Cancer risk prediction and personalized vaccine preparation system and method based on single sample DNA sequencing

Abstract

The invention belongs to the technical field of biological medicine, and discloses a cancer risk prediction and personalized vaccine preparation system and method based on single sample DNA sequencing, and the system comprises: a, a microsampling module; b, a multi-omics analysis engine; the engine synchronously executes the following operations: identifying a cfDNA terminal motif mode through a convolutional neural network; the presentation probability of the HLA-II type molecules is predicted on the basis of a Transform architecture; quantifying tumor mutation load (TMB) and microsatellite instability (MSI); c, a vaccine synthesis device; the device comprises: a programmable microfluidic mRNA synthesis chip; a self-assembly unit of lipid nanoparticles (LNP); and d, a closed-loop feedback system: monitoring the IFN-gamma through a wearable device to dynamically adjust the vaccine dosage. According to the invention, the early canceration detection rate can be improved, the antigen prediction accuracy is improved, the timeliness demand of early cancer intervention is met, and the adverse reaction rate is reduced.

Description

Technical Field

[0001] The present invention belongs to the field of biomedical technologies, and particularly relates to a cancer risk prediction and personalized vaccine preparation system and method based on single-sample DNA sequencing. Background Art

[0002] Cancer is the second leading cause of death globally. In 2023, there were over 20 million newly diagnosed cancer cases worldwide, and nearly 50% of the patients were in the middle or advanced stage at the time of diagnosis, with a 5-year survival rate of less than 30%. According to statistics in *Nature Medicine*, approximately 60% of cancer patients develop drug resistance due to tumor evolution caused by treatment delay, ultimately leading to treatment failure. Therefore, there is an urgent need for a non-invasive, highly sensitive, and rapidly responsive integrated diagnosis and treatment solution.

[0003] Existing Technical Bottlenecks and Breakthrough Directions: 1. Cancer Early Screening Technology: Limitations of One-Dimensional Analysis Current liquid biopsy technologies mainly focus on the detection of somatic mutations in ctDNA (such as TMB, MSI), but the abundance of driver mutations in early cancer is extremely low (VAF < 0.1%), and there are the following defects: Absence of epigenetic signals: DNA methylation abnormalities occur earlier than mutations (for example, hypermethylation of the SFRP2 gene in colorectal cancer is 5 - 8 years earlier than APC mutations), but existing technologies (such as PCR-targeted methylation detection) cover limited sites (<1% CpG islands).

[0004] Fragmentation patterns not utilized: Tumor-derived ctDNA has a specific nucleosome occupancy pattern (such as short fragment enrichment), but traditional sequencing technologies cannot analyze the terminal motif characteristics.

[0005] 2. Personalized Vaccine Development: Long Cycle and Low Efficiency There are two major shortcomings in the existing neoantigen vaccine preparation process: Low antigen prediction accuracy: It only relies on the HLA-I molecule presentation model (covering 40% of the immune response), ignoring the HLA-II molecule (key for CD4+ T cell activation) and the immune escape mechanism of epigenetic regulation.

[0006] Low preparation efficiency: mRNA in vitro synthesis relies on manual operation (error rate > 1%), and the encapsulation efficiency of the LNP encapsulation process is less than 80%, resulting in high production costs and a long cycle. Summary of the Invention

[0007] To overcome the above technical problems, the present invention provides a cancer risk prediction and personalized vaccine preparation system and method based on single-sample DNA sequencing.

[0008] The present invention adopts the following technical solutions: A cancer risk prediction and personalized vaccine preparation system based on single-sample DNA sequencing, comprising: a. A micro-sampling module; the module integrates a nano-needle array and a surface-enhanced Raman spectroscopy (SERS) probe; b. A multi-omics analysis engine; the engine synchronously performs the following operations: Identifying cfDNA terminal motif patterns through a convolutional neural network; Predicting the presentation probability of HLA class II molecules based on the Transformer architecture; Quantifying tumor mutational burden (TMB) and microsatellite instability (MSI); c. A vaccine synthesis device; the device includes: A programmable microfluidic mRNA synthesis chip; A self-assembly unit of lipid nanoparticles (LNP); d. A closed-loop feedback system: monitoring IFN-γ through a wearable device and dynamically adjusting the vaccine dose.

[0009] Preferably, the silicon-based nano-needles have a needle diameter of 80±5 nm, a cone angle of 15-20°, a needle density of 200-500 needles / cm², and the surface is modified with aptamers SEQ ID NO: 6-10.

[0010] Preferably, the mRNA synthesis chip: modular assembly: 5'-cap (CleanCap®)-UTR-antigen coding region (codon-optimized CAI>0.8)-3'-tail (polyA), error rate <0.1%.

[0011] Preferably, the LNP encapsulation unit: Lipid formulation: DSPC: cholesterol: PEG-DMG = 50:40:10 (particle size 80-100 nm, encapsulation efficiency>95%); Buffer: 10 mM Tris-HCl (pH 7.4) + 150 mM NaCl.

[0012] Preferably, wearable monitoring: a serum IFN-γ sensor, optimizing vaccine injection through a dose adjustment algorithm: .

[0013] The present invention also discloses a cancer risk prediction method based on single-sample DNA sequencing, comprising the following steps: S1. Collect 50-200 μL of peripheral blood, and use a micro-column containing silanized magnetic beads to separate ctDNA, with a recovery rate ≥90%; S2. Perform whole-genome methylation sequencing (covering ≥5M CpG sites) and somatic mutation detection (low-frequency mutation identification limit 0.01%) on ctDNA; S3. Calculate the cancer risk through the following algorithm model:

[0014] where W i is the pathway weight (trained with TCGA data), F i is the feature frequency, and λ is the immune regulation coefficient.

[0015] Among them, known pathogenic mutations in the BRCA1 / 2 genes are excluded.

[0016] The present invention also discloses a preparation method for personalized vaccines based on single-sample DNA sequencing, including the following steps: S1. Antigen screening and optimization Priority rules: Exclude mutant peptides with a homology > 80% to self-antigens; preferably select clonal mutations (tumor cell proportion > 20%); Combination design: Use the greedy algorithm to select 5 - 10 antigens covering 90% of tumor clones; S2. mRNA vaccine synthesis Sequence design: Insert the Kozak sequence (GCCACC) into the 5' UTR and integrate the miR-122 inhibitory element into the 3' UTR; Avoid forming a stable secondary structure (ΔG > -10 kcal / mol) in the antigen coding region; LNP encapsulation verification: Detect the particle size distribution by dynamic light scattering (DLS) (PDI < 0.2); Verify IFN-γ secretion by in vitro ELISpot, and > 50 SFC / 10^6 PBMC is qualified.

[0017] Compared with the prior art, the beneficial effects of the present invention are: Multi-omics integrated analysis: Simultaneously analyze the methylation map of ctDNA (covering ≥ 5M CpG sites), fragmentation pattern (CNN identifies the terminal motif), and somatic mutations (VarScan2 algorithm, detection limit 0.01%), improving the early cancer detection rate; Multi-modal antigen prediction model: Based on the Transformer architecture, fuse the HLA-I / II presentation probability, methylation-regulated immunogenicity score (IC50 < 500 nM), and the proportion of clonal mutations, improving the antigen prediction accuracy; Fully automated microfluidic synthesis: Integrate a programmable mRNA synthesis chip (error rate < 0.1%) and an LNP self-assembly unit (encapsulation efficiency > 95%), compress the vaccine preparation cycle to 72 hours, and meet the timeliness requirements for early cancer intervention; Dynamic Closed-Loop Therapy: From "Static" to "Real-Time" Continuously monitor the serum IFN-γ level through a wearable device, and combine it with a dose adjustment algorithm to dynamically optimize the efficacy of the vaccine and reduce the adverse reaction rate. Specific Embodiment

[0018] The embodiments of the present invention are described in detail below. Unless otherwise specified, the raw materials and equipment used can be purchased from the market or are commonly used in the art. The methods in the embodiments, unless otherwise specified, are conventional methods in the art. The embodiments described below are exemplary and are only used to explain the present invention and should not be construed as a limitation to the present invention.

[0019] I. System Architecture 1. Micro-Sampling Module Nanoneedle Array: Needle Body Parameters: Silicon-based nanoneedles (needle diameter 80±5nm, cone angle 15 - 20°, needle density 200 - 500 needles / cm²), surface modified with aptamers (SEQ ID NO:6 - 10).

[0020] Function Integration: SERS probe for real-time monitoring of ctDNA capture efficiency (sensitivity 0.1pg / μL).

[0021] ctDNA Separation: Silanized magnetic column (recovery rate ≥90%) combined with a microfluidic chip to remove blood cell debris.

[0022] 2. Multi-Omics Analysis Engine Epigenetic Analysis: Whole Genome Bisulfite Sequencing (WGBS): Covering ≥5M CpG sites, detecting promoter region hypermethylation (threshold Δβ≥0.25).

[0023] cfDNA Fragmentation Pattern: CNN to identify abnormal nucleosome occupancy (such as enrichment of 5'-CCWGG-3' end motif).

[0024] Genomic Variant Detection: Somatic Mutation Screening: VarScan2 algorithm (low-frequency mutation detection limit 0.01%), quantifying TMB (≥10 mut / Mb is high risk) and MSI status.

[0025] Neoantigen Prediction: Multi-Modal AI Model: Transformer architecture to predict the HLA-II presentation probability (threshold >0.8).

[0026] Fusion of NetMHCpan 4.0 to calculate the immunogenicity score (IC50 <500nM is positive).

[0027] 3. Vaccine Preparation Device mRNA Synthesis Chip: Modular Assembly: 5' Cap (CleanCap®) - UTR - Antigen Encoding Region (Codon Optimized CAI > 0.8) - 3' Tail (polyA), Error Rate < 0.1%.

[0028] LNP Encapsulation Unit: Lipid Formula: DSPC: Cholesterol: PEG - DMG = 50:40:10 (Particle Size 80 - 100nm, Encapsulation Efficiency > 95%).

[0029] Buffer: 10mM Tris - HCl (pH7.4) + 150mM NaCl.

[0030] 4. Dynamic Feedback System Wearable Monitoring: Serum IFN - γ Sensor (Detection Limit 0.5pg / mL), Optimizing Vaccine Injection through Dose Adjustment Algorithm: .

[0031] II. Cancer Risk Prediction Method S1. Collect 50 - 200μL of peripheral blood, separate ctDNA using a microcolumn with silanized magnetic beads, and the recovery rate ≥ 90%; S2. Perform whole - genome methylation sequencing (covering ≥ 5M CpG sites) and somatic mutation detection (low - frequency mutation recognition limit 0.01%) on ctDNA; Exclude known pathogenic mutations in the BRCA1 / 2 genes; S3. Calculate the carcinogenesis risk through the following algorithm model: .

[0032] Parameter Definition: Wi: Weights of 20 pathways trained based on TCGA (e.g., the weight of the PI3K - AKT pathway = 1.32).

[0033] Fi: Methylation Entropy (threshold ≥ 0.35), TMB, MSI, and cfDNA concentration (> 5ng / mL).

[0034] λ: HLA Diversity Regulation Coefficient (heterozygous HLA - I / II genotype λ = 1.2).

[0035] Risk Classification: Low Risk (Score < 30): Annual Follow - up.

[0036] Medium Risk (30 ≤ Score < 60): 3 - month Dynamic Monitoring of ctDNA.

[0037] High risk (Score≥60): Initiate the preparation of preventive vaccines.

[0038] III. Personalized Vaccine Preparation Method 1. Antigen Screening and Optimization Priority Rules: Exclude mutant peptides with a homology of >80% to self-antigens.

[0039] Preferably select clonal mutations (tumor cell proportion >20%).

[0040] Combinatorial Design: Use the greedy algorithm to select 5 - 10 antigens that cover 90% of tumor clones.

[0041] 2. mRNA Vaccine Synthesis Sequence Design: Insert the Kozak sequence (GCCACC) into the 5'UTR and integrate the miR-122 inhibitory element into the 3'UTR.

[0042] Avoid the formation of stable secondary structures (ΔG > -10 kcal / mol) in the antigen coding region.

[0043] LNP Encapsulation Verification: Use dynamic light scattering (DLS) to detect the particle size distribution (PDI < 0.2).

[0044] Verify IFN-γ secretion by in vitro ELISpot (>50 SFC / 10^6 PBMC is qualified).

[0045] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to the above embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the claims and their equivalents.

Claims

1. A cancer risk prediction and personalized vaccine preparation system based on single-sample DNA sequencing, characterized in that Comprising: a. Micro-sampling module; the module integrates a nano-needle array and a surface-enhanced Raman spectroscopy (SERS) probe; b. Multi-omics analysis engine; The engine synchronously performs the following operations: Identifying cfDNA terminal motif patterns through a convolutional neural network; Predicting the presentation probability of HLA class II molecules based on the Transformer architecture; Quantifying tumor mutational burden (TMB) and microsatellite instability (MSI); c. Vaccine synthesis device; the device includes: A programmable microfluidic mRNA synthesis chip; Self-assembly unit of lipid nanoparticles (LNP); d. Closed-loop feedback system: Monitoring IFN-γ through a wearable device to dynamically adjust the vaccine dose.

2. The cancer risk prediction and personalized vaccine preparation system based on single-sample DNA sequencing according to claim 1, wherein The silicon-based nano-needles have a needle diameter of 80 ± 5 nm, a cone angle of 15 - 20°, a needle density of 200 - 500 needles / cm², and the surface is modified with aptamers SEQ ID NO: 6 - 10.

3. The cancer risk prediction and personalized vaccine preparation system based on single-sample DNA sequencing according to claim 1, characterized in that mRNA synthesis chip: Modular assembly: 5'-cap (CleanCap®)-UTR-antigen coding region (codon-optimized CAI > 0.8)-3'-tail (polyA), error rate < 0.1%.

4. The cancer risk prediction and personalized vaccine preparation system based on single-sample DNA sequencing according to claim 1, characterized in that LNP encapsulation unit: Lipid formulation: DSPC:cholesterol:PEG-DMG = 50:40:10 (particle size 80 - 100 nm, encapsulation efficiency > 95%); Buffer: 10 mM Tris-HCl (pH 7.4) + 150 mM NaCl.

5. The cancer risk prediction and personalized vaccine preparation system based on single-sample DNA sequencing according to claim 1, characterized in that Wearable monitoring: Serum IFN-γ sensor, optimizing vaccine injection through a dose adjustment algorithm: 。 6. A cancer risk prediction method based on single-sample DNA sequencing, characterized in that, Including the following steps: S1. Collect 50 - 200 μL of peripheral blood, and use a micro-column containing silanized magnetic beads to separate ctDNA, with a recovery rate ≥ 90%; S2. Perform whole-genome methylation sequencing (covering ≥ 5M CpG sites) and somatic mutation detection (low-frequency mutation identification limit 0.01%) on ctDNA; S3. Calculate the carcinogenesis risk through the following algorithm model: Among them, W i is the pathway weight (trained with TCGA data), F i is the feature frequency, and λ is the immune regulation coefficient.

7. The cancer risk prediction method based on single-sample DNA sequencing according to claim 6, wherein Excluding known pathogenic mutations in the BRCA1 / 2 genes.

8. A preparation method for personalized vaccine preparation based on single-sample DNA sequencing, characterized in that, Including the following steps: S1. Antigen screening and optimization Priority rules: Excluding mutant peptides with a homology > 80% to self-antigens; preferably clonal mutations (tumor cell proportion > 20%); Combination design: Using a greedy algorithm to select 5 - 10 antigens covering 90% of tumor clones; S2. mRNA vaccine synthesis Sequence design: Insert the Kozak sequence (GCCACC) into the 5'UTR, and integrate the miR-122 inhibitory element into the 3'UTR; Avoid forming a stable secondary structure in the antigen coding region (ΔG > -10 kcal / mol); LNP encapsulation verification: Detecting the particle size distribution by dynamic light scattering (DLS) (PDI < 0.2); Verifying IFN-γ secretion in vitro by ELISpot, with > 50 SFC / 10^6 PBMC being qualified.

Images