Automatic construction method of metabolite pathway expansion network based on structural features

Abstract

The present application relates to a kind of metabolic pathway extension network based on structural feature automatical establishment method, belong to bioinformatics and computational chemistry technical field.The purpose is to solve the problem of existing KEGG database incomplete, annotation lag of metabolite.Method includes: obtaining KEGG skeleton and modified group construction seed library;Through substructure matching and set difference calculation to determine the attribution of derivative;Iterative expansion forms structure association set;Combining stereochemistry rule and mass difference verification screening derivative;Merging KEGG reaction and extended relationship constructs network, and remove redundant path by graph theory method, finally output extended pathway network.The method greatly improves the annotation rate of metabolite pathway, supports in-depth functional analysis of metabolomics data.

Description

Technical Field

[0001] This invention belongs to the fields of bioinformatics and computational chemistry, and involves metabolomics, cheminformatics, pathway analysis and graph network modeling. Specifically, it relates to an automated method for establishing extended metabolite pathway networks based on structural features. Background Technology

[0002] Currently, functional annotation of metabolomics data mainly relies on the KEGG pathway database. This method maps metabolites to a pre-defined static metabolic pathway network in KEGG to obtain their affiliation and interaction relationships.

[0003] The KEGG pathway database has a limited number of metabolites, which cannot meet the needs of large-scale metabolomics annotation. As a result, many metabolites are excluded from functional analysis due to missing matching information. The pathway maps in the KEGG database are artificially constructed static templates, and newly discovered metabolites and molecular relationships in the metabolomics cannot be dynamically included in the database, resulting in annotation results that do not reflect the latest research findings. Summary of the Invention

[0004] To address the aforementioned problems in the existing technology, this invention provides an automated method for establishing extended metabolite pathway networks based on structural features. The objective of this invention can be achieved through the following technical solutions: Automated methods for constructing extended metabolite pathway networks based on structural features include: S1: Obtain the basic scaffold structures of metabolites from the KEGG pathway database, characterize them using SMILES, and construct a seed structure library; obtain and integrate representative modification group structures from the literature, and construct a modification group library; obtain metabolite structure information from the reference structure database as a candidate derivative library; S2: Using a substructure matching algorithm, the candidate derivative structure is matched with the seed skeleton and modifying groups to obtain a set of matching atomic numbers; by calculating the set difference and combining it with the modification reaction characteristics, it is determined whether the candidate derivative is derived from a specific skeleton and groups. S3: Using the derivative obtained in the first round of judgment as the new seed structure, iteratively execute step S2 to perform multiple rounds of modification chain extension structure matching and judgment until no new derivatives are generated, forming an extended metabolite structure association set. S4: During the substructure matching process, chiral matching rules are dynamically set based on the stereochemical information of the skeleton and the derivative; further, the molecular weight difference rule library of modification type is used to verify the molecular weight difference between the derivative and the skeleton, and to screen derivatives that meet the quality consistency requirements. S5: Merge the original KEGG reaction pairs with the extended skeleton-derived pairs to construct a structural extension network; use graph theory to identify and remove redundant direct derivation pathways, retain indirect pathways with multiple derivation steps, and finally output a redundant metabolite pathway extension network.

[0005] Specifically, in step S2, the set of atomic numbers is obtained by executing the substructure search function of the substructure matching algorithm.

[0006] Specifically, in step S2, the determination further includes: calculating the difference between the set of atomic numbers of the candidate derivative and the set of atomic numbers matching the skeleton to obtain the set of atomic numbers of residues; traversing the set of atomic numbers matching the modifying group, if there is a set that has only one more atomic number than the set of atomic numbers of residues, identifying the element corresponding to the extra atomic number, and determining that the candidate derivative is derived from the skeleton and the modifying group through the atomic connection of the element.

[0007] Specifically, in step S3, the iterative execution is provided with a termination condition: when a round of expansion process ends and no new derivative structure that does not repeat the existing structure is generated, the iteration terminates.

[0008] Specifically, the process of dynamically setting the chiral matching rules in step S4 is as follows: if the skeleton structure contains stereochemical information, but the candidate derivative structure does not contain stereochemical information, then the substructure matching parameters are changed.

[0009] Specifically, the stereochemical information is characterized by chiral symbols in the SMILES string.

[0010] Specifically, in step S4, the self-built rule library for molecular weight difference of modification type pre-stores the precise molecular weight difference value corresponding to the modification group; the verification process includes: calculating the precise molecular weight difference between the derivative and the skeleton, comparing the difference value with the molecular weight difference value of the modification group pre-stored in the rule library; if the difference value is successfully matched within the preset error range, the verification is passed.

[0011] Specifically, in step S5, the identification and deletion of redundant direct derivative paths using graph theory methods includes: Traverse every direct edge in the network; temporarily remove a direct edge, and then determine whether there is another indirect path consisting of multiple edges between the two nodes it connects; If at least one of the indirect paths exists, the corresponding direct edge is deleted in the final network; otherwise, the corresponding direct edge is retained.

[0012] Specifically, the substructure matching algorithm is implemented using a cheminformatics toolkit, including but not limited to RDKit; the substructure search function is the GetSubstructMatches function or other functions with equivalent functionality.

[0013] Specifically, the graph theory method is a path redundancy elimination algorithm based on directed graphs; the structure extension network is constructed with metabolite structures as nodes and derivation relationships as directed edges, wherein the direction of the edges points from the modified skeleton structure to the derived structure.

[0014] The beneficial effects of this invention are as follows: This invention overcomes the limitations of traditional reliance on the static KEGG database through an automated substructure matching and iterative extension mechanism, significantly expanding the coverage of metabolite pathway networks. It integrates a large number of derivatives not included in KEGG into the network through structural derivation relationships, improving the completeness of functional annotation of metabolomics data. The combination of dynamic chiral matching rules and molecular weight difference verification ensures the accuracy and specificity of derivative identification, effectively reducing erroneous associations caused by stereochemical differences or mass mismatches. A graph-based redundant path elimination strategy optimizes the network structure, preserving more biologically meaningful multi-step derivation paths, making the extended network more concise and reflecting the true hierarchical derivation relationships between metabolites. The overall method achieves fully automated construction from seed structure to extended network, providing efficient and systematic tool support for the functional analysis of large-scale metabolomics data, and contributing to the discovery of new metabolic pathway associations and potential biomarkers. Attached Figure Description

[0015] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0016] Figure 1 This is a schematic diagram of the technical route of the present invention; Figure 2 This is a schematic diagram illustrating the substructure matching between the skeleton and groups and the derivatives in this invention; Figure 3 This is a schematic diagram illustrating the iterative expansion of metabolite structures in this invention; Figure 4 This is a schematic diagram of redundancy removal in the structural extension network of this invention; Figure 5 This is a schematic diagram of the KEGG structure extended network in this invention. Detailed Implementation

[0017] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this disclosure more comprehensive and complete, and to fully convey the concept of example embodiments to those skilled in the art. Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a full understanding of embodiments of this disclosure. However, those skilled in the art will recognize that the technical solutions of this disclosure can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring aspects of this disclosure. The blocks shown in the drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices. The flowcharts shown in the drawings are merely illustrative and do not necessarily include all contents and operations / steps, nor do they necessarily have to be performed in the order described. For example, some operations / steps can be broken down, while others can be combined or partially combined. Therefore, the actual execution order may change depending on the actual situation.

[0018] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided.

[0019] Please see Figures 1-5 An automated method for constructing extended metabolite pathway networks based on structural features includes: S1: Obtain the basic scaffold structures of metabolites from the KEGG pathway database, characterize them using SMILES, and construct a seed structure library; obtain and integrate representative modification group structures from the literature, and construct a modification group library; obtain metabolite structure information from the reference structure database as a candidate derivative library; S2: Using a substructure matching algorithm, the candidate derivative structure is matched with the seed skeleton and modifying groups to obtain a set of matching atomic numbers; by calculating the set difference and combining it with the modification reaction characteristics, it is determined whether the candidate derivative is derived from a specific skeleton and groups. S3: Using the derivative obtained in the first round of judgment as the new seed structure, iteratively execute step S2 to perform multiple rounds of modification chain extension structure matching and judgment until no new derivatives are generated, forming an extended metabolite structure association set. S4: During the substructure matching process, chiral matching rules are dynamically set based on the stereochemical information of the skeleton and the derivative; further, the molecular weight difference rule library of modification type is used to verify the molecular weight difference between the derivative and the skeleton, and to screen derivatives that meet the quality consistency requirements. S5: Merge the original KEGG reaction pairs with the extended skeleton-derived pairs to construct a structural extension network; use graph theory to identify and remove redundant direct derivation pathways, retain indirect pathways with multiple derivation steps, and finally output a redundant metabolite pathway extension network.

[0020] Specifically, in step S2, the set of atomic numbers is obtained by executing the substructure search function of the substructure matching algorithm.

[0021] Specifically, in step S2, the determination further includes: calculating the difference between the set of atomic numbers of the candidate derivative and the set of atomic numbers matching the skeleton to obtain the set of atomic numbers of residues; traversing the set of atomic numbers matching the modifying group, if there is a set that has only one more atomic number than the set of atomic numbers of residues, identifying the element corresponding to the extra atomic number, and determining that the candidate derivative is derived from the skeleton and the modifying group through the atomic connection of the element.

[0022] Specifically, in step S3, the iterative execution is provided with a termination condition: when a round of expansion process ends and no new derivative structure that does not repeat the existing structure is generated, the iteration terminates.

[0023] Specifically, the process of dynamically setting the chiral matching rules in step S4 is as follows: if the skeleton structure contains stereochemical information, but the candidate derivative structure does not contain stereochemical information, then the substructure matching parameters are changed.

[0024] Specifically, the stereochemical information is characterized by chiral symbols in the SMILES string.

[0025] Specifically, in step S4, the self-built rule library for molecular weight difference of modification type pre-stores the precise molecular weight difference value corresponding to the modification group; the verification process includes: calculating the precise molecular weight difference between the derivative and the skeleton, comparing the difference value with the molecular weight difference value of the modification group pre-stored in the rule library; if the difference value is successfully matched within the preset error range, the verification is passed.

[0026] Specifically, in step S5, the identification and deletion of redundant direct derivative paths using graph theory methods includes: Traverse every direct edge in the network; temporarily remove a direct edge, and then determine whether there is another indirect path consisting of multiple edges between the two nodes it connects; If at least one of the indirect paths exists, the corresponding direct edge is deleted in the final network; otherwise, the corresponding direct edge is retained.

[0027] Specifically, the substructure matching algorithm is implemented using a cheminformatics toolkit, including but not limited to RDKit; the substructure search function is the GetSubstructMatches function or other functions with equivalent functionality.

[0028] Specifically, the graph theory method is a path redundancy elimination algorithm based on directed graphs; the structure extension network is constructed with metabolite structures as nodes and derivation relationships as directed edges, wherein the direction of the edges points from the modified skeleton structure to the derived structure.

[0029] In this embodiment, the Substructure Searching function of RDKit (i.e., the GetSubstructMatches function) is used to obtain the structure matching results of the candidate derivative (derivative A) with the KEGG seed skeleton (skeleton B) and the candidate modifying group (group C) respectively (e.g. Figure 2 (As shown). The results are presented as sets of matching atom indexes, where AB matched atom index set represents the set of matching atom indexes between derivative A and skeleton B, AC matched atom index set represents the set of matching atom indexes between derivative A and group C, and A atom index set represents the set of atom indexes corresponding to the candidate derivative structure. Subsequently, it is determined whether derivative A consists of skeleton B and group C according to the following rules.

[0030] (1) Obtain the difference between the atom index set of A and the matched atom index set of AB, to represent the atom set (residue atom index set) corresponding to the derived structure of derivative A based on skeleton B. Figure 1 In the example, the residual atom index set is (0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 39, 40, 41, 42).

[0031] (2) In this embodiment, the selected derivatives are defined as those formed by the condensation of the skeleton and groups through O-acylation or etherification reactions, and the group structure must have one more oxygen atom than the group residues in the derivative. Each AC matched atom indexset is traversed. If there exists a set that has only one more atom number than the residual atom index set, and the atom corresponding to that atom number is an oxygen atom, then it indicates that derivative A consists of two parts: skeleton B and group C. Figure 2 In the example, the difference between the AC matched atom index set and the residual atom index set is (18), which corresponds to the oxygen atom at the skeleton and group connection position.

[0032] like Figure 3 As shown, this invention is applicable to a multi-round derivation and extension method guided by the skeleton of modified chain extension. Metabolites usually undergo a multi-round chain extension process of modification in metabolic reactions. This patent uses the derivative determined in the first round (Round 1) as a seed, iteratively determines the derivative of the next round (Round 2), and extends for n rounds (Round n) until no new derivatives are generated, at which point the extension process terminates.

[0033] In this embodiment, metabolites widely exhibit chiral characteristics in the stereochemical dimension. The chiral information is applied to the structure matching process between the skeleton and derivatives by setting the `useChirality` parameter of the `GetSubstructMatches` function. Existing reference structure databases contain SMILES in two storage formats: some are 3D SMILES with stereochemical information, which can accurately distinguish molecular stereoisomers; the rest are 2D SMILES without stereochemical annotations, which cannot identify or distinguish stereoisomers. During substructure matching, specific matching rules need to be established based on the stereochemical annotations of the skeleton and derivatives. For cases where the skeleton contains stereochemical information but the derivative does not, `useChirality = False` is set to relax the chiral matching conditions. In all other cases, `useChirality = True` is set to restrict the matching of chiral molecules; that is, molecules with inconsistent chirality will not be successfully matched.

[0034] The screening rule based on molecular weight difference in this invention is as follows: After the derivative determination step based on substructure matching, this patent uses a self-built representative modification type molecular weight difference rule library to further screen candidate derivatives by comparing the consistency between the molecular weight difference of the derivative and the skeleton and the molecular weight difference corresponding to specific groups in the rule library (error set to 1ppm).

[0035] In this embodiment, the automatic network redundancy removal method based on graph theory and the reaction principle is as follows: Metabolite reaction pairs from the KEGG pathway database and backbone-derived pairs obtained through multiple rounds of structural modification expansion are merged to establish a KEGG structure expansion network. This patent utilizes the following steps to remove redundancy from the network. In the network, for cases where multiple derivation pathways exist for a given backbone and downstream derivative, long indirect paths involving multiple derivation steps are retained. The method involves traversing the direct paths in the network, pre-deleting these direct paths, and then determining whether there are other long indirect paths between the two nodes corresponding to these direct paths. If such paths exist, the direct paths are deleted from the final result. If no long indirect paths exist, the direct paths are retained. Figure 4 In the example, the indirect long path where node1 is modified by Hexopyranoside to become node2, and node2 is further modified by Galloyl to become node3 is preserved.

[0036] Figure 5 This diagram illustrates the representation of the KEGG structure extension network. Solid blue arrows represent pathway relationships within KEGG, while dashed yellow arrows represent derivation relationships based on the backbone and modifying groups. Taking met02 as an example, one of the basic flavonoid metabolites in KEGG, the metabolic relationship of met02 being derived into met04 via coumaroyl glycoside modification was extended through substructure matching. In downstream functional analysis, metabolite expression data are mapped onto the network, with node colors representing changes in metabolite expression. The KEGG structure extension network can visually present the distribution, enrichment levels, and metabolic flux changes of the core backbone and downstream derivatives, providing a basis for identifying key pathways and biomarkers.

[0037] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An automated method for establishing extended metabolite pathway networks based on structural features, characterized in that, include: S1: Obtain the basic scaffold structures of metabolites from the KEGG pathway database, characterize them using SMILES, and construct them into a seed structure library. Obtain representative modification group structures from integrated literature and construct them into a modification group library; Obtain metabolite structure information from the reference structure database to serve as a candidate derivative library; S2: Using a substructure matching algorithm, the candidate derivative structure is matched with the seed skeleton and modifying groups to obtain a set of matching atomic numbers; by calculating the set difference and combining it with the modification reaction characteristics, it is determined whether the candidate derivative is derived from a specific skeleton and groups. S3: Using the derivative obtained in the first round of judgment as the new seed structure, iteratively execute step S2 to perform multiple rounds of modification chain extension structure matching and judgment until no new derivatives are generated, forming an extended metabolite structure association set. S4: During the substructure matching process, chiral matching rules are dynamically set based on the stereochemical information of the skeleton and the derivative; further, the molecular weight difference rule library of modification type is used to verify the molecular weight difference between the derivative and the skeleton, and to screen derivatives that meet the quality consistency requirements. S5: Merge the original KEGG reaction pairs with the extended skeleton-derived pairs to construct a structural extension network; use graph theory to identify and remove redundant direct derivation pathways, retain indirect pathways with multiple derivation steps, and finally output a redundant metabolite pathway extension network.

2. The method according to claim 1, characterized in that, In step S2, the set of atomic numbers is obtained by executing the substructure search function of the substructure matching algorithm.

3. The method according to claim 1, characterized in that, In step S2, the determination further includes: calculating the difference between the set of atomic numbers of the candidate derivative and the set of atomic numbers matching the skeleton to obtain the set of atomic numbers of residues; traversing the set of atomic numbers matching the modifying group, if there is a set that has only one more atomic number than the set of atomic numbers of residues, identifying the element corresponding to the extra atomic number, and determining that the candidate derivative is derived from the skeleton and the modifying group through the atomic connection of the element.

4. The method according to claim 1, characterized in that, In step S3, the iterative execution is provided with a termination condition: when a round of expansion process ends and no new derivative structure that does not repeat the existing structure is generated, the iteration terminates.

5. The method according to claim 1, characterized in that, The specific process of dynamically setting the chiral matching rules in step S4 is as follows: if the skeleton structure contains stereochemical information, but the candidate derivative structure does not contain stereochemical information, then the substructure matching parameters are changed.

6. The method according to claim 1 or 5, characterized in that, The stereochemical information is characterized by chiral symbols in the SMILES string.

7. The method according to claim 1, characterized in that, In step S4, the self-built rule library for molecular mass difference of modification type pre-stores the precise molecular mass difference value corresponding to the modification group; The verification process includes: calculating the precise molecular weight difference between the derivative and the backbone, comparing the difference with the molecular weight difference of the modified groups pre-stored in the rule library; if the difference is successfully matched within a preset error range, the verification is passed.

8. The method according to claim 1, characterized in that, In step S5, the identification and deletion of redundant direct derived paths using graph theory methods specifically includes: Traverse every direct edge in the network; temporarily remove a direct edge, and then determine whether there is another indirect path consisting of multiple edges between the two nodes it connects; If at least one of the indirect paths exists, the corresponding direct edge is deleted in the final network; otherwise, the corresponding direct edge is retained.

9. The method according to claim 1, characterized in that, The substructure matching algorithm is implemented using a cheminformatics toolkit, including but not limited to RDKit; the substructure search function is the GetSubstructMatches function or other functions with equivalent functionality.

10. The method according to claim 1, characterized in that, The graph theory method is a path redundancy elimination algorithm based on directed graphs; the structure extension network is constructed with metabolite structures as nodes and derivation relationships as directed edges, where the direction of the edges points from the modified skeleton structure to the derived structure.

Images