Targeted senile degenerative bone disease key lesion regulation factor mRNA therapy recommendation evaluation method, electronic equipment and program product

Abstract

The invention discloses a targeted senile degenerative bone disease key lesion regulation factor mRNA therapy recommendation evaluation model method, and the model comprises a data input layer which receives bone disease genetic association and regulation data, including GWAS summary data, space transcriptome and single cell transcriptome data; the intervention target priority ordering module is used for integrating the obtained genetic evidence, regulation evidence and network evidence and carrying out priority ordering on intervention targets; and the intervention target identification network module is used for predicting a key intervention target according to an intervention target network generated by acquiring a gene interaction relationship from the pathway, and obtaining a potential mRNA intervention therapy. The genetic evidence comprises a genetic risk site set annotated through multi-modal regulation genomics data; the regulation evidence comprises functional genomics data related to diseases; the network evidence comprises a high-credibility protein interaction relationship. And the intervention target identification network module is used for analyzing an intervention target network and further comprises disturbance removal analysis and regulation and control hierarchy analysis.

Description

TECHNICAL FIELD

[0001] The present disclosure belongs to the technical field of bioinformatics, and particularly relates to a method for intelligent recommendation and evaluation of mRNA therapy targeting key lesion regulatory factors of senile degenerative bone disease, an electronic device and a program product. BACKGROUND

[0002] Aging is a complex biological process characterized by the progressive decline of physiological functions of various systems in the body. With this systemic functional degradation, the risk of age-related chronic degenerative diseases significantly increases, including musculoskeletal diseases such as osteoarthritis, neurodegenerative diseases such as Alzheimer's disease, and metabolic diseases such as diabetes. Among them, senile degenerative bone disease represented by osteoarthritis has a continuously rising prevalence in people over 60 years old, and has become the main cause of pain, dysfunction and even disability in the elderly, bringing a heavy economic burden to the global public health system. However, current clinical intervention methods are still mainly limited to symptom relief, and are difficult to effectively reverse disease progression.

[0003] Studies have shown that degenerative bone disease has significant pathological complexity and clinical heterogeneity, and its occurrence and development involve the joint action of multiple factors such as genetic susceptibility, environmental exposure and lifestyle. Current clinical treatment strategies mainly focus on symptom management, including the use of non-steroidal anti-inflammatory drugs (NSAIDs) and opioid analgesics for pain control, and lifestyle interventions such as weight management to improve joint function load. However, these methods are all palliative treatments, which can relieve the suffering of patients to a certain extent, but cannot change the natural course of the disease. Long-term use of NSAIDs can cause gastrointestinal damage, cardiovascular events and other adverse reactions, while opioid drugs have the risk of addiction, further limiting their clinical application.

[0004] Currently, there is no disease-modifying osteoarthritis drug (DMOAD) that can effectively delay, block or reverse joint structure destruction on the market. Most DMOAD candidates in the clinical development stage have not shown clear disease-modifying effects, reflecting the significant shortcomings in the current understanding of the key drivers of degenerative bone disease and their dynamic network interaction mechanisms. Traditional drug treatment models mainly rely on small molecule compounds or monoclonal antibodies targeting 'druggable' target proteins with clear active sites, such as specific enzymes or receptors. However, genomic studies have shown that more than 85% of disease-related proteins lack suitable active domains for small molecule or antibody binding, and are classified as 'undruggable' targets. Targeted protein degradation technology is expected to break through this treatment bottleneck. The proteolysis-targeting chimera (PROTAC) and molecular glue strategies guide E3 ubiquitin ligases to ubiquitinate target proteins, and then target proteins are degraded through the proteasome degradation pathway. Unlike traditional inhibition, this mechanism does not depend on whether the target protein has an active site, so it greatly expands the range of targetable targets. This technology provides a new approach to addressing the heterogeneity challenge of degenerative bone disease, and is expected to achieve tissue-specific treatment and reduce systemic side effects by precisely degrading key regulatory proteins in specific cell types or pathological links, opening up new paths for developing truly disease-modifying therapies.

[0005] Emerging treatment strategies represented by mRNA therapy also do not rely on the active site or binding pocket of the target protein, bypassing many limitations of traditional drugs in the field of degenerative bone disease and showing breakthrough potential. This technology designs specific messenger ribonucleic acid (mRNA) sequences to enable cells to express proteins or antigens with therapeutic functions, thereby triggering specific immune responses or achieving protein function supplementation. Currently, mRNA technology has made important progress in protein replacement therapy, tumor neoantigen therapy, gene editing tool delivery, and vaccine design, showing a wide range of application prospects. mRNA technology has multiple revolutionary advantages: first, it has high coding flexibility and can express any protein or regulatory factor with therapeutic function, such as anti-inflammatory cytokines, cartilage repair growth factors, and gene editing tools, allowing for intervention at the root of the disease; second, mRNA is transiently expressed in the cytoplasm without entering the nucleus, avoiding the risk of genomic integration and having high safety; third, sequence design is simple, and production processes are easy to standardize and scale up, which can significantly shorten the drug development cycle. In the field of degenerative bone disease such as osteoarthritis, mRNA therapy shows diversified application potential, mainly including the following forms:

[0006] The first is mRNA drugs based on local delivery. By using new delivery systems such as LNP and hydrogel, mRNA encoding osteogenic or chondrogenic factors is directly injected into the joint to achieve targeted protein expression, thereby enhancing efficacy while significantly reducing systemic toxicity.

[0007] The second is mRNA vaccine strategies. By designing mRNA sequences targeting senescent cells or key pro-inflammatory factors, the body can induce specific immune responses to eliminate pathogenic cells or alleviate joint inflammation.

[0008] The third is cell therapy. Drawing on the in vivo CAR-T technology that has made progress in systemic lupus erythematosus, mRNA-LNP systems are used to directly modify immune cells in vivo, targeting specific antigens in diseased bone and joint tissues, providing a new direction for immunotherapy in degenerative bone diseases.

[0009] Although mRNA technology has made important progress in infectious disease vaccines and tumor immunotherapy, its application in bone disease treatment still faces key challenges, with the systematicity and precision of target selection being the most prominent bottleneck. Existing methods rely on scattered knowledge or single databases, lacking a computational framework that can integrate multi-dimensional omics data and intelligently identify key upstream regulatory factors suitable for mRNA intervention, making it difficult to efficiently lock in truly therapeutic molecular targets. In addition, the delivery specificity of mRNA in bone tissue and the preclinical efficacy evaluation system also need to be further improved. Therefore, building an intelligent target recommendation and evaluation framework for mRNA therapy is of great significance for promoting the conversion of this technology in the field of degenerative bone diseases. Integrating genetic evidence and multi-omics data to establish a systematic mRNA therapy intelligent recommendation and evaluation framework is key to moving this field from concept discovery to clinical translation. Such a framework will not only help to accurately identify actionable targets for degenerative bone diseases, but also provide a basis for developing treatment strategies with disease-modifying effects.

[0010] Converting genetic and regulatory genomic knowledge into effective mRNA therapeutic targets is a key challenge in current research on degenerative bone diseases in the elderly. Although genome-wide association studies (GWAS) have identified a large number of genetic variations in osteoarthritis research, due to the limited degree of explanation of disease heritability, the high complexity of disease genetic structure, and the difficulty in analyzing the association mechanism between non-coding region risk loci and effector genes, there are still many difficulties in applying GWAS findings to intervention target identification and clinical translation. The main problems include,

[0011] First, most of the variations discovered by GWAS are located in non-coding regions. These risk loci may affect gene expression through enhancer-promoter interactions and other long-range regulatory mechanisms, making it difficult to accurately define their effector genes.

[0012] Secondly, the regulatory effects of these non-coding variants are cell type-specific, with differential expression patterns in different joint tissues such as cartilage and synovium. Therefore, it is necessary to combine tissue or cell-specific functional genomics data (such as chromatin accessibility and histone modification) for analysis.

[0013] In addition, GWAS essentially reflects statistical association, making it difficult to distinguish causal relationships in disease occurrence and unable to capture the effects of environmental factors (such as obesity and mechanical load) on gene expression. Therefore, the integration of single-cell and spatial transcriptome data is particularly crucial: single-cell technology reveals the specific gene expression patterns of different cell subpopulations in cartilage tissue, defining the cell type-specific effects of non-coding variants at a higher resolution; spatial transcriptome can capture expression differences in different areas of the affected joint to reflect the effects of the local microenvironment during cartilage degradation, making up for the shortcomings of GWAS in analyzing spatial heterogeneity. To overcome these limitations, a multi-modal target prioritization system needs to be established:

[0014] 1) Use three-dimensional genomics data (such as promoter capture Hi-C) to analyze the physical interaction between non-coding variants and target genes; integrate tissue-specific expression quantitative trait locus (eQTL) data to establish a genetic variation-gene expression correlation map;

[0015] 2) Integrate disease-related functional genomics data (single-cell transcriptome sequencing and spatial transcriptomics sequencing) to analyze how environmental risk factors affect gene expression and explore the roles of these genes in disease occurrence;

[0016] 3) Combine protein interaction network analysis to identify new targets that interact with known therapeutic targets in the protein interaction network, thereby constructing a more complete targeted intervention network. This strategy of integrating genomic regulatory features and protein network topology can significantly improve the efficiency of therapeutic target discovery and clinical translation potential.

[0017] Current mainstream target prioritization methods (such as the Open Targets platform) integrate eQTL, pQTL, and other genomics data, but still have significant shortcomings, failing to systematically integrate disease-related functional genomics features and protein interaction network information. This deficiency leads to two key limitations:

[0018] On the one hand, it is difficult to capture environmentally dependent pathological gene expression changes, and on the other hand, it is unable to identify potential targets with indirect genetic associations and their synergistic mechanisms, thereby severely restricting the scope of target discovery and the development of network pharmacology strategies. SUMMARY

[0019] One of the embodiments of the present disclosure is an intelligent recommendation and evaluation framework for mRNA therapy targeting key pathogenesis regulators of senile degenerative bone diseases, comprising:

[0020] A data input layer receives genetic association and regulatory data of bone diseases, including GWAS, spatial transcriptome and single-cell transcriptome data;

[0021] An intervention target prioritization module integrates the obtained genetic evidence, regulatory evidence and network evidence to prioritize the intervention targets;

[0022] An intervention target identification network module predicts key intervention targets and obtains potential mRNA intervention therapies according to the intervention target network generated from the gene interaction relationship obtained from the pathway. BRIEF DESCRIPTION OF DRAWINGS

[0023] The above and other objects, features and advantages of the exemplary embodiments of the present disclosure will be readily understood through reading the detailed description of the exemplary embodiments of the present disclosure below in conjunction with the accompanying drawings. In the drawings, several embodiments of the present disclosure are illustrated by way of example in which:

[0024] Figure 1 The principle schematic diagram of the intelligent recommendation and evaluation framework for mRNA therapy targeting key pathogenesis regulators of senile degenerative bone diseases according to one of the embodiments of the present disclosure.

[0025] Figure 2 The intervention target network schematic diagram of osteoarthritis according to one of the embodiments of the present disclosure.

[0026] Figure 3 The perturbation removal analysis schematic diagram based on the intervention target network according to one of the embodiments of the present disclosure.

[0027] Figure 4 The regulatory hierarchy analysis schematic diagram based on the intervention target network according to one of the embodiments of the present disclosure.

[0028] Figure 5 The summary schematic diagram of multiple analyses according to one of the embodiments of the present disclosure. DETAILED DESCRIPTION

[0029] The present disclosure finds that there are still significant limitations in the process of converting genetics and regulatory genomics knowledge into effective mRNA treatment targets through research on prior art and literature.

[0030] Firstly, although genome-wide association studies (GWAS) have identified a large number of genetic loci associated with degenerative bone diseases, most of them are located in non-coding regions, which may affect the expression of genes through long-range regulation of chromatin and play a role through tissue-specific regulatory mechanisms, thus it is difficult to accurately analyze the regulatory target genes and functional pathways of non-coding variations in specific joint cell types.

[0031] Secondly, such methods cannot capture the dynamic changes in gene expression driven by environmental factors (such as inflammation, mechanical load, etc.), which may be important factors leading to the occurrence and progression of degenerative bone diseases. At the same time, the existing evaluation system lacks consideration of the function of target points and their importance in the network. Relying solely on statistical correlation strength cannot infer causal relationships in the disease occurrence process or identify key intervention nodes in the disease regulatory network. These shortcomings collectively limit the development of efficient mRNA therapies based on genetic findings, highlighting the need for integration of functional genomics data.

[0032] To solve the above technical problems, the present disclosure proposes an mRNA therapy intelligent recommendation and evaluation framework targeting key pathological regulatory factors of senile degenerative bone diseases. This method is based on genetics, highlights the integration of functional genomics information, and relies on gene network interactions to identify key intervention target networks and their regulatory hierarchy, thereby converting disease genetic findings (GWAS summary data) and environmentally dependent functional genomics changes (gene expression changes) into intervention targets with genetic evidence or environmental regulation characteristics to clarify and accelerate the identification and application of mRNA treatment targets.

[0033] According to one or more embodiments, an mRNA therapy intelligent recommendation and evaluation model targeting key pathological regulatory factors of senile degenerative bone diseases includes the following steps:

[0034] Step S1: input GWAS summary statistics and multi-modal regulatory genomic data sets of degenerative bone diseases, combine genomic and network evidence, and define core genes and peripheral genes with genetic evidence support;

[0035] Step S2: input spatial transcriptome and single-cell transcriptome data of degenerative bone diseases, combine functional genomics and network evidence, and define core genes and peripheral genes regulated by the environment;

[0036] Step S3: calculate the affinity score of the gene using the restart random walk algorithm, form a gene-predictor matrix, and integrate the score through sequential statistical methods to systematically prioritize candidate genes, achieving the integration of genetics and functional genomics;

[0037] Step S4: Intervention target network analysis based on prioritization results, using reward collection Stanley tree algorithm to identify key intervention sub-networks enriched in high-score nodes in the network integrated by KEGG pathways;

[0038] Step S5: Perturbation removal and causal network analysis of genes in the intervention sub-network, to assess the impact of single or multiple genes on network stability, and to analyze the regulatory hierarchy of pathological mechanisms, and to identify core intervention nodes;

[0039] Step S6: Comprehensive evaluation of key intervention targets, combined with genetic evidence and functional genomics regulatory features, to ultimately determine targets with mRNA treatment potential.

[0040] The following is further described in different steps.

[0041] (1) Prioritization of intervention targets for senile degenerative bone disease

[0042] First, based on multi-modal regulatory omics data such as genomic proximity, QTL and PCHi-C, the lead SNPs and LD SNPs obtained from GWAS are annotated, and the core genes supported by genetic evidence are defined; based on spatial and single-cell transcriptome data resources, genes with regional and cell type-specific expression disorders are defined as core genes supported by regulatory evidence. Next, through random walk algorithm, the core genes are defined as seed genes (starting nodes), the genes with high affinity to the seed genes in the STRING protein interaction network are defined as peripheral genes, and the affinity scores of the core genes and peripheral genes are calculated to obtain the gene-predictor matrix. Further use of order statistic-based combination meta-analysis method to combine different predictors, to realize the integration of genetic evidence, functional genomics evidence and network evidence. Further including:

[0043] (1-1) Degenerative bone disease genetic association and regulatory data preprocessing.

[0044] Obtain disease-related GWAS summary data from NHGRI-EBI GWAS Catalog, and select SNPs with p-value <5×10 -8 as lead SNPs, then calculate LD SNPs (R 2 ≥0.8) with lead SNPs in linkage disequilibrium, and score lead SNPs and LD SNPs through genetic association p-value, significance threshold, and LD strength R 2 .

[0045] The spatial and single-cell transcriptomic datasets were standardized, and the differentially expressed genes (FDR < 0.05, |Fold Change| ≥1.5) in different spatial regions and different cell types (such as inflammatory chondrocytes, fibrochondrocytes, etc.) were analyzed by the methods of DESeq2 and Seurat.

[0046] (1-2) Identification of core genes using genetic evidence and regulatory evidence.

[0047] Based on the lead SNP and LD SNP described above, the present disclosure identifies core genes through genomic evidence, including genomic proximity, QTL, and promoter capture Hi-C (PCHi-C), etc. The score of the core genes takes into account the following factors: (i) SNP distance window, representing the degree of genomic proximity; (ii) genetic association significance level with gene expression of eQTL dataset or protein abundance of pQTL dataset; (iii) physical interaction strength between gene promoter and SNP-containing genomic region (CHiCAGO score) of PCHi-C dataset. In addition, to ensure that the score range is between 0-1, the present disclosure normalizes all SNP-gene pairs by empirical cumulative distribution function (eCDF). In summary, the present disclosure identifies a set of core genes supported by genetic evidence and scores them;

[0048] Based on the region- and cell type-specific dysregulated genes described above, scoring is performed using FDR and Fold Change to quantify the degree and significance of differential expression. Thus, the present disclosure identifies a set of core genes supported by regulatory evidence and scores them.

[0049] (1-3) Identification of peripheral genes using network evidence, and construction of gene-predictor matrix.

[0050] The present disclosure takes core genes as seed nodes, and identifies affected non-seed nodes, i.e. peripheral genes interacting with core genes, in the protein interaction network through a re-starting random walk (RWR) algorithm. The RWR algorithm assigns affinity scores to genes according to their connection degree with seed nodes, and the process is as follows: starting from each seed gene, randomly walking along the network edges to iterate neighbor nodes, at each iteration, the algorithm faces a choice: either randomly moving to a neighboring node or jumping back to the seed node, which is controlled by a restart probability, allowing the algorithm to not only maintain the connection with seed genes, but also explore peripheral genes located far away. After a finite number of iterations, the algorithm obtains a steady-state probability vector, which stores the affinity scores of each gene relative to core genes, ranging from 0 to 1, for measuring the overall influence of seed nodes on the network. For each type of evidence (or different datasets under the same type), the gene list containing these affinity scores constitutes different predictors, respectively.

[0051] In summary, the present disclosure uses the above scoring process (SNPs / transcriptome-core genes-peripheral genes) to construct a gene-predictor matrix, with genes as rows and predictors as columns. In this matrix, affinity scores can be combined to integrate genetic evidence, regulatory evidence, and network evidence. The present disclosure combines the transformed p-values of different predictors by three integration methods (Fisher's combination method, logistic combination method, order statistic combination method), and selects the best method. Finally, the combined p-value is standardized to obtain a priority ranking from 0 to 5.

[0052] (2) Identification of intervention target network

[0053] This analysis identifies a set of high-priority and strongly associated intervention targets. The present disclosure uses the KEGG pathway-merged gene interaction network to identify intervention target subnetworks by heuristically solving the Steiner tree problem. To evaluate the statistical significance (p-value) of the identified intervention target network, the present disclosure performs 100 iterations of degree node permutation test.

[0054] (3) Analysis based on intervention target network genes, further comprising:

[0055] (3-1) Perturbation removal analysis based on intervention target network

[0056] Perturbation removal analysis aims to systematically evaluate the impact of different nodes on network stability and further reveal potential key intervention targets. This analysis proceeds from two dimensions:

[0057] (i) Single node removal, i.e. removing each node in the network one by one, quantifying the disruptive effect of its absence on network structure;

[0058] (ii) Combined node removal, i.e. removing two or more nodes simultaneously, to investigate their synergistic disruptive effects on network integrity.

[0059] The core principle of the analysis is that if a node (or combination of nodes) is critical to the network structure, its removal will disconnect a large number of nodes from the largest connected component (LCC) in the network. The present disclosure quantitatively describes this effect by calculating the disconnecting proportion (i.e. the proportion of nodes that lose connection to the LCC after removal of a specified node). Nodes or combinations of nodes with a significantly elevated disconnecting proportion are defined as critical regulatory factors necessary for maintaining network connectivity. The analysis is divided into three progressive steps to optimize target identification layer by layer:

[0060] (1) Single node removal: each node in the largest connected component is removed in turn, and the disconnecting proportion is calculated, thereby identifying core nodes that can disrupt network connectivity by their absence alone;

[0061] (2) Combined node removal: on the basis of step 1, the core node pairs and their combinations with other genes are removed simultaneously to evaluate their synergistic disruptive effects on network stability, and thereby identify the most influential node combinations;

[0062] (3) Clinical target integration: clinical bone disease-related targets that have entered phase III and above are included in the combined removal analysis to explore the best target combination that can maximally disrupt the network in structure while having clinical safety and effectiveness.

[0063] Through this layer-by-layer progressive perturbation removal analysis, not only single key nodes but also target combinations that play a core role in regulating the network can be identified, thereby providing a solid basis for determining mRNA therapeutic intervention targets.

[0064] (3-2) Regulation level analysis based on intervention target network

[0065] The present disclosure uses QIAGEN Ingenuity Pathway Analysis (IPA) to systematically identify upstream regulatory factors (including various biological entities such as individual genes, compounds, protein families, molecular complexes, and microRNAs, etc.), aiming to map the upstream regulatory network of pathway crosstalk genes. The analysis process involves two steps:

[0066] (1) Priority ranking and conversion of connectivity scores

[0067] The present disclosure converts two core indicators of pathway crosstalk genes, their priority scores and connectivity scores (from disconnect scores in network attack analysis), into two IPA-compatible indicators: scores of P-like (to reflect the significance of prioritization) and Fold Change-like (FC-like) (to reflect the degree of node connectivity).

[0068] , (Equation 1)

[0069] represents the original PI - Pro priority score of each crosstalk gene; The higher the value (the higher the priority of the gene), the smaller the P-like value, indicating the higher significance.

[0070] , (Equation 2)

[0071] represents the original connectivity score (disconnect score) of each crosstalk gene; each parameter ( and ) normalizes the output: A baseline fold change threshold is set (to ensure that FC-like has biological significance in connectivity), The input values are normalized to maintain the proportional relationship and prevent the FC-like score from being excessively inflated.

[0072] (2) System identification of upstream regulators

[0073] The present disclosure uses the converted FC-like and P-like indicators as inputs to systematically identify the upstream regulators of pathway crosstalk genes using the upstream regulator analysis tool of IPA. Two independent scores are calculated to assess their relevance (enrichment score) and regulatory status (directional Z score). For key regulators (such as top-ranked cytokines, transcription factors, or chemical drugs), further delineation of their specific regulatory effects (activation or inhibition) on individual target genes in the crosstalk network is provided.

[0074] (4) Prediction of mRNA therapy for senile degenerative bone disease

[0075] Various analyses performed on the present disclosure can find potential intervention strategies for different key network genes and their products:

[0076] (4-1) Application of mRNA in delivering gene therapy effect molecules

[0077] Gene editing therapy, as an important direction of gene therapy, has shown the potential for radical treatment of single-gene mutation related diseases (such as thalassemia). The core of gene editing therapy is to restore the normal function of genes by precise base addition, deletion or replacement. At present, the most commonly used gene editing tool is the CRISPR-Cas9 system developed based on the bacterial innate immune system. This system relies on guide RNA to guide Cas9 nuclease to recognize and cut specific DNA sites, thereby achieving repair or replacement of mutant genes. Although delivery methods based on viruses or plasmids have been widely used, there are still safety concerns. The delivery strategy of mRNA-based Cas9 or dCas9 is considered a safer and more flexible solution, and is expected to become an important direction for future precise gene editing therapy.

[0078] The present disclosure focuses on the role of core genes with genetic evidence supporting in bone diseases, and further proposes an mRNA-based gene editing delivery strategy to achieve precise repair of key genes related to senile degenerative bone diseases, to restore their function of regulating cell cycle and inhibiting abnormal proliferation, and to slow down the progression of bone diseases. Through such a method, the safety and controllability of gene editing are improved, and at the same time, the precise intervention of degenerative bone diseases is achieved by combining the core target genes identified by the present disclosure with genetic evidence.

[0079] (4-2) Application of mRNA in protein replacement therapy

[0080] mRNA-based protein replacement therapy has become an important direction of precision medicine, which delivers mRNA encoding specific functional proteins to make cells synthesize the desired proteins directly in the body, thereby making up for the loss of function caused by gene mutations or protein defects. This strategy has shown significant efficacy in ischemic diseases, hemophilia and various single-gene genetic diseases, providing a new idea for the treatment of degenerative diseases. mRNA therapy usually uses delivery systems such as lipid nanoparticles (LNP), polymer nanoparticles or exosomes to specifically deliver mRNA encoding therapeutic proteins to target cells, and express functional proteins in cells to regulate pathological processes.

[0081] The present disclosure focuses on the mechanism of core genes supported by regulatory evidence in bone diseases, and for their abnormal expression characteristics, mRNA delivery technology can be used for precise intervention. For example, for environment-dependent dysregulated genes, protective proteins can be supplemented to restore chondrocyte homeostasis. Among them, LNP-mRNA is the most promising delivery system for clinical application at present, which forms nanoparticles to encapsulate mRNA by ionizable lipids, and realizes efficient delivery by endosome escape mechanism, has the advantages of high loading capacity, low immunogenicity and repeatable administration, and has been confirmed to be effective in treating degenerative diseases and regulating key pathogenic pathways in multiple clinical trials.

[0082] (4-3) Application of mRNA in in vivo CAR-T cell therapy

[0083] CAR-T cell therapy is an immunotherapy that targets and eliminates pathological cells by modifying patient's own T cells to obtain specific killing ability. In recent years, in vivo CAR-T technology based on mRNA-LNP has gradually emerged as an important complementary solution to autologous CAR-T therapy. This strategy directly delivers mRNA encoding chimeric antigen receptor (CAR) to CD8+ T cells through lipid nanoparticles to generate functionally active CAR-T cells in situ, thereby avoiding the limitations of high cost, complex production process and chemotherapy pretreatment in traditional autologous CAR-T therapy. The representative product HN2301 uses engineered cell-targeted lipid nanoparticles (EnC-LNP) to efficiently and selectively deliver CD19 CAR mRNA to CD8+ T cells, showing significant B cell clearance effect in preclinical studies and good safety and therapeutic potential in autoimmune disease models such as systemic lupus erythematosus.

[0084] The present disclosure further discovers core intervention targets in degenerative bone diseases, which can be targeted by CAR-T mRNA delivery. By designing CAR constructs targeting these intervention targets, precise recognition and elimination of pathological cells can be achieved, thereby combining in vivo CAR-T technology with the intervention targets discovered by the present disclosure to provide a safe, efficient and controllable treatment strategy for immune intervention of degenerative diseases.

[0085] (4-4) Application of mRNA in vaccine design

[0086] mRNA-based immunotherapy strategies have become an important direction in the field of tumor intervention, including individualized mRNA vaccines, combined immune checkpoint inhibitor therapy, and combination approaches to deliver immune-activating factors, which are precise treatment methods with both specificity and universality. Usually, by encoding tumor-associated antigens or tumor-specific antigens, mRNA is targeted to deliver to dendritic cells and other efficient antigen-presenting cells, thereby activating T cell and B cell immune responses, and thus achieving tumor clearance. At the same time, the delivery of mRNA encoding IL-2, IL-12, IL-15 or GM-CSF and other immune factors can also enhance immune cell function and further enhance anti-tumor efficacy. Vaccine-based anti-aging intervention strategies have become a new direction for the prevention and treatment of age-related diseases, including preventive vaccines, therapeutic vaccines, and adjuvant enhancement therapies, which are comprehensive methods with both preventive and therapeutic potential. Vaccine therapy usually targets specific antigens (such as mutant proteins, abnormally expressed proteins, etc.) or combined immune-regulating adjuvants to activate specific immune responses to clear pathogenic factors or repair damaged tissue function.

[0087] The present disclosure focuses on the key role of disease-related antigens verified by multi-omics in disease progression, and can develop precise vaccine treatment strategies, such as "antigen-adjuvant" combination therapy, to enhance its function of inducing protective immune response, inhibiting pathological cartilage degeneration, and delaying the progression of bone disease.

[0088] According to one or more embodiments, the present disclosure proposes an mRNA therapy intelligent recommendation and evaluation framework targeting key lesion regulatory factors of senile degenerative bone disease, mainly including four parts: genetic intervention target prioritization, intervention target network identification, and advanced analysis based on intervention target network (perturbation removal analysis, regulatory hierarchy analysis), and the specific process is as shown in Figure 1 The input of the model includes genetic and regulatory data related to orthopedic diseases, including genetic association data, tissue-specific eQTL and PCHi-C data, single-cell and spatial transcriptome data.

[0089] The output 1 of the model includes prioritized intervention targets. The model prioritizes the intervention targets based on the obtained genetic evidence, regulatory evidence, and network evidence, generates a ranking list including genes and their corresponding priority indices, and the higher the priority index, the more likely the gene becomes a therapeutic target. Among them,

[0090] Genetic evidence includes the use of multi-modal regulatory genomics data (such as proximity, eQTL, PCHi-C) to identify genetic variations that may affect disease;

[0091] Regulatory evidence includes analyzing disease-related genomics data (single-cell and spatial transcriptome) to determine the genes that are active in specific cell types and tissue environments;

[0092] Network evidence includes assessing high-confidence interaction relationships between genes through methods such as restarting random walk algorithm.

[0093] The output 2 of the model includes intervention target networks and potential intervention means, i.e., obtaining gene interaction relationships from pathways to construct intervention target networks. Through key intervention target prediction and potential intervention means prediction, it can be determined which gene variations are critical for treatment and possible intervention methods (such as cell therapy, vaccination, etc.). Here, further analysis is also included, including perturbation removal analysis and regulatory hierarchy analysis. Perturbation removal analysis analyzes the stability and function of the entire gene network by simulating the removal of the influence of a specific gene. Regulatory hierarchy analysis determines the hierarchical position of genes in the regulatory network to identify master regulators and executors.

[0094] The following takes osteoarthritis as an example to further illustrate the implementation process of the present disclosure.

[0095] (1) Prioritization function of osteoarthritis genetic intervention targets, the process of target prioritization includes two key steps:

[0096] The first step is the preparation of predictors, which utilizes multi-modal regulatory genomic datasets to construct genomic evidence through proximity, QTL, and PCHi-C, etc., constructs regulatory genomics data through spatial and single-cell transcriptome resources, and constructs network evidence through high-quality protein interaction knowledge using RWR algorithm.

[0097] The second step is the combination of predictors, and benchmarking is performed on different prioritization schemes, i.e., the scheme of meta-analysis combination. By comparing different schemes to identify the ability of the target supported by clinical theory, the experimental results show that the performance of the order statistic scheme is significantly better under the same parameter setting (optimized with a restart probability of 0.7).

[0098] (2) Identification of osteoarthritis intervention target networks. The present disclosure utilizes prioritized genes for intervention target network analysis to identify sub-networks related to osteoarthritis disease progression to determine potential intervention targets, such as Figure 2The intervention target network for osteoarthritis shown, results in an intervention target network of 50 genes with high priority scores and interconnections. Among these 50 intervention target network genes, a subset of genes have been validated to be associated with the pathogenesis of osteoarthritis or existing therapeutic strategies. For example, TNF is a key driver of osteoarthritis, and anti-TNF-a strategies have shown dual anti-inflammatory and cartilage-protective effects in preclinical models. For example, adalimumab, a TNF-a inhibitor, has entered phase II clinical trials for osteoarthritis. Another example is YAP1, which is highly expressed in fibrocartilage chondrocytes. YAP / TAZ transcriptional coactivators are central hubs that translate mechanical signals into cellular responses. Under pathological mechanical loading characteristic of osteoarthritis (e.g., excessive joint stress), Hippo pathway inhibition leads to YAP / TAZ nuclear translocation, promoting chondrocyte apoptosis and extracellular matrix (ECM) degradation, while also modulating the differentiation potential of mesenchymal stem cells (MSCs). These data demonstrate the effectiveness of network identification.

[0099] Figure 2 The intervention target network for osteoarthritis, which is derived from pathway- derived gene interactions, is limited by intervention target priority information. The network contains a set of highly prioritized and interconnected genes, with nodes labeled as "Gene Name@Priority Score" and colored by priority score. TNF and YAP1 are bolded.

[0100] (3) Advanced analysis based on intervention target networks, including

[0101] (301) Perturbation removal analysis

[0102] Perturbation removal analysis assesses the extent of disruption to the molecular network integrity by single node or node combination removal based on network connectivity, and removal of key nodes or node combinations leads to significant fragmentation of the largest connected component (LCC) of the molecular network. The present disclosure can systematically assess the vulnerability of the network to single node, double node, and triple node combination removal through a three-stage node attack analysis, calculate the stability of the network while identifying core targets and their combinations for osteoarthritis treatment, and determine the most disruptive set of key genes to the integrity of the molecular network Figure 3). Notably, in the single-node removal phase, the removal of YAP1 led to the greatest network fragmentation, with 6% of nodes disconnected from the LCC, which is consistent with known research: YAP1 plays an important hub role in sensing mechanical changes in the cellular microenvironment and mediating multi-level stress responses, so this gene can be a potential intervention target for osteoarthritis. Further, in the double-node removal phase, simultaneous targeting of YAP1 and AKT1 can produce the greatest synergistic disruption effect (12% of nodes disconnected from the LCC). Finally, in the three-node removal phase, the introduction of clinically validated targets improves the expected efficacy and safety. Among them, the removal of YAP1+AKT1+TNF can make the network fragmentation reach a peak (14% of nodes disconnected from the LCC), fully demonstrating that these gene combinations play a core role in maintaining the stability of the osteoarthritis molecular network through synergistic regulation, providing a theoretical basis for the development of multi-target combination therapy.

[0103] As Figure 3 Based on the perturbation removal analysis of the intervention target network, the x-axis represents the removal of a single node or node combination, and the y-axis shows the proportion of nodes disconnected from the network after the corresponding nodes are removed individually or jointly. For the removal of two or three nodes at the same time, only the removal combination with the greatest impact is shown. The left insert is a visualization of the intervention target network, which has the same layout as Figure 2 but only marks the genes with the best removal effect. The right insert is an explanation of the three scenarios of perturbation removal analysis.

[0104] (302) Analysis of the regulatory hierarchy of intervention targets

[0105] The present disclosure conducts a regulatory hierarchy analysis of the intervention target network (FIG. 2B). Figure 4 Ingenuity Pathway Analysis (QIAGEN-IPA) was used to analyze upstream regulatory factors and to analyze the hierarchical regulatory pattern of signal pathway crosstalk in osteoarthritis. First, focus on the target with therapeutic prospects. YAP1 was determined to be the transcription factor with the highest degree of activation, directly or indirectly regulating 23 genes in the network. Further analysis showed that YAP1 not only has self-activation ability, but also directly regulates key nodes in multiple signal pathways. These nodes include cytoskeletal regulatory factors (ACTB, SRC), inflammation hubs (TNF, STAT1, STAT3, RELA, NFKBIA), catabolic regulators (SMAD3), transcriptional coactivator complexes (EP300 and CTNNB1), and it indirectly activates JUN. The role of YAP1 in sensing mechanical changes in the cellular microenvironment and mediating multi-level stress responses suggests that it is a valuable target for intervention in the treatment of osteoarthritis in the elderly.

[0106] The present disclosure simultaneously analyzes and identifies chemical drugs that are promising for the treatment of osteoarthritis in the elderly. Metformin, a drug that has been approved for the treatment of type 2 diabetes, shows a special drug repurposing potential for osteoarthritis. In obesity-related knee osteoarthritis, which involves the interaction between biomechanical stress and metabolic inflammation, animal studies have shown that metformin reduces pain and protects joints through multiple mechanisms. These mechanisms include anti-inflammatory effects, metabolic regulation, and improved insulin resistance. Clinical evidence further confirms its therapeutic effect on pain and joint function in overweight or obese patients with knee osteoarthritis, supporting its cartilage protection and symptom improvement. The analysis of the present disclosure shows that metformin acts on multiple components of the osteoarthritis crosstalk network, notably the inhibition of TNF highlights its anti-inflammatory properties, further highlighting its potential for elderly treatment in the treatment of osteoarthritis.

[0107] As Figure 4 Regulatory hierarchy analysis based on intervention target network. The regulatory effects of upstream activator YAP1 and small molecule inhibitor metformin on genes in the intervention target network are shown in the figure.

[0108] Based on the above analysis results, as Figure 5 The summary of multiple analyses, through the flowchart, shows that the aforementioned perturbation removal analysis, druggability analysis, and regulatory hierarchy analysis all point to YAP1 as the most potential osteoarthritis intervention target and potential intervention means. Therefore, the following conclusions can be drawn:

[0109] 1) YAP1 exhibits a significant disruptive effect on the osteoarthritis molecular network, and the regulatory hierarchy analysis reveals that YAP1, as a core upstream regulator, is involved in the fine mechanical-biochemical signal cascade regulation. Multiple evidences establish the important value of YAP1 in the treatment of osteoarthritis;

[0110] 2) Targeted degradation of SMAD3 and EP300 using PROTACs can break through the technical bottleneck that traditional drugs are difficult to inhibit transcriptional co-activation complexes, providing a new solution for osteoarthritis disease modification treatment;

[0111] 3) Metformin mediates multiple effects in osteoarthritis, including anti-inflammatory effects, metabolic reprogramming, and cartilage protection, and has important direct repurposing potential. In the future, new interventions for osteoarthritis can be achieved through targeted means of key genes and their products, for example:

[0112] (1) Application of mRNA in delivering gene therapy effect molecules: CRISPR-Cas9 is the most widely used gene editing technology at present, which restores the normal function of genes by precise base addition, deletion or replacement and is currently widely used in the treatment of various genetic diseases. Compared with traditional delivery methods relying on adeno-associated virus (AAV) or plasmids, the use of mRNA as a delivery vector avoids the safety hazards of random DNA integration, while avoiding the shortcomings of RNA-protein complex delivery in tissue penetration and in vivo efficiency. For the priority genes identified in the present disclosure, such as SMAD3 and EP300, which have genetic evidence support, CRISPR sequences can be used as guide RNA to accurately recognize and cut the specific position of DNA through Cas9 protein, thereby repairing or replacing the gene at this position and slowing down the disease progression.

[0113] (2) Application of mRNA in protein replacement therapy: Protein replacement technology is usually used to treat genetic diseases or protein deficiency or dysfunction caused by exogenous factors. mRNA-based protein replacement therapy has unique advantages in precise genetic intervention, including high safety, high efficiency, controllable delivery and non-immunogenicity, which is based on the breakthrough of structural optimization. mRNA therapy usually delivers mRNA encoding therapeutic proteins to target cells through delivery systems such as lipid nanoparticles (LNP), polymer nanoparticles or exosomes, and expresses functional proteins in cells to regulate pathological processes. For the environment-dependent dysregulated genes identified in the present disclosure, such as YAP1, the pathogenic gene can be silenced and protective proteins can be supplemented simultaneously to achieve dynamic regulation, thereby restoring chondrocyte homeostasis. Local delivery of cartilage anabolic factors (IGF-1) or anti-inflammatory / immune modulating factors (such as TGF-β) is expected to promote the repair and regeneration of cartilage while inhibiting chronic inflammation that exacerbates cartilage degradation.

[0114] (3) Application of mRNA in in vivo CAR-T cell therapy: CAR-T cell therapy is an immunotherapy that targets and eliminates pathological cells by modifying patient's own T cells to obtain specific killing ability. In recent years, in vivo CAR-T technology based on mRNA-LNP has gradually emerged as an important supplement to autologous CAR-T therapy. This strategy delivers mRNA encoding chimeric antigen receptor (CAR) to CD8+ T cells via lipid nanoparticles (LNP) to generate functionally active CAR-T cells in situ, thereby avoiding the limitations of high cost, complex production process and chemotherapy pretreatment in traditional autologous CAR-T therapy. In this disclosure, the disclosure found that KRAS and HRAS were significantly overexpressed in osteoarthritis patients through the integration of regulatory genomics data and gene expression analysis, which may serve as potential neoantigens and become candidate targets for targeted immune intervention. Based on this finding, these intervention targets can be used as objects for CAR-T mRNA delivery, and by designing CAR constructs targeting KRAS, HRAS or other core pathological factors, precise recognition and elimination of pathological cells can be achieved.

[0115] (4) Application of mRNA in vaccine design: mRNA-based vaccines are a promising strategy for preventing and treating age-related diseases, including preventive vaccines, therapeutic vaccines, and adjuvant-enhanced therapies. They are a comprehensive approach that combines prevention and treatment potential. The intervention targets identified in this disclosure can be used to design mRNA encoding key pathological proteins or antigens to achieve precise vaccine intervention for osteoarthritis and related degenerative diseases. Combined with the "vaccine + adjuvant therapy" strategy, the immune effect of mRNA vaccine can be further enhanced. Metformin is used as a potential adjuvant in this disclosure, which not only has direct potential for reapplication, but also can play an immunomodulatory role: enhancing memory-like CD8+ T cell responses and improving vaccine immunogenicity in animal models; restoring B cell function and improving immune aging phenotype in elderly people with type 2 diabetes; enhancing influenza vaccine response in non-diabetic elderly people. These results highlight the potential of metformin as an mRNA vaccine adjuvant, enabling it to enhance immune responses, improve immune function related to degenerative diseases, and provide new ideas for the design of osteoarthritis anti-aging vaccines.

[0116] The present disclosure proposes an mRNA therapy intelligent recommendation and evaluation model targeting key lesion regulatory factors of senile degenerative bone disease, which has the following core advantages: comprehensive integration of regulatory genomics data sets, including genetic information, protein interaction networks, and disease-associated functional genomics information, to identify core and peripheral genes that mediate environmental risk factors and pathological processes in specific tissues or cell types. This integration strategy not only overcomes the target identification bias caused by single data source analysis in the prior art, but also captures key pathological regulatory hubs driven by non-genetic factors such as aging, obesity, and joint damage, enabling high-value intervention target prediction for degenerative diseases such as osteoarthritis. Based on the core target, the present disclosure further proposes the application of various mRNA therapies:

[0117] Application of mRNA in gene therapy effector molecule delivery: By delivering mRNA encoding key genes or regulatory factors, rapid and controllable protein expression can be achieved in target cells, thereby repairing pathological gene mutations or restoring the function of abnormally expressed core factors. Compared to traditional viral or plasmid delivery methods, the mRNA vector used in the present disclosure has low cost, simple structure, can be repeatedly administered, and eliminates the risk of random DNA integration, ensuring efficient expression and safety in vivo;

[0118] Application of mRNA in protein replacement therapy: For joint degeneration caused by the absence or impaired function of key proteins, protein replacement can be achieved by delivering mRNA molecules encoding functional proteins, directly supplementing the missing functional proteins, thereby restoring chondrocyte homeostasis and metabolic balance and improving tissue repair capacity;

[0119] Application of mRNA in in vivo CAR-T cell therapy: The present disclosure can deliver mRNA encoding CAR structures that recognize pathological targets to CD8+ T cells to generate functional CAR-T cells in situ in vivo, thereby accurately eliminating pathogenic cells, overcoming the high cost and chemotherapy pretreatment limitations of traditional autologous CAR-T therapy, and improving the operability and safety of treatment;

[0120] Application of mRNA in vaccine design: For abnormal proteins or pathological antigens related to bone disease, the present disclosure can design mRNA vaccines to achieve immune intervention of key pathological pathways, activate T cells and B cells to produce specific immune responses, and regulate the local inflammatory environment to delay disease progression;

[0121] In addition, the disclosure also proposes an advanced analysis method based on intervention target network, including perturbation removal analysis and regulatory hierarchy analysis, to ensure the functional certainty of the identified target. Network attack analysis can identify synergistic combination targets to maximize intervention effects, while optimizing mRNA delivery design and improving treatment efficiency. In summary, the disclosure as a regulatory omics data-driven target prediction tool driven by network interaction, can help identify intervention targets for a series of senile degenerative diseases, and promote the development of various mRNA-based treatment strategies. These strategies target the underlying mechanisms of joint degeneration associated with aging, rather than just managing symptoms, accelerating the development of precise geriatric treatment methods.

[0122] It should be understood that in the embodiments of the disclosure, the term "and / or" only describes the association relationship of the associated objects, which means that there can be three relationships. For example, A and / or B can mean that A exists alone, A and B exist together, and B exists alone. In addition, the character " / " herein generally represents that the front and rear associated objects are in an "or" relationship.

[0123] If the integrated unit is implemented in the form of a software function unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the disclosure essentially or say the part of the prior art that contributes to the prior art, or the entire or part of the technical solutions can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute all or part of the steps of the method described in various embodiments of the disclosure. The aforementioned storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0124] The above is only a specific implementation of the disclosure, but the protection scope of the disclosure is not limited thereto. Any person skilled in the art can easily think of various equivalent modifications or replacements within the technical scope disclosed by the disclosure, and these modifications or replacements should be covered within the protection scope of the disclosure. Therefore, the protection scope of the disclosure should be subject to the protection scope of the claims.

Claims

1. A method for evaluating and recommending mRNA therapy targeting key regulatory factors in age-related degenerative bone diseases, characterized in that, By recommending and evaluating models, intervention targets for age-related orthopedic degenerative diseases are predicted, leading to mRNA-based therapies that target key regulatory factors in age-related degenerative bone diseases. This model includes... The data input layer receives genetic association and regulatory data on bone diseases, including GWAS summary data, spatial transcriptomics, and single-cell transcriptomics data. The intervention target prioritization module integrates the obtained genetic evidence, regulatory evidence, and network evidence to prioritize intervention targets. The intervention target identification network module predicts key intervention targets and obtains potential mRNA intervention therapies based on the intervention target network generated from gene interaction relationships obtained from pathways.

2. The method according to claim 1, characterized in that, Genetic evidence includes sets of genetic risk sites annotated through multimodal regulatory genomics data; regulatory evidence includes disease-related functional genomics data; and network evidence includes high-confidence protein-protein interactions.

3. The method according to claim 2, characterized in that, The multimodal regulatory genomics data include proximity, eQTL, and PCHi-C; the regulatory evidence includes single-cell and spatial transcriptomes; and the network evidence includes high-confidence interactions identified through a restart random walk algorithm.

4. The method according to claim 1, characterized in that, The intervention target identification network module further includes perturbation removal analysis and regulation level analysis in its analysis of the intervention target network.

5. The method according to claim 1, characterized in that, The mRNA intervention therapy is used in cell therapy and vaccination.

6. The method according to claim 1, characterized in that, The intervention target priority ranking module uses a restarted random walk algorithm to calculate the affinity score of genes, forming a gene-predictor matrix, and integrates the scores through sequential statistics to prioritize candidate genes.

7. The method according to claim 6, characterized in that, The intervention target identification network module performs intervention target network analysis based on priority ranking results, and uses the reward collection Stanley tree algorithm to identify key intervention subnetworks with high-scoring nodes in the network integrated by the KEGG pathway.

8. The method according to claim 7, characterized in that, The intervention target identification network module performs perturbation removal and causal network analysis on genes in the intervention sub-network, assesses the impact of single or multiple genes on network stability, analyzes the regulatory hierarchy of pathological mechanisms, and identifies core intervention nodes; combining genetic evidence and functional genomics regulatory characteristics, it identifies targets with mRNA therapeutic potential.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the method as described in any one of claims 1 to 8.

10. A computer program product, comprising a computer program, characterized in that, The computer program is executed by a processor to implement the method according to any one of claims 1 to 8.

Images