belongs to the fragment structure cluster, as a matrix of the fragment sequence and the structure cluster. This is a certainty matrix creation means.
- the query sequence input unit 102 f is a query sequence input unit that allows a user to input a query sequence.
- the query fragment sequence creation unit 102g is a query fragment sequence creation unit that divides the query sequence input by the query sequence input unit into predetermined lengths to create a query fragment sequence.
- the query fragment sequence similarity search unit 102h is a query fragment sequence similarity search unit that performs a sequence similarity search with the fragment sequence for the query fragment sequence created by the query fragment sequence creation unit.
- the fragment structure probability calculation unit 102 i generates a query fragment sequence based on the confidence matrix created by the confidence matrix creation unit and the search result of the query fragment sequence similarity search unit. Is a fragment structure probability calculating means for calculating a probability belonging to.
- the fragment sequence structure prediction unit 102j is a fragment sequence structure prediction unit that predicts the fragment structure of the query sequence based on the probability calculated by the fragment structure probability calculation unit.
- the whole structure optimization unit 102 k is a whole structure optimization unit that performs a predetermined optimization on the initial whole structure determined by the fragment structure having the maximum certainty factor. The details of the processing performed by these units will be described later.
- FIG. 5 is a flowchart showing an example of a fragment structure prediction process of the present system in the present embodiment.
- the protein structure prediction device 100 accesses the protein structure database 106a by the processing of the fragment structure cluster creation unit 102a, and obtains protein sequence information. (For example, amino acid sequence information, etc.) and three-dimensional structure information are obtained, and a fragment sequence obtained by dividing the sequence information into a predetermined length and a fragment structure corresponding thereto are generated (step S A-1).
- FIG. 6 is a conceptual diagram showing an example in which the fragment structure cluster creating section 102a acquires a fragment sequence and its corresponding fragment structure from the protein structure database 106a.
- the fragment structure cluster creating section 102a divides the sequence for each fragment sequence of a predetermined length (in FIG. 6, 7 amino acid residues) and takes the fragment sequence. It is stored in the storage unit 106 in association with the fragment structure.
- the length of the fragment is not limited to 7 residues, and the fragment structure can be divided into various lengths.
- FIG. 7 is a diagram showing an example of a fragment structure cluster of the fragment sequence created by the fragment structure cluster creating unit 102a.
- the fragment structure cluster creation unit 102a uses a self-organizing map (SOM; se 1 forganized map) using the similarity of fragment structures (for example, rmsd or dme) as an index of similarity. ), k-means (k-means), clustering using known clustering methods such as hierarchical clustering.
- SOM self-organizing map
- k-means k-means
- FIG. 8 is a diagram showing an example of a case where a fragment structure cluster is created by using a hierarchical cluster method.
- the fragment structure cluster creation unit 102a performs clustering by calculating the distances of all fragment structures and sequentially grouping the closest distances.
- the distance between the clusters is calculated by, for example, calculating the distances of all members belonging to each cluster and taking an average.
- the protein structure prediction apparatus 100 performs similarity of the existing sequence such as b1 ast search with the surrounding fragment sequences in the sequence space for all the fragment sequences by the processing of the fragment sequence similarity search section 102b.
- a similar fragment sequence, a similarity score, and a fragment structure cluster to which the fragment sequence belongs are obtained by a gender search technique (step SA-3).
- Fig. 9 shows the similarity of fragment sequence A (D, F, G, S, I, etc.) and similarity score (50, 30, 28, 25, 20).
- a fragment structure cluster ( ⁇ , / 3, ⁇ , ⁇ , etc.) to which the fragment sequence belongs.
- the protein structure prediction device 100 uses the similarity matrix creation unit 102 c to perform a similarity search on the fragment sequence, and displays a similarity 1 ”raw matrix as a fragment sequence matrix.
- Create 106 b (step SA-4)
- Fig. 10 is a diagram showing an example of information stored in the similarity matrix 106 b.
- the similarity matrix 106 b the result of performing a similarity search on each fragment sequence is stored.
- the protein structure predicting apparatus 100 creates a structural cluster information matrix 106 c indicating to which fragment structural cluster the fragment sequence belongs by the processing of the structural cluster information matrix creating unit 102 d.
- FIG. 11 is a diagram showing an example of information stored in the structural cluster information matrix 106c. As shown in FIG. 11, the structure cluster information “1” is set in the fragment structure cluster to which the fragment sequence belongs.
- the protein structure prediction apparatus 100 when the information that a certain fragment sequence is similar to another fragment sequence is obtained by the processing of the confidence matrix creating unit 102e, the fragment sequence is A confidence matrix 106 d indicating the confidence that is the probability of belonging to the structural cluster of another fragment sequence is created (step SA-6).
- FIG. 12 shows that the confidence matrix creation unit 102 e creates a confidence matrix 106 d based on the similarity matrix 106 b and the structural cluster information matrix 106 c.
- the confidence matrix creation unit 102 e generates the confidence matrix by taking the product of the normalized similarity matrix 106 b and the structural cluster information matrix 106 c. Create 1 06 d.
- the protein structure prediction device 100 executes the processing of the query sequence input unit 102 f. Then, the user inputs a query sequence (step SA-7). This sequence may be input by allowing the user to select a desired sequence from a database storing external amino acid sequences, or the user may directly input the desired sequence.
- the protein structure prediction apparatus 100 divides the query sequence into fragment sequences of a predetermined length (for example, 7 amino acid residues) by the processing of the query fragment sequence creating unit 102 g, and divides the fragment sequence (query fragment sequence). It is stored in the storage unit 106 (step SA-8).
- the length of the fragment is not limited to 7 residues, and the fragment structure may be divided into various lengths.
- the protein structure prediction apparatus 100 searches the sequence similarity for each fragment sequence (query fragment sequence) of the query sequence by the processing of the query fragment sequence similarity search unit 102h (step SA-9), and the search result Based on the above, the probability of the fragment structure to which the fragment sequence belongs is calculated by the processing of the fragment structure probability calculating unit 102i (step SA-10).
- FIG. 13 is a concept showing an example of performing a similarity search on a query sequence (query fragment sequence) X and multiplying the search result by a certainty matrix 106 d to calculate the probability of a fragment structure to which the query belongs.
- FIG. As shown in Fig. 13, by multiplying the standardized similarity vector of the query sequence X by the confidence matrix 106 d, the probability that the query sequence X belongs to each fragment structure cluster (the confidence ) Can be calculated.
- FIG. 14 is a conceptual diagram showing an example of fragment structure prediction by the fragment sequence structure prediction section 102 j.
- the fragment sequence structure predicting unit 102 j sorts the query sequence X into the fragment structure by sorting according to the reliability of the structural gaster to which the similar sequence of the query sequence X belongs. Anticipate. This completes the fragment structure prediction processing.
- FIG. 15 is a flowchart showing an example of the overall structure prediction processing of the present system in the present embodiment.
- the user inputs a query (query) array (step SB-1).
- the protein structure predicting apparatus 100 divides the query (query) sequence into fragment sequences of a predetermined length by the processing of the query fragment sequence creating unit 102g (Step SB-2).
- fragment sequences of multiple patterns divided by different lengths are created (two patterns are created in Fig. 15).
- the protein structure prediction device 100 predicts the fragment structure by the above-described method (step SB-3).
- the protein structure prediction device 100 creates an initial overall structure from the fragment structure having the maximum certainty factor by the processing of the fragment sequence structure prediction unit 102j (step SB-4).
- the protein structure prediction device 100 optimizes the entire structure using the statistical potential method, the MC method, and the simulated annealing (SA) by the processing of the entire structure optimization unit 102k. (Step SB-5).
- SA simulated annealing
- the fragment structure is randomly selected from the predicted fragment structures. , Calculate the energy value (E ne ;) of the confidence factor (P ne ;) after the replacement, and calculate the probability / 0 that the fragment structure after the replacement is adopted in the next step,
- the protein structure prediction apparatus 100 performs the processing in a stand-alone form has been described as an example.However, in response to a request from a client terminal configured in a separate housing from the protein structure prediction apparatus 100, It may be configured to perform processing according to the request and return the processing result to the client terminal.
- all or some of the processes described as being performed automatically can be manually performed, or the processes described as being performed manually can be performed. All or a part of the processing can be automatically performed by a known method.
- each processing function performed by the control unit the whole or any part thereof is described by C It can be implemented by a PU (Central Processing Unit) and a program interpreted and executed by the CPU, or it can be implemented as hardware by wire-and-logic.
- the program is recorded on a recording medium described later, and is mechanically read by the protein structure prediction device 100 as necessary. That is, a computer program for giving instructions to the CPU and performing various processes in cooperation with an OS (operating system) is recorded in the storage unit 106 such as a ROM or an HD.
- the computer program is executed by being loaded into the RAM, and forms a control unit in cooperation with the CPU.
- this computer program may be recorded on an application program server connected to the protein structure prediction device 100 via an arbitrary network, and download all or part of the computer program as needed. This is also possible.
- the program according to the present invention can be stored in a computer-readable recording medium.
- this “recording medium” refers to any “portable physical medium” such as a flexible disk, a magneto-optical disk, ROM, EPROM, EEPROM, CD-ROM, MO, DVD, etc., and various computer systems.
- any ⁇ fixed physical medium '' such as built-in ROM, RAM, HD, etc., or a communication line or carrier wave when transmitting a program via a network represented by LAN, WAN, Internet, etc.
- a “program” is a data processing method described in any language or description method, regardless of the format of source code or binary code. Note that the “program” is not necessarily limited to a single program, but may be distributed with multiple module libraries or a separate program typified by an operating system (OS). Including those that work to achieve that function. Note that a specific structure for reading a recording medium in each device described in the embodiment is described. Known configurations and procedures can be used for the configuration, reading procedure, or installation procedure after reading.
- the network 300 has a function of interconnecting the protein structure prediction device 100 and the external system 200, and includes, for example, the Internet, an intranet, a LAN (including both wired and wireless), a VAN, and a personal computer.
- Telecommunications networks public telephone networks (both analog and digital), leased line networks (both analog and digital), CATV networks, and mobile lines such as IMT2000, GSM, or PDCZPDC-P Switching network
- a portable bucket switching network a radio paging network, a local radio network such as B 1 uetooth, a PHS network, and a satellite communication network such as CS, BS or IS DB may be included. That is, the present system can transmit and receive various data via any network regardless of whether it is wired or wireless.
- the protein structure database 106a to the certainty matrix 106d include memory devices such as RAM and ROM, fixed disk devices such as a node disk, flexible disks, optical disks and the like. It is a storage means and stores various programs and tables, files, data bases, files for web pages, etc. used for various processes and websites.
- the protein structure prediction device 100 connects a peripheral device such as a printer monitor or an image scanner to an information processing device such as an information processing terminal such as a known personal computer or workstation, and connects the information processing device of the present invention to the information processing device. It may be realized by implementing software (including programs, data, etc.) that realizes the above.
- each database may be independently configured as an independent database device, and a part of the processing may be realized by a CGI (Comm on Gateway Interlace). You may.
- CGI Common Gateway Interlace
- a fragment sequence obtained by dividing sequence information into a predetermined length based on protein sequence information and three-dimensional structure information and a fragment structure corresponding thereto are obtained.
- a fragment structure cluster is created based on the similarity of the fragment structure, and a sequence similarity search is performed on the fragment sequence with the surrounding fragment sequences in the sequence space, and the similar sequence of the fragment sequence is converted into the fragment structure cluster.
- the user inputs a query sequence to the IJ user, divides the input query sequence into predetermined lengths, creates a query fragment sequence, and searches the fragment sequence and sequence similarity for the created query fragment sequence.
- the probability that the query fragment sequence belongs to the fragment structure cluster is calculated based on the created confidence matrix and the search result, and the fragment structure of the query sequence is predicted based on the calculated probability. It is possible to calculate the correlation of substructures from subsequences and predict the substructures so that the complex manifolds and correlations can be expressed. That is, according to the present invention, when calculating the structure, the probabilities (confidence) of a plurality of structure candidates are given and given according to the degree of correlation (function of the confidence is used as the probability of structural change).
- a protein structure prediction device, a protein structure prediction method, a program, and a recording medium that can be provided.
- the present apparatus first forms a cluster of partial structures and takes into account the complex shape of the structure sequence correlation polymorphism and queries ( query) It is possible to provide a protein structure prediction device, a protein structure prediction method, a program, and a recording medium that can dynamically form a sequence correlation cluster after a sequence is given.
- a number of structural clusters are created from different viewpoints (eg, fragment sequence length, fragment structure resolution, structural cluster size, degree of correlation, etc.), and the structure from each data set is created.
- a recording medium can be provided.
- similarity matrix creating means for creating a similarity matrix in which a result of similarity search for a fragment sequence is displayed as a matrix of fragment sequences, and to which fragment structure cluster the fragment sequence belongs
- a structural cluster information matrix is created by displaying the structural cluster information indicating the above as a matrix of fragment sequences and structural clusters, and a confidence matrix is created based on the created similarity matrix and structural cluster one information matrix. It is possible to provide a protein structure prediction device, a protein structure prediction method, a program, and a recording medium that can easily and precisely calculate a certainty factor based on similarity search results using a matrix calculation method.
- a predetermined optimization is performed on the initial overall structure determined by the fragment structure having the maximum certainty factor. It becomes possible to divide and integrate the optimal prediction results. Further, by further optimizing the initial structure, it is possible to provide a protein structure prediction device, a protein structure prediction method, a program, and a recording medium that can further improve the accuracy of the overall structure prediction.
- the protein structure predicting apparatus, the protein structure predicting method, the program, and the recording medium according to the present invention provide a three-dimensional structure prediction of a protein, analysis of a site for mutual use of a protein, and creation using the analysis result. It can be used for medicines and the like.
Landscapes
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- Biotechnology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Chemical & Material Sciences (AREA)
- Data Mining & Analysis (AREA)
- Crystallography & Structural Chemistry (AREA)
- Proteomics, Peptides & Aminoacids (AREA)
- Analytical Chemistry (AREA)
- Evolutionary Computation (AREA)
- Databases & Information Systems (AREA)
- Artificial Intelligence (AREA)
- Epidemiology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Bioethics (AREA)
- Public Health (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Investigating Or Analysing Biological Materials (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP02792060A EP1460559A4 (en) | 2001-12-27 | 2002-12-27 | DEVICE FOR PREDICTING A PROTEIN STRUCTURE, METHOD FOR PREDICTING A PROTEIN STRUCTURE, PROGRAM AND RECORDING MEDIUM |
US10/846,622 US20050026217A1 (en) | 2001-12-27 | 2004-05-17 | Protein structure prediction device, protein structure prediction method, program, and recording medium |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2001-398569 | 2001-12-27 | ||
JP2001398569A JP4084040B2 (ja) | 2001-12-27 | 2001-12-27 | 蛋白質構造予測装置、蛋白質構造予測方法、プログラム、および、記録媒体 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US10/846,622 Continuation-In-Part US20050026217A1 (en) | 2001-12-27 | 2004-05-17 | Protein structure prediction device, protein structure prediction method, program, and recording medium |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2003056461A1 true WO2003056461A1 (fr) | 2003-07-10 |
Family
ID=19189362
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2002/013832 WO2003056461A1 (fr) | 2001-12-27 | 2002-12-27 | Dispositif de prediction de structure proteique, procede de prediction de structure proteique, programme et support d'enregistrement associes |
Country Status (4)
Country | Link |
---|---|
US (1) | US20050026217A1 (ja) |
EP (1) | EP1460559A4 (ja) |
JP (1) | JP4084040B2 (ja) |
WO (1) | WO2003056461A1 (ja) |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1796009A3 (en) * | 2005-12-08 | 2007-08-22 | Electronics and Telecommunications Research Institute | System for and method of extracting and clustering information |
JP5732697B2 (ja) * | 2010-07-02 | 2015-06-10 | Necソリューションイノベータ株式会社 | 予測装置、予測方法、プログラムおよび記録媒体 |
CN104200130B (zh) * | 2014-07-23 | 2017-08-11 | 浙江工业大学 | 一种基于树结构副本交换和片段组装的蛋白质结构预测方法 |
JP2017037377A (ja) * | 2015-08-07 | 2017-02-16 | 富士通株式会社 | 情報処理装置、シミュレーション方法、およびシミュレーションプログラム |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3012441B2 (ja) * | 1993-10-01 | 2000-02-21 | 日本電気株式会社 | タンパク質立体構造予測方法 |
JP3585613B2 (ja) * | 1995-12-08 | 2004-11-04 | 富士通株式会社 | 蛋白質の二次構造予測方法及び装置 |
-
2001
- 2001-12-27 JP JP2001398569A patent/JP4084040B2/ja not_active Expired - Fee Related
-
2002
- 2002-12-27 WO PCT/JP2002/013832 patent/WO2003056461A1/ja not_active Application Discontinuation
- 2002-12-27 EP EP02792060A patent/EP1460559A4/en not_active Withdrawn
-
2004
- 2004-05-17 US US10/846,622 patent/US20050026217A1/en not_active Abandoned
Non-Patent Citations (1)
Title |
---|
YI T.-M. ET AL.: "Protein secondary structure prediction using nearest-neighbor methods", JOURNAL OF MOLECULAR BIOLOGY, vol. 232, no. 4, 1993, pages 1117 - 1129, XP002965206 * |
Also Published As
Publication number | Publication date |
---|---|
JP4084040B2 (ja) | 2008-04-30 |
JP2003196575A (ja) | 2003-07-11 |
EP1460559A4 (en) | 2007-01-24 |
US20050026217A1 (en) | 2005-02-03 |
EP1460559A1 (en) | 2004-09-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Pagès et al. | Protein model quality assessment using 3D oriented convolutional neural networks | |
Mulder et al. | The InterPro database and tools for protein domain analysis | |
CA2700033A1 (en) | Method, apparatus and computer program product for performing a visual search using grid-based feature organization | |
Selvitopi et al. | Distributed many-to-many protein sequence alignment using sparse matrices | |
JP2002245048A (ja) | 画像検索方法および画像検索装置 | |
JP2007334402A (ja) | クラスタリングされたベクトルデータを検索するサーバ、システム及び方法 | |
CA2942106A1 (en) | Aligning and clustering sequence patterns to reveal classificatory functionality of sequences | |
Ellis et al. | diBELLA: Distributed long read to long read alignment | |
WO2008156773A1 (en) | Biological database index and query searching | |
JPWO2003001409A1 (ja) | 構造化データ処理装置 | |
CN111274455A (zh) | 图数据处理方法、装置、电子设备及计算机可读介质 | |
WO2003056461A1 (fr) | Dispositif de prediction de structure proteique, procede de prediction de structure proteique, programme et support d'enregistrement associes | |
JP2003330934A (ja) | 変異体配列解析装置、変異体配列解析方法、プログラム、および、記録媒体 | |
Vaddadi et al. | Read mapping on genome variation graphs | |
Woo et al. | MONACO: accurate biological network alignment through optimal neighborhood matching between focal nodes | |
EP1471444A1 (en) | Rna sequence analyzer, and rna sequence analysis method, program and recording medium | |
EP1460560A1 (en) | Apparatus for predicting interaction site, method of predicting interaction site, program and recording medium | |
JP2003242154A (ja) | 遺伝子発現情報管理装置、遺伝子発現情報管理方法、プログラム、および、記録媒体 | |
KR100478792B1 (ko) | 2차원 젤 이미지를 이용한 유사 단백질 검색 장치 및 방법 | |
JP4054574B2 (ja) | 相互作用部位予測装置、相互作用部位予測方法、プログラム、および、記録媒体 | |
JP4250443B2 (ja) | 抗体設計装置、抗体設計方法、プログラム、および、記録媒体 | |
Coatney et al. | Motifminer: A general toolkit for efficiently identifying common substructures in molecules | |
Plewczynski et al. | Molecular modeling of phosphorylation sites in proteins using a database of local structure segments | |
Tapinos et al. | De novo assembly of nucleotide sequences in a compressed feature space | |
JP2004109053A (ja) | 結合部位予測方法、結合部位予測装置、プログラム、および、記録媒体 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AK | Designated states |
Kind code of ref document: A1 Designated state(s): US |
|
AL | Designated countries for regional patents |
Kind code of ref document: A1 Designated state(s): AT BE BG CH CY CZ DE DK EE ES FI FR GB GR IE IT LU MC NL PT SE SI SK TR |
|
DFPE | Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101) | ||
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2002792060 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 10846622 Country of ref document: US |
|
WWP | Wipo information: published in national office |
Ref document number: 2002792060 Country of ref document: EP |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: 2002792060 Country of ref document: EP |