RAF co-mutation drug resistance mechanism analysis method based on organoid transcriptome

Abstract

The invention discloses an organoid transcriptome-based RAF co-mutation drug resistance mechanism analysis method, and belongs to the technical field of biomedicine. The method comprises the following steps: collecting and standardizing multi-source heterogeneous organoid transcriptome data, and constructing a dynamic detection timeline covering the whole process of organoid culture and drug treatment; defining dynamic expression characteristics and three core sets of RAF co-mutation, differential expression genes and drug-resistant phenotypes based on a time window; analyzing a dynamic regulation and control relationship among the three by utilizing a basic inference rule set driven by literatures, calculating an analysis weight by counting the locking times of a time window, quantitatively evaluating the analysis effectiveness and endowing a state value; and finally, by constructing an analysis state matrix, mining analysis relevancy among different detection events, and screening an optimal related analysis object. According to the method, the problems of static analysis, fuzzy regulation and control relationship and the like in the traditional drug resistance mechanism research are solved, and the dynamic, accurate and efficient analysis of the RAF comutation drug resistance mechanism is realized.

Description

Technical Field

[0001] This invention relates to the field of biomedical technology, specifically to a method for elucidating the RAF co-mutation drug resistance mechanism based on organoid transcriptome. Background Technology

[0002] Mutations in RAF family genes (including BRAF, NRAS, MEK, etc.) are key drivers in the development and progression of human malignancies such as colorectal cancer, melanoma, and thyroid cancer. The BRAF V600E mutation is the most common, accounting for over 80% of all RAF mutations. Targeted inhibitors against RAF mutations (such as vemurafenib and dabrafenib) have become first-line treatments in clinical practice. However, many patients develop secondary resistance after 6-12 months of treatment, severely impacting treatment efficacy and becoming a core bottleneck limiting the benefits of targeted therapy.

[0003] Organoids, as three-dimensional culture models derived from patient tumor tissue that can accurately replicate the pathological features and gene expression profiles of primary tumors, have been proven to be ideal tools for studying tumor drug resistance mechanisms. Compared to traditional cell lines, they more closely resemble the in vivo tumor microenvironment, and compared to animal models, they offer advantages such as shorter experimental cycles, higher throughput, and lower cost. However, current research on drug resistance mechanisms based on organoids still has significant limitations, including:

[0004] Traditional methods are mostly based on transcriptome data from a single time point for analysis, which ignores the dynamic evolution of RAF co-mutation status, gene expression level and drug resistance phenotype under drug action, making it difficult to capture the temporal causal relationship among the three.

[0005] Organoid transcriptome research involves multi-source heterogeneous data such as RNA-seq, single-cell transcriptome, and methylation. Existing methods lack standardized data cleaning procedures, and batch effects, noisy data, and outliers can easily lead to biased analysis results.

[0006] RAF mutations often exist in the form of co-mutations (such as BRAF V600E+NRAS Q61K, BRAF V600D+MEK1 K57N). Existing methods mostly focus on single mutation types and do not systematically analyze the regulatory association between co-mutations and differentially expressed genes and drug resistance phenotypes. Furthermore, they lack methods to quantitatively assess the reliability of the analysis and do not utilize the biological similarity between events detected by different transcriptomes. The complete analysis process needs to be repeated for each event, which increases experimental costs and prolongs the analysis cycle.

[0007] Therefore, developing a method that can integrate multi-source dynamic data, quantify the regulatory relationship of RAF co-mutation drug resistance, and improve the analysis efficiency is of great significance for overcoming the bottleneck of drug resistance in targeted cancer therapy. Summary of the Invention

[0008] The purpose of this invention is to provide a method for elucidating the RAF co-mutation drug resistance mechanism based on organoid transcriptome, so as to solve the problems mentioned in the background art.

[0009] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0010] A method for elucidating the RAF co-mutation resistance mechanism based on organoid transcriptomes, which includes the following steps:

[0011] Step S1: Collect and process multi-source heterogeneous organoid transcriptome data and establish a relevant database;

[0012] Step S2: Assign a unique identifier and time record to each transcriptome detection event during organoid culture and drug treatment to construct a dynamic timeline for organoid detection;

[0013] Step S3: Based on the constructed dynamic detection timeline and the set time window, identify the analytical features related to dynamic expression, as well as the relevant sets corresponding to RAF co-mutations, differentially expressed genes, and drug resistance phenotypes;

[0014] Step S4: Establish a basic inference rule set, and analyze the dynamic regulatory relationship between RAF co-mutation type, differentially expressed gene type and drug resistance phenotype based on the basic inference rule set;

[0015] Step S5: Take the relevant entities corresponding to the dynamic control relationship as the core entity group, lock the corresponding time window of the core entity group in the dynamic detection timeline, count the number of related events and calculate the parsing weight;

[0016] Step S6: Evaluate the parsing effectiveness of the core entity group under the basic reasoning rules based on the parsing weights, and determine the corresponding parsing state value;

[0017] Step S7: Based on the resolution validity results, establish a resolution status matrix for each transcriptome detection event, analyze the resolution correlation between different transcriptome detection events through the matrix, and select the transcriptome detection event with the highest correlation as the relevant resolution object of the corresponding detection event.

[0018] As a preferred embodiment of the present invention, the specific implementation process of step S1 includes:

[0019] Multi-source heterogeneous organoid transcriptome data were collected, and an organoid transcriptome database was established. The organoid transcriptome database records RNA-seq data, single-cell transcriptome data, methylation data, and drug resistance phenotype detection data. The organoid transcriptome data was cleaned to remove batch effects, noise data, and outliers, and the data format was standardized. A preliminary drug resistance mechanism association network was constructed.

[0020] The preliminary construction of the drug resistance mechanism association network includes: constructing a basic association network structure based on core entities and regulatory relationships between entities; the core entities include RAF co-mutation sites, differentially expressed genes, drug resistance phenotypes, and signaling pathways; the regulatory relationships between entities are keywords related to the interaction between different entities, such as "activation", "inhibition", "mediation", and "dependence".

[0021] As a preferred embodiment of the present invention, the specific implementation process of step S2 includes:

[0022] For each transcriptome detection event during organoid culture and drug processing Assign a unique identifier and timestamp , recorded as Where i represents the transcriptome detection event number, Indicates transcriptome detection events Time points (e.g., 0h, 24h, 48h, 72h after drug treatment);

[0023] Constructing a dynamic detection timeline for organoids , where n represents the total number of transcriptome detection events recorded during organoid culture and drug treatment.

[0024] As a preferred embodiment of the present invention, the specific implementation process of step S3 includes:

[0025] Based on organoid dynamic detection timeline and time window Define dynamic expression analysis features The dynamic expression analysis feature is defined within a specified time window. The combination of co-mutation states occurring within the body and gene expression levels;

[0026] Define the RAF comutation set ,in, Indicates the RAF co-mutation type identified in transcriptome detection event i (e.g., BRAF V600E+NRAS Q61K, BRAF V600D+MEK1 K57N, etc.); defines the set of differentially expressed genes. ,in, Indicate the differentially expressed gene types identified in transcriptome detection event i; define the set of drug resistance phenotypes. ,in, This indicates the drug resistance phenotype identified in transcriptome detection event i (such as increased IC50 value, decreased apoptosis rate, etc.).

[0027] As a preferred embodiment of the present invention, the specific implementation process of step S4 includes:

[0028] A basic inference rule set was established based on existing literature on RAF signaling pathways and drug resistance. Where m represents the total number of basic rules. This represents the d-th basic reasoning rule;

[0029] Using basic reasoning rules During analysis, if the RAF co-mutation type Differentially expressed gene types within a time window after their emergence Developing drug resistance phenotype Changes in this type lead to the establishment of co-mutation types. Differentially expressed gene types and drug resistance phenotype The dynamic regulation relationship is denoted as .

[0030] As a preferred embodiment of the present invention, the specific implementation process of step S5 includes:

[0031] In the dynamic regulation relationship Dedicated to a group of core entities , where h is the group number of the core entity;

[0032] Through dynamic regulation of relationships In the timeline of organoid dynamic detection The core entity is locked in. time window The core entity is obtained by counting the number of times the time window is locked. In basic reasoning rules The parsing weights below Number of time window locks and organoid dynamic detection timeline The ratio of the total number of time windows.

[0033] As a preferred embodiment of the present invention, the specific implementation process of step S6 includes:

[0034] Based on analytical weights, evaluate core entities. In basic reasoning rules The following is the validity of the parsing:

[0035] like This indicates the core entity. In basic reasoning rules The following parsing is valid, if This indicates the core entity. In basic reasoning rules The following parsing is invalid;

[0036] When the parsing is valid, let If the parsing is invalid, let ,and To resolve the state value.

[0037] As a preferred embodiment of the present invention, the specific implementation process of step S7 includes:

[0038] Based on the results of valid or invalid parsing, a parsing state matrix for transcriptome detection event i is established. The row numbers of the parsing state matrix correspond to the group numbers of the core entities, and the column numbers correspond to the numbers of the basic inference rules. The parsing state values ​​are determined using the group numbers of the core entities and the numbers of the basic inference rules, and these values ​​are recorded in the corresponding row and column positions in the matrix to obtain the parsing state matrix for transcriptome detection event i. ;

[0039] Based on the method for generating the parse state matrix, the parse state matrix for each transcriptome detection event is obtained. Let the parse state matrix of transcriptome detection event j be... Given that i ≠ j, assess the analytical correlation between transcriptome detection event i and transcriptome detection event j. In the formula, Analytical state matrix for transcriptome detection event j The transpose of the matrix, Analyzing the state matrix for transcriptome detection event i With transpose matrix The sum of each element in the matrix obtained after the product operation;

[0040] Selecting the relevance of the analysis Transcriptome detection events at maximum , which serves as the relevant parsing time object for transcriptome detection event i.

[0041] The analytical system for elucidating the RAF co-mutation resistance mechanism based on organoid transcriptomes includes:

[0042] A preliminary module for data processing and drug resistance mechanism association network construction is used to collect RNA-seq data, single-cell transcriptome data, methylation data, and drug resistance phenotype detection data related to organoid transcriptomes; it is also used to store all cleaned and standardized organoid transcriptome data to establish a complete organoid transcriptome database.

[0043] The dynamic detection timeline construction module is used to assign a unique identifier to each transcriptome detection event during organoid culture and drug processing, and record the corresponding time node for each event; it is used to integrate the identifiers and corresponding time nodes of all transcriptome detection events to form a dynamic detection timeline covering the entire process of organoid culture and drug processing.

[0044] The dynamic expression feature and core set definition module is used to set the time window parameters required for dynamic analysis; to define dynamic expression analysis features that include combinations of co-mutation states and gene expression levels based on the dynamic detection timeline and the set time window; to determine the RAF co-mutation types in each transcriptome detection event and integrate them to form a RAF co-mutation set; to determine the differentially expressed gene types in each transcriptome detection event and integrate them to form a differentially expressed gene set; and to determine the drug resistance phenotypes in each transcriptome detection event and integrate them to form a drug resistance phenotype set.

[0045] The module for constructing basic inference rules and analyzing dynamic regulatory relationships is used to organize a set of basic inference rules containing multiple basic inference rules based on existing RAF signaling pathway and drug resistance research literature; it is used to determine the correlation between changes in differentially expressed gene types and drug resistance phenotypes within a time window after the emergence of RAF comutation types based on the basic inference rules, and to establish the dynamic regulatory relationship between RAF comutation types, differentially expressed gene types and drug resistance phenotypes.

[0046] The core entity group locking and parsing weight calculation module is used to classify the combination of RAF co-mutation type, differentially expressed gene type and drug resistance phenotype in the dynamic regulatory relationship into groups of core entities; it is used to lock the time window containing each core entity group in the dynamic detection timeline and count the number of time window locking for each core entity group; it is used to calculate the parsing weight of each core entity group under each basic inference rule, which is the ratio of the number of time window locking for the core entity group to the total number of time windows in the dynamic detection timeline;

[0047] The parsing validity evaluation module is used to calculate the average threshold of the parsing weight of all core entity groups under all basic inference rules, compare the parsing weight of each core entity group under the corresponding basic inference rule with the average threshold, determine whether the parsing is valid, and assign a parsing status value.

[0048] The module for constructing the parsing state matrix and analyzing the correlation of detected events is used to fill the parsing state values ​​into the corresponding matrix positions, using the group number of the core entity group as the row number and the number of the basic inference rule as the column number, to generate the parsing state matrix for each transcriptome detection event. It performs transpose and matrix multiplication operations on the parsing state matrices of different transcriptome detection events and calculates the sum of all elements in the product matrix. Based on the matrix operation results, it calculates the parsing correlation between different transcriptome detection events and selects other transcriptome detection events with the highest parsing correlation with the target transcriptome detection event as the relevant parsing objects of the target detection event.

[0049] Compared with existing technologies, the beneficial effects achieved by this invention are as follows: By collecting and standardizing multi-source heterogeneous organoid transcriptome data, a dynamic detection timeline covering the entire process of organoid culture and drug treatment is constructed; dynamic expression characteristics and three core sets—RAF co-mutations, differentially expressed genes, and drug resistance phenotypes—are defined based on time windows; the dynamic regulatory relationships among these three elements are analyzed using a literature-driven basic reasoning rule set; the analysis weight is calculated by locking the number of times the time window is statistically analyzed, the analysis effectiveness is quantitatively evaluated, and state values ​​are assigned; finally, by constructing an analysis state matrix, the correlation between different detection events is mined, and the optimal related analysis objects are selected. This invention solves the problems of static analysis, insufficient data integration, and ambiguous regulatory relationships in traditional drug resistance mechanism research, achieving dynamic, accurate, and efficient analysis of RAF co-mutation drug resistance mechanisms. It provides a scientific basis for targeted drug combination development and drug resistance reversal strategy formulation, and has significant clinical translational value. Attached Figure Description

[0050] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof.

[0051] Figure 1 This is a schematic diagram illustrating the steps of the method for elucidating the RAF co-mutation drug resistance mechanism based on organoid transcriptome in this invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] Please see Figure 1 In this first embodiment, a method for elucidating the RAF co-mutation resistance mechanism based on organoid transcriptomics is provided. This embodiment uses "elucidation of the resistance mechanism of BRAF / NRAS co-mutations to the BRAF inhibitor vemurafenib in organoids derived from colorectal cancer patients" as the application scenario. The experimental subjects are primary tumor cells isolated from surgical resection specimens of 10 colorectal cancer patients. Organoid models are constructed by three-dimensional culture. Each sample is set with 3 replicates. The drug treatment concentration is the vemurafenib IC50 value (1.2 μM). Transcriptome and drug resistance phenotype detection are performed at 0h (control group), 24h, 48h, and 72h after drug treatment, respectively. A total of 10×3×4=120 transcriptome detection events (n=120) are obtained.

[0054] The method includes the following steps:

[0055] Step S1: Collect and process multi-source heterogeneous organoid transcriptome data and establish a relevant database;

[0056] For example, multi-source heterogeneous organoid transcriptome data are collected, and an organoid transcriptome database is established. The organoid transcriptome database records RNA-seq data, single-cell transcriptome data, methylation data, and drug resistance phenotype detection data. The organoid transcriptome data is cleaned to remove batch effects, noise data, and outliers, and the data format is standardized to initially construct a drug resistance mechanism association network.

[0057] The preliminary construction of the drug resistance mechanism association network includes: constructing a basic association network structure based on core entities and regulatory relationships between entities; the core entities include RAF co-mutation sites, differentially expressed genes, drug resistance phenotypes, and signaling pathways; the regulatory relationships between entities are keywords related to the interaction between different entities, such as "activation", "inhibition", "mediation", and "dependence".

[0058] Step S2: Assign a unique identifier and time record to each transcriptome detection event during organoid culture and drug treatment to construct a dynamic timeline for organoid detection;

[0059] For example, for each transcriptome detection event during organoid culture and drug treatment. Assign a unique identifier and timestamp , recorded as Where i represents the transcriptome detection event number, Indicates transcriptome detection events Time points (e.g., 0h, 24h, 48h, 72h after drug treatment);

[0060] Constructing a dynamic detection timeline for organoids , where n represents the total number of transcriptome detection events recorded during organoid culture and drug treatment.

[0061] Step S3: Based on the constructed dynamic detection timeline and the set time window, identify the analytical features related to dynamic expression, as well as the relevant sets corresponding to RAF co-mutations, differentially expressed genes, and drug resistance phenotypes;

[0062] For example, based on the organoid dynamic detection timeline and time window Define dynamic expression analysis features The dynamic expression analysis feature is defined within a specified time window. The combination of co-mutation states occurring within the body and gene expression levels;

[0063] Define the RAF comutation set ,in, Indicates the RAF co-mutation type identified in transcriptome detection event i (e.g., BRAF V600E+NRAS Q61K, BRAF V600D+MEK1 K57N, etc.); defines the set of differentially expressed genes. ,in, Indicate the differentially expressed gene types identified in transcriptome detection event i; define the set of drug resistance phenotypes. ,in, This indicates the drug resistance phenotype identified in transcriptome detection event i (such as increased IC50 value, decreased apoptosis rate, etc.).

[0064] For example, time window h (determined based on the conventional response time of vemurafenib in tumor cells’ metabolic cycle and gene expression changes).

[0065] Dynamic expression characteristics: DEP(TL, 24h) = combination of comutation status within a specified 24h window + differential gene expression level (e.g., "BRAF V600E + NRAS Q61K comutation + MEK1 log2FC = 1.8").

[0066] RAF comutation set MU: Validated by Sanger sequencing, a total of 8 comutation types were identified, MU={mu_1(BRAFV600E+NRAS Q61K), mu_2(BRAF V600D+MEK1 K57N), mu_3(BRAF V600E+PI3KCA E545K), …, mu_8(NRAS Q61K+MEK2 Y130C)};

[0067] The differentially expressed gene set GE: 124 differentially expressed genes were obtained by screening with DESeq2 (threshold: FDR<0.05, |log2FC|>1), GE={ge_1(MEK1), ge_2(ERK2), ge_3(PI3KCA), ..., ge_124(JNK)};

[0068] Drug resistance phenotype set RE: Based on phenotypic detection results, RE = {re_1 (IC50 increased by 2 times), re_2 (IC50 increased by 3 times), re_3 (apoptosis rate decreased by 30%), re_4 (apoptosis rate decreased by 50%)}.

[0069] Step S4: Establish a basic inference rule set, and analyze the dynamic regulatory relationship between RAF co-mutation type, differentially expressed gene type and drug resistance phenotype based on the basic inference rule set;

[0070] For example, a basic inference rule set is established based on existing literature on RAF signaling pathways and drug resistance research. Where m represents the total number of basic rules. This represents the d-th basic reasoning rule;

[0071] Using basic reasoning rules During analysis, if the RAF co-mutation type Differentially expressed gene types within a time window after their emergence Developing drug resistance phenotype Changes in this type lead to the establishment of co-mutation types. Differentially expressed gene types and drug resistance phenotype The dynamic regulation relationship is denoted as ;

[0072] For example, br_1: BRAF V600E co-mutation → MEK1 upregulation → increased IC50;

[0073] br_2: BRAF V600D co-mutation → ERK2 upregulation → decreased apoptosis rate;

[0074] br_3: NRAS Q61K co-mutation → PI3KCA upregulation → increased IC50;

[0075] br_4: MEK1 K57N co-mutation → JNK downregulation → decreased apoptosis rate;

[0076] br_5: BRAF V600E+NRAS Q61K co-mutation → STAT3 upregulation → increased IC50;

[0077] Criteria for determining regulatory relationships: If, after the comutation type mu_i appears, the expression change of the differentially expressed gene ge_i conforms to the rules within the Δt=24h window, and is accompanied by the emergence of the drug resistance phenotype re_i, then a dynamic regulatory relationship Δt: (mu_i, ge_i, re_i) is established.

[0078] Step S5: Take the relevant entities corresponding to the dynamic control relationship as the core entity group, lock the corresponding time window of the core entity group in the dynamic detection timeline, count the number of related events and calculate the parsing weight;

[0079] For example, in the dynamic regulation relationship Dedicated to a group of core entities , where h is the group number of the core entity;

[0080] Through dynamic regulation of relationships In the timeline of organoid dynamic detection The core entity is locked in. time window The core entity is obtained by counting the number of times the time window is locked. In basic reasoning rules The parsing weights below Number of time window locks and organoid dynamic detection timeline The ratio of the total number of time windows.

[0081] Step S6: Evaluate the parsing effectiveness of the core entity group under the basic reasoning rules based on the parsing weights, and determine the corresponding parsing state value;

[0082] For example, core entities are evaluated based on parsed weights. In basic reasoning rules The following is the validity of the parsing:

[0083] like This indicates the core entity. In basic reasoning rules The following parsing is valid, if This indicates the core entity. In basic reasoning rules The following parsing is invalid;

[0084] When the parsing is valid, let If the parsing is invalid, let ,and To resolve the state value.

[0085] Step S7: Based on the resolution validity results, establish a resolution status matrix for each transcriptome detection event, analyze the resolution correlation between different transcriptome detection events through the matrix, and select the transcriptome detection event with the highest correlation as the relevant resolution object of the corresponding detection event;

[0086] For example, based on the results of valid or invalid parsing, a parsing state matrix for transcriptome detection event i is established, where the row numbers of the parsing state matrix correspond to the group numbers of the core entities, and the column numbers correspond to the numbers of the basic inference rules. The parsing state values ​​are determined by the group numbers of the core entities and the numbers of the basic inference rules, and these parsing state values ​​are recorded in the corresponding row and column positions in the matrix to obtain the parsing state matrix for transcriptome detection event i. ;

[0087] Based on the method for generating the parse state matrix, the parse state matrix for each transcriptome detection event is obtained. Let the parse state matrix of transcriptome detection event j be... Given that i ≠ j, assess the analytical correlation between transcriptome detection event i and transcriptome detection event j. In the formula, Analytical state matrix for transcriptome detection event j The transpose of the matrix, Analyzing the state matrix for transcriptome detection event i With transpose matrix The sum of each element in the matrix obtained after the product operation;

[0088] Selecting the relevance of the analysis Transcriptome detection events at maximum , which serves as the relevant parsing time object for transcriptome detection event i.

[0089] In this second embodiment, an analytical system is provided for the analysis of RAF co-mutation drug resistance mechanisms based on organoid transcriptomes. This system includes:

[0090] A preliminary module for data processing and drug resistance mechanism association network construction is used to collect RNA-seq data, single-cell transcriptome data, methylation data, and drug resistance phenotype detection data related to organoid transcriptomes; it is also used to store all cleaned and standardized organoid transcriptome data to establish a complete organoid transcriptome database.

[0091] The dynamic detection timeline construction module is used to assign a unique identifier to each transcriptome detection event during organoid culture and drug processing, and record the corresponding time node for each event; it is used to integrate the identifiers and corresponding time nodes of all transcriptome detection events to form a dynamic detection timeline covering the entire process of organoid culture and drug processing.

[0092] It should be noted that tumor drug resistance is a biological process involving dynamic changes in gene expression and phenotype. By assigning a unique identifier and timestamp to each transcriptome detection event, a dynamic detection timeline covering the entire drug treatment process can be constructed to ensure that the temporal correlation between co-mutations, gene expression, and drug resistance phenotypes at different time points can be captured, thus avoiding the one-sidedness of static analysis.

[0093] The dynamic expression feature and core set definition module is used to set the time window parameters required for dynamic analysis; to define dynamic expression analysis features that include combinations of co-mutation states and gene expression levels based on the dynamic detection timeline and the set time window; to determine the RAF co-mutation types in each transcriptome detection event and integrate them to form a RAF co-mutation set; to determine the differentially expressed gene types in each transcriptome detection event and integrate them to form a differentially expressed gene set; and to determine the drug resistance phenotypes in each transcriptome detection event and integrate them to form a drug resistance phenotype set.

[0094] The module for constructing basic inference rules and analyzing dynamic regulatory relationships is used to organize a set of basic inference rules containing multiple basic inference rules based on existing RAF signaling pathway and drug resistance research literature; it is used to determine the correlation between changes in differentially expressed gene types and drug resistance phenotypes within a time window after the emergence of RAF comutation types based on the basic inference rules, and to establish the dynamic regulatory relationship between RAF comutation types, differentially expressed gene types and drug resistance phenotypes.

[0095] It should be noted that a basic inference rule set was constructed based on existing RAF signaling pathway and drug resistance research literature. Using time windows as constraints, the causal chain of "co-mutation occurrence → differential gene expression change → drug resistance phenotype" was screened to ensure the biological rationality of the regulatory relationship.

[0096] The core entity group locking and parsing weight calculation module is used to classify the combination of RAF co-mutation type, differentially expressed gene type and drug resistance phenotype in the dynamic regulatory relationship into groups of core entities; it is used to lock the time window containing each core entity group in the dynamic detection timeline and count the number of time window locking for each core entity group; it is used to calculate the parsing weight of each core entity group under each basic inference rule, which is the ratio of the number of time window locking for the core entity group to the total number of time windows in the dynamic detection timeline;

[0097] It should be noted that by counting the number of times the core entity group is locked in the time window in the dynamic timeline, calculating the parsing weight, and then comparing it with the global average threshold, the reliability of the parsing results is quantitatively evaluated, thus avoiding false positives caused by random association.

[0098] The parsing validity evaluation module is used to calculate the average threshold of the parsing weight of all core entity groups under all basic inference rules, compare the parsing weight of each core entity group under the corresponding basic inference rule with the average threshold, determine whether the parsing is valid, and assign a parsing status value.

[0099] The module for constructing the parsing state matrix and analyzing the correlation of detected events is used to fill the parsing state values ​​into the corresponding matrix positions, using the group number of the core entity group as the row number and the number of the basic inference rule as the column number, to generate the parsing state matrix for each transcriptome detection event. It performs transpose and matrix multiplication operations on the parsing state matrices of different transcriptome detection events and calculates the sum of all elements in the product matrix. Based on the matrix operation results, it calculates the parsing correlation between different transcriptome detection events and selects other transcriptome detection events with the highest parsing correlation to the target transcriptome detection event as the relevant parsing objects for the target detection event.

[0100] It should be noted that the parsing status (valid / invalid) of each detected event is transformed into a matrix form. Through matrix transpose, multiplication and accumulation operations, the parsing similarity between different events is quantified, highly correlated events are selected as parsing objects, and the reuse of effective parsing results is realized, thereby improving the overall parsing efficiency.

[0101] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0102] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for analyzing RAF co-mutation drug resistance mechanism based on organoid transcriptome, characterized in that, The method comprises the following steps: Step S1: Collect and process multi-source heterogeneous organoid transcriptome data, and establish a related database; Step S2: Assign a unique identifier and time record to each transcriptome detection event during organoid culture and drug treatment, and construct an organoid dynamic detection timeline; Step S3: Based on the constructed dynamic detection timeline and the set time window, determine the analysis features related to dynamic expression, and the related sets corresponding to RAF co-mutation, differentially expressed genes and drug resistance phenotype; Step S4: Establish a set of basic reasoning rules to analyze the dynamic regulation relationship between RAF co-mutation type, differentially expressed gene type and drug resistance phenotype; Step S5: Take the related entities corresponding to the dynamic regulation relationship as the core entity group, lock the corresponding time window of the core entity group in the dynamic detection timeline, count the number of related times and calculate the analysis weight; Step S6: According to the analysis weight, evaluate the analysis effectiveness of the core entity group under the basic reasoning rules, and determine the corresponding analysis state value; Step S7: Based on the analysis effectiveness result, establish an analysis state matrix for each transcriptome detection event, analyze the analysis correlation between different transcriptome detection events through matrix analysis, and select the transcriptome detection event with the highest correlation as the related analysis object of the corresponding detection event. 2.The RAF co-mutation resistance mechanism analysis method based on organoid transcriptome according to claim 1, wherein, The specific implementation process of step S1 comprises: Collect multi-source heterogeneous organoid transcriptome data and establish an organoid transcriptome database, which records RNA-seq data, single-cell transcriptome data, methylation data and drug resistance phenotype detection data; clean the organoid transcriptome data, remove batch effects, noise data and outliers, and standardize the data format to preliminarily construct a drug resistance mechanism correlation network; The preliminary construction of the drug resistance mechanism correlation network comprises: constructing a basic correlation network structure according to the core entities and the regulation relationship between entities; the core entities include RAF co-mutation sites, differentially expressed genes, drug resistance phenotypes and signal pathways; the regulation relationship between entities is the key word of the action relationship between different entities. 3.The RAF co-mutation resistance mechanism analysis method based on organoid transcriptome according to claim 2, characterized in that, The specific implementation process of step S2 comprises: For each transcriptomic detection event during organoid culture and drug treatment process Assigning unique identifiers and timestamps , denoted as where i denotes the number of the transcriptomic detection event, denotes the time node of the transcriptomic detection event . Constructing organoid dynamic detection timelines where n represents the total number of transcriptome detection events recorded during the organoid culture and drug treatment process. 4.The RAF co-mutation resistance mechanism analysis method based on organoid transcriptome according to claim 3, characterized in that, The specific implementation process of step S3 comprises: Dynamic detection timeline based on organoids and time windows defining dynamic expression analysis features ; the dynamic expression analysis features are combinations of co-mutation status and gene expression levels occurring within a specified time window ​ defining a set of RAF co-mutations wherein, represents a RAF co-mutation type determined in the transcriptomic test event i; defining a set of differentially expressed genes wherein, represents a differentially expressed gene type determined in the transcriptomic test event i; defining a set of drug resistance phenotypes wherein, represents a drug resistance phenotype determined in the transcriptomic test event i.

5. The method according to claim 4, wherein, The specific implementation process of step S4 comprises: A set of basic reasoning rules is established based on the existing RAF signal pathway and drug resistance research literature wherein m represents the total number of basic rules, represents the dth basic reasoning rule; Using basic reasoning rules During analysis, if the RAF co-mutation type Differentially expressed gene types within a time window after their emergence Developing drug resistance phenotype Changes in this type lead to the establishment of co-mutation types. Differentially expressed gene types and drug resistance phenotype The dynamic regulation relationship is denoted as .

6. The method according to claim 5, wherein, The specific implementation process of step S5 comprises: In the dynamic regulation relationship denoted as a set of core entities where h is the group number of the core entities; Through dynamic regulation of relationships In the timeline of dynamic detection of organoids The core entity is locked in. time window The core entity is obtained by counting the number of times the time window is locked. In basic reasoning rules The following parsing weight Number of time window locks and organoid dynamic detection timeline The ratio of the total number of time windows.

7. The organoid-transcriptome-based RAF co-mutation resistance mechanism resolving method according to claim 6, characterized in that, The specific implementation process of step S6 comprises: Based on the resolved weights, the core entities are evaluated Under the base inference rules of the resolved validity: If , then the core entity is valid under the base reasoning rule ; if , then the core entity is invalid under the base reasoning rule ; When the parsing is valid, let If the parsing is invalid, let ,and To resolve the state value. 8.The RAF co-mutation resistance mechanism analysis method based on organoid transcriptome according to claim 1, wherein, The specific implementation process of step S7 comprises: Based on the analytical effective or analytical invalid result, an analytical state matrix of the transcriptome detection event i is established, and the row number of the analytical state matrix corresponds to the group number of the core entity, the column number of the analytical state matrix corresponds to the number of the basic reasoning rule, the analytical state value is determined through the group number of the core entity and the number of the basic reasoning rule, and the analytical state value is recorded to the matrix position corresponding to the row number and the column number, to obtain the analytical state matrix of the transcriptome detection event i ; Based on the generating mode of the analytical state matrix, an analytical state matrix of each transcriptomic detection event is obtained, let the analytical state matrix of the transcriptomic detection event j be , and i≠j, the analytical correlation between the transcriptomic detection event i and the transcriptomic detection event j is evaluated , wherein, is the analytical state matrix of the transcriptomic detection event j is the transpose matrix of the analytical state matrix of the transcriptomic detection event j is the analytical state matrix of the transcriptomic detection event i is the transpose matrix of the analytical state matrix of the transcriptomic detection event i is the cumulative sum of each matrix element in the matrix obtained after the product operation of the analytical state matrix of the transcriptomic detection event i and the transpose matrix of the analytical state matrix of the transcriptomic detection event j. Selecting the analytical correlation Maximal transcriptome detection event , as the analytical correlation time object of the transcriptome detection event i.

9. An analysis system for performing the organoid transcriptome-based RAF co-mutation drug resistance mechanism analysis method according to claim 1, comprising: A data processing and drug resistance mechanism correlation network preliminary construction module for collecting organoid transcriptome-related RNA-seq data, single-cell transcriptome data, methylation data and drug resistance phenotype detection data; for storing all cleaned and standardized organoid transcriptome data, and establishing a complete organoid transcriptome database; A dynamic detection timeline construction module is configured to assign a unique identifier to each transcriptome detection event in the process of organoid culture and drug treatment, and record the time node corresponding to each event; and is configured to integrate the identifiers of all transcriptome detection events and the corresponding time nodes to form a dynamic detection timeline covering the whole process of organoid culture and drug treatment; A dynamic expression feature and core set definition module is configured to set a time window parameter required for dynamic analysis; is configured to define a dynamic expression analysis feature including a co-mutation state and a gene expression level combination based on the dynamic detection timeline and the set time window; is configured to determine the RAF co-mutation type in each transcriptome detection event, and integrate to form a RAF co-mutation set; is configured to determine the differential expression gene type in each transcriptome detection event, and integrate to form a differential expression gene set; and is configured to determine the drug resistance phenotype in each transcriptome detection event, and integrate to form a drug resistance phenotype set; A basic reasoning rule construction and dynamic regulation relationship analysis module is configured to sort and form a basic reasoning rule set including a plurality of basic reasoning rules based on existing RAF signaling pathways and drug resistance research literature; and is configured to determine the change correlation between the differential expression gene type and the drug resistance phenotype within the time window after the appearance of the RAF co-mutation type according to the basic reasoning rules, and establish a dynamic regulation relationship between the RAF co-mutation type, the differential expression gene type and the drug resistance phenotype; A core entity group locking and analysis weight calculation module is configured to divide the combination of the RAF co-mutation type, the differential expression gene type and the drug resistance phenotype in the dynamic regulation relationship into a group of core entities; and is configured to lock the time window containing each core entity group in the dynamic detection timeline, and count the number of time window lockings corresponding to each core entity group; is configured to calculate the analysis weight of each core entity group under each basic reasoning rule, which is the ratio of the number of time window lockings corresponding to the core entity group to the total number of time windows in the dynamic detection timeline; An analysis effectiveness evaluation module is configured to calculate the average threshold of the analysis weights of all core entity groups under all basic reasoning rules, compare the analysis weight of each core entity group under the corresponding basic reasoning rule with the average threshold, determine whether the analysis is effective, and assign an analysis state value; An analysis state matrix construction and detection event correlation analysis module is configured to fill the analysis state value into the corresponding matrix position with the group number of the core entity group as the row number and the number of the basic reasoning rule as the column number, generate the analysis state matrix of each transcriptome detection event, perform transpose operation and matrix multiplication operation on the analysis state matrices of different transcriptome detection events, and calculate the cumulative sum of all elements in the product matrix; and is configured to calculate the analysis correlation between different transcriptome detection events according to the matrix operation result, select other transcriptome detection events with the largest analysis correlation with the target transcriptome detection event as the relevant analysis objects of the target detection event.

Images