Multimodal in vivo genetic collaborative transformation system

Abstract

This utility model discloses a multimodal live genetic collaborative transformation system, relating to the fields of plant genetic transformation and molecular breeding technology. Specifically, it is a multimodal live genetic collaborative transformation system, comprising a housing with a glass observation door at the front end and a glass window on the end face of the observation door. A plastic frame is provided at the connection between the observation door and the glass window, and a handle is provided on the end face of the observation door. A display screen is located on the upper right side of the observation door, and the display screen is integrated with the housing. Control buttons are located below the display screen. This system enables visual observation of the experimental process, facilitating real-time monitoring of the plant material's status. The casters provide the system with flexible mobility to adapt to different experimental scenarios. The overall structure balances sealing, observation, and mobility, significantly improving the convenience of experimental operation.

Description

Technical Field

[0001] This invention relates to the field of plant genetic transformation and molecular breeding technology, specifically a multimodal living genetic synergistic transformation system. Background Technology

[0002] Plant genetic transformation and molecular breeding are core technologies for modern crop improvement. Their goal is to improve plant traits such as stress resistance, yield, and quality through methods such as the introduction of exogenous genes, gene editing, or germplasm resource hybridization.

[0003] However, existing patents have the following drawbacks: (1) Low efficiency and insufficient targeting of genetic transformation: Traditional gene introduction methods (such as Agrobacterium-mediated transformation and gene gun method) rely on random insertion, which can easily lead to gene silencing or abnormal expression; and the transformation process is mostly an in vitro operation, which makes it difficult to achieve in situ transformation of living plants, thus limiting the assessment of the impact on the physiological state of intact plants.

[0004] (2) Lack of multimodal collaborative processing capability: Existing equipment focuses on a single function (such as gene transformation or mutagenesis treatment only), lacking the ability to coordinate and regulate multiple links such as gene introduction, cell hybridization, and radiation mutagenesis; and the environmental parameters (such as temperature, humidity, and gas) between different treatments are difficult to control synchronously, and the experimental results are easily affected by environmental fluctuations.

[0005] (3) Insufficient precision of mutagenesis treatment: Traditional radiation mutagenesis (such as gamma rays and X-rays) relies on fixed dose irradiation and cannot dynamically adjust the dose according to the tolerance of plant materials, which can easily lead to excessive cell damage or even death, making it difficult to achieve the balance between "mild mutagenesis and effective mutation".

[0006] (4) Single phenotypic data collection: Existing phenotypic analysis relies on manual observation or a single sensor (such as only spectral imaging), making it difficult to simultaneously acquire multi-dimensional data such as physiological traits (such as chlorophyll content) and mechanical traits (such as stem strength), and thus cannot fully reflect the relationship between genotype and phenotype.

[0007] (5) Low data processing and analysis efficiency: In traditional breeding, phenotypic-genotypic association analysis requires manual integration of multi-source data (such as sequencing results and phenotypic records), which is time-consuming, labor-intensive, and prone to subjective errors, making it difficult to quickly screen mutants of the target trait. Utility Model Content

[0008] The purpose of this invention is to provide a multimodal live genetic collaborative transformation system. By integrating modules such as gene introduction, cell hybridization, radiation mutagenesis, phenotypic acquisition, and intelligent analysis, and combining environmental regulation and precision operation technology, it can achieve the whole-process collaborative processing of plant live cell genetic transformation, improve transformation efficiency, accuracy, and data integration capabilities, and provide efficient technical support for modern molecular breeding.

[0009] To achieve the above objectives, this utility model provides the following technical solution: a multimodal in vivo genetic collaborative transformation system, comprising a housing, a glass observation door at the front end of the housing, a glass window on the end face of the glass observation door, a plastic frame at the connection between the glass observation door and the glass window, a handle on the end face of the glass observation door, a display screen on the upper right side of the glass observation door, the display screen being integrated with the housing cabinet, control buttons below the display screen, a cabinet door on the end face of the housing, and casters at the lower end of the housing.

[0010] Furthermore, the enclosure integrates a gene introduction module, a cell hybridization device, a radiation treatment unit, a high-throughput phenotypic acquisition system, and a data analysis platform. The gene introduction module is used to introduce exogenous genes into target plant cells through in vivo transfection technology. The cell hybridization device is configured to achieve protoplast fusion of different species or varieties in a sterile environment. The radiation treatment unit uses a gamma-ray or X-ray generator to induce mutagenesis in plant materials. The high-throughput phenotypic acquisition system integrates spectral imaging and multispectral sensors to monitor plant physiological traits in real time. The data analysis platform includes a genome database and a machine learning algorithm module for analyzing phenotypic-genotypic associations and optimizing breeding strategies.

[0011] Furthermore, the radiation treatment unit is equipped with a pulsed light control synchronization device, which controls the intermittent irradiation intensity and period of gamma rays or X-rays through a preset program to achieve precise dose adjustment and avoid excessive mutagenesis leading to plant death.

[0012] Furthermore, the data analysis platform integrates an AI-driven phenotypic-genotype association prediction model, uses deep learning algorithms to perform three-dimensional reconstruction of spectral imaging data, automatically identifies mutant phenotypic characteristics and matches corresponding gene loci, and accelerates the efficiency of target trait screening.

[0013] Furthermore, the live transfection module is equipped with a gene delivery system mediated by magnetic nanobeads, which precisely controls the exogenous gene vector to target and enter the plant cell nucleus through a magnetic field, significantly improving gene editing efficiency and accuracy.

[0014] Furthermore, the cell hybridization device incorporates a sterile gas circulation filtration system and a negative pressure isolation chamber design to prevent cross-contamination and support the parallel processing of germplasm resources with different safety levels.

[0015] Furthermore, the high-throughput phenotyping system is equipped with a multi-dimensional mechanical sensor array to monitor plant mechanical traits such as lodging resistance and disease resistance in real time, and to construct a multimodal phenotyping database by combining spectral information.

[0016] Furthermore, the multimodal table database adopts a distributed cloud storage architecture, is implemented based on a NoSQL database, supports standardized storage of JSON-LD data format, and provides a RESTful API interface for the integration, querying, and sharing of multi-source table data. The API interface allows external systems to perform data read and write operations via the HTTP protocol, achieving interoperability with public databases.

[0017] Furthermore, the training data for the AI-driven phenotype-genotype association prediction model comes from publicly available plant genome and phenotype databases, including but not limited to NCBI, Ensembl Plants, or public plant phenomics platforms. These data are updated regularly through standardized data interfaces to ensure the transparency and reproducibility of the model training data.

[0018] This invention provides a multimodal in vivo genetic collaborative transformation system, which has the following beneficial effects: 1. The enclosure structure, through the sealed design of a glass observation door, transparent glass window, and rubber frame, enables visual observation of the experimental process, facilitating real-time monitoring of the plant material's condition. The handle at the front and the display screen on the upper right enhance the intuitiveness of the operation, while the centralized layout of the control buttons further simplifies the parameter setting process. The casters at the bottom give the system flexible mobility to adapt to the needs of different experimental scenarios. The overall structure balances sealing, observation, and mobility, significantly improving the convenience of experimental operation.

[0019] 2. By integrating gene introduction modules, cell hybridization devices, radiation treatment units, high-throughput phenotyping systems, and data analysis platforms, a closed-loop system covering the entire process of "gene manipulation - cell fusion - mutagenesis treatment - phenotyping monitoring - strategy optimization" was constructed.

[0020] 3. The pulsed light-controlled synchronization device optimized for the radiation processing unit controls the irradiation intensity and cycle through a preset program, avoiding the problem of excessive mutagenesis caused by traditional continuous irradiation, and realizing precise adjustment of the mutagenesis dose. It can preserve more usable mutant resources while ensuring mutagenesis efficiency.

[0021] 4. An integrated AI-driven phenotypic-genotype association prediction model uses deep learning algorithms to reconstruct 3D spectral imaging data, which can automatically identify subtle phenotypic features of mutants and quickly match the corresponding gene loci, shortening the traditional manual screening cycle of several weeks or even months to several hours or days, and greatly accelerating the screening efficiency of target traits.

[0022] 5. The nanomagnetic bead-mediated gene delivery system precisely controls the exogenous gene vector to target and enter the plant cell nucleus through magnetic fields, avoiding the randomness of traditional transfection technology and significantly improving the efficiency and accuracy of gene editing, especially suitable for editing complex genomes.

[0023] 6. The cell hybridization device effectively blocks external microbial contamination through a sterile gas circulation filtration system and a negative pressure isolation chamber design. At the same time, it supports the parallel processing of germplasm resources with different safety levels, avoids the risk of cross-contamination, meets biosafety requirements, and expands the application scenarios of the system.

[0024] 7. The newly added multi-dimensional mechanical sensor array can monitor the physical traits of plants in real time, such as lodging resistance and resistance to disease and mechanical damage. It can also be integrated with the physiological trait data from spectral imaging to construct a three-dimensional multimodal phenotypic database. This provides richer data dimensions for a comprehensive analysis of the "gene-environment-phenotype" relationship, further improving the scientific nature and accuracy of breeding strategies. Attached Figure Description

[0025] To more clearly illustrate the embodiments of this utility model or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings in the following description are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the overall structure of this utility model; Figure 2 This is a flowchart illustrating the overall structure of this utility model; Figure 3 This is a schematic diagram of the internal flow of the overall structure of this utility model; Figure 4 This is a schematic diagram of the inter-module signals of the overall structure of this utility model. Detailed Implementation

[0027] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses consistent with some aspects of this disclosure as detailed in the appended claims.

[0028] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present utility model without creative effort are within the scope of protection of the present utility model.

[0029] Example 1: A plant gene editing synergistic transformation system based on nanomagnetic bead delivery and pulsed irradiation

[0030] This embodiment provides a multimodal in vivo genetic co-transformation system, the specific structure of which is as follows: Cabinet structure: The cabinet is made of stainless steel 1, with a double-layer hollow glass observation door 11 installed at the front. The inner side of the glass observation door 11 is embedded with a transparent tempered glass window 12, and the edges are sealed with a food-grade silicone frame 112. The upper right side of the door is integrated with a touch screen display 12, and the lower side is equipped with function control buttons 13. The side of the cabinet is equipped with a sealed cabinet door 14 for equipment maintenance, and the bottom is equipped with four casters 15 with brakes for easy movement and fixation.

[0031] Core functional modules: Gene delivery module: Configured with a gene delivery system mediated by magnetic nanobeads, including a magnetic bead carrier storage chamber, a magnetic field generator, and an electroporation auxiliary device.

[0032] Cell hybridization device: It is a transparent and sealed chamber with a sterile gas circulation system and a negative pressure isolation chamber to support the fusion experiment of rice and wheat protoplasts.

[0033] Radiation treatment unit: Equipped with a gamma ray generator and a new pulsed light control synchronization device to avoid continuous irradiation causing Arabidopsis seedling death.

[0034] High-throughput phenotypic acquisition system: integrates a spectral imager and a multispectral sensor, and adds a multi-dimensional mechanical sensor array to record plant stem bending force and root tension data in real time.

[0035] Data analysis platform: Built-in genome database and machine learning algorithm module, which can automatically match mutant plant height, abnormal leaf color phenotypes and candidate genes.

[0036] Workflow: Rice callus tissue is placed in a cell hybridization device, and drought-resistant genes are introduced through the nanomagnetic bead system of the gene introduction module. At the same time, the radiation treatment unit is activated to induce mutations. The high-throughput phenotyping system records the growth status of the callus tissue in real time. The data analysis platform analyzes the data through an AI model, screens out positive clones with drought-resistant phenotypes, and finally outputs the optimized breeding strategy.

[0037] Example 2: AI-driven multi-species genetic transformation and precise phenotypic analysis system

[0038] This embodiment focuses on the deep integration of AI models and multimodal data, and the system architecture enhances data analysis capabilities based on Embodiment 1: Key improvements: Data analysis platform: Upgraded AI-driven phenotypic-genotype association prediction model, adopting a hybrid architecture combining convolutional neural networks and graph neural networks. CNN is used for local feature extraction of spectral images, and GNN is used for topological analysis of gene regulatory networks. The model training data includes phenotypic-genotype control data of 100,000 Arabidopsis mutants.

[0039] High-throughput phenotypic acquisition system: The spectral imager is equipped with a polarized light module, the mechanical sensing array is extended to three dimensions, and the sampling frequency is increased to 50Hz to capture the rapid growth process.

[0040] Work Scenario: Taking maize lodging resistance breeding as an example, protoplasts of maize inbred line B73 were introduced into lodging resistance-related genes. After fusion using a cell hybridization device, the protoplasts were transferred to a culture medium. The radiation treatment unit induced slight mutations using continuous low-dose irradiation. A high-throughput phenotypic acquisition system simultaneously recorded three-dimensional mechanical data and hyperspectral images of plant stems. The AI ​​model of the data analysis platform automatically identified a strong correlation between "30% increase in stem vascular bundle density" and "2-fold upregulation of ZmPROPEP1 gene expression" by comparing the phenotypic differences between wild-type and transgenic lines. Ultimately, this gene was recommended as the core target, and it was suggested that subsequent site-specific editing using CRISPR-Cas9 technology be performed to enhance lodging resistance.

[0041] Example 3: Collaborative Transformation System for Parallel Processing of Germplasm Resources with Multiple Security Levels

[0042] This embodiment addresses the need for germplasm resource safety by focusing on optimizing the aseptic and isolation design of the cell hybridization device: Key improvements: Cell hybridization device: Built-in three-level biosafety protection, including: ① outer negative pressure chamber, with air discharged through a high-efficiency air filter; ② inner sterile chamber, equipped with a dual-person, dual-lock control system, which can only be opened by authorized personnel through password and fingerprint verification; ③ physical isolation baffle, which supports the simultaneous processing of highly pathogenic host and non-pathogenic germplasm to avoid cross-contamination.

[0043] Gene transfer module: Added temperature gradient control function to adapt to the optimal transfection conditions for cells from different sources.

[0044] Example of work: To conduct a genetic transformation experiment on cassava resistance to Fusarium wilt, it is necessary to treat cassava germplasm infected with Fusarium wilt and healthy banana germplasm at the same time.

[0045] In the experiment: The cell hybridization device activates a physical isolation barrier; area A processes infected cassava, while area B processes healthy bananas. The gene delivery module delivers disease-resistant genes to cassava callus tissue in region A and stress-resistant genes to banana tissue in region B; The radiation treatment unit applies low-dose irradiation to area A to eliminate bacteria, while area B is not irradiated to avoid damaging banana cells. The high-throughput phenotypic acquisition system distinguishes the phenotypic differences between region A and region B through spectral imaging. The data analysis platform, combined with mechanical data, ultimately screens out disease-resistant mutants with "lesion area < 5%" in region A and superior plants with "improved drought resistance" in region B, providing data support for subsequent variety approval.

[0046] Example 4: Safety Protection Measures for Radiation Processing Unit

[0047] The radiation processing unit (e.g., a gamma-ray or X-ray generator) integrates multiple active and passive safety protection systems to ensure the safety of operators and the environment.

[0048] Shielding structure design: Main shielding layer: The core irradiation chamber of the radiation processing unit has a double-layer shielding structure on its inner wall. The inner layer is a high-density lead-tungsten alloy plate (thickness ≥ 50 mm) used to absorb and shield primary radiation; the outer layer is a boron-containing polyethylene composite material (thickness ≥ 30 mm) used to effectively attenuate secondary radiation and absorb neutrons (if generated).

[0049] Door Interlocking Shielding: The sample entrance / exit door connected to the irradiation chamber is hydraulically or electrically driven, and its interior also integrates a lead-tungsten alloy shielding layer with specifications consistent with the inner wall of the chamber. This door is physically interlocked with the high-voltage power supply system of the irradiation device; that is, the high-voltage power supply can only be activated and emit radiation after the shielding door is confirmed to be completely sealed by sensors; conversely, if the door is abnormally opened during irradiation, the high-voltage power supply will be instantly cut off, and irradiation will stop immediately.

[0050] Emergency safety control system: The unit is equipped with an independent emergency stop button, which is located in a prominent position outside the housing (1) and is connected by a hard wire. When triggered, it can bypass the main control system and directly cut off all power supply to the irradiation device.

[0051] The control system has built-in redundant dose monitoring sensors to monitor the radiation dose rate inside and outside the chamber in real time. Once the dose rate is detected to exceed the preset safety threshold (e.g., >2.5 μSv / h on the outer surface of the chamber), the system will automatically trigger an alarm and immediately stop irradiation.

[0052] Warning and Indication System: The irradiation unit is equipped with an external audible and visual alarm device (such as a rotating warning light and a buzzer). Before and during irradiation, the device continuously emits visual and audible warning signals.

[0053] The display screen 12 is equipped with a dedicated safety status indicator interface, which dynamically displays the irradiation status, real-time dose rate, and shielding door status, providing operators with clear safety status information.

[0054] Example: Efficiency comparison experiment between the multimodal in vivo genetic co-transformation system and the traditional Agrobacterium method.

[0055] Experimental Objective

[0056] This embodiment aims to compare the efficiency, accuracy, and operation cycle of the multimodal live genetic co-transformation system (hereinafter referred to as "this system") with the traditional Agrobacterium-mediated transformation method in plant genetic transformation. The performance of the gene delivery module (nanomagnetic bead-mediated system) is evaluated in detail, and quantitative data are provided to demonstrate the superiority of this system, incorporating radiation treatment, phenotypic acquisition, and data analysis functions.

[0057] Materials and methods

[0058] Plant materials: Embryogenic callus from the model plant rice (Oryza sativa L.) variety “Nipponbare” was used as transformation material, and all materials were cultured under the same aseptic conditions.

[0059] Gene construct: A CRISPR-Cas9 editing vector carrying a GFP reporter gene and a hygromycin resistance gene (targeting the rice EPSPS gene) was used to evaluate transformation efficiency and editing precision.

[0060] Experimental Groups: Experimental group: Genetic transformation was performed using this system. The specific procedure is as follows: Gene delivery: The aforementioned magnetic nanoparticle-mediated gene delivery system was used. Gene carriers were combined with magnetic nanoparticles (50 nm in diameter) and precisely controlled by a magnetic field (0.5 T) to be targeted and delivered to the nuclei of callus cells (processing time 10 minutes).

[0061] Radiation treatment: A pulsed optical synchronization device was activated to apply a low dose of gamma rays (total dose 5 Gy, pulse period 1s on / 2s off) to induce controlled mutagenesis.

[0062] Phenotypic acquisition: Through a high-throughput phenotypic acquisition system, GFP fluorescence spectra and mechanical properties (such as cell stiffness) are automatically collected daily, and the data is uploaded to a distributed cloud database in real time.

[0063] Data analysis: AI-driven models (based on deep learning algorithms) are used to automatically identify GFP-positive cells and predict editing sites.

[0064] Control group: The traditional Agrobacterium method (Agrobacterium strain EHA105) was used, and the co-culture protocol was followed (infection time 30 minutes, co-culture for 2 days).

[0065] Evaluation indicators: Transformation efficiency: Calculate the percentage of GFP-positive callus (number of positive callus / total number of treated callus × 100%).

[0066] Editing accuracy: Target editing rate and off-target rate (number of off-target events / total number of editing events × 100%) were analyzed using high-throughput sequencing (Illumina platform).

[0067] Operation cycle: Record the time (in days) required from transformation to obtaining stable transgenic plants.

[0068] Cell survival rate: The proportion of surviving callus tissue was calculated 7 days after transformation.

[0069] Statistical methods: Each experiment was repeated 3 times (n=200 callus tissues each time). Data are expressed as mean ± standard deviation. The t-test was used for significance analysis (p<0.05 was considered significant).

[0070] result

[0071] The table below summarizes the key comparative data:

[0072] Conversion efficiency: This system is significantly higher than the Agrobacterium method (88.5% vs 42.1%), mainly due to the targeted delivery of the nanomagnetic beads, which avoids the host limitation of Agrobacterium.

[0073] Editing accuracy: This system has a higher target editing rate (94.3% vs 65.7%) and a lower off-target rate (1.2% vs 5.8%), which is due to the precise control of the magnetic field and the dose optimization of the radiation unit.

[0074] Operation cycle: This system shortens the time by more than 50% because the phenotypic acquisition system monitors GFP expression in real time, eliminating the need for long-term screening.

[0075] Cell survival rate: This system is higher (92.0% vs 70.5%) because pulsed radiation reduces excessive damage.

[0076] All electrical components mentioned in this article are connected to an external main controller and 220V AC mains power. The main controller can be a conventional known device such as a computer for control. The detailed description of known functions and components is omitted in the specific implementation of this disclosure. To ensure the compatibility of the device, the operating methods used are consistent with the parameters of commercially available instruments.

[0077] Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the present invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multimodal in vivo genetic co-transformation system, comprising a housing (1), characterized in that: The front end of the box (1) is provided with a glass observation door (11), the end face of the glass observation door (11) is provided with a glass window (111), the connection between the glass observation door (11) and the glass window (111) is provided with a plastic frame (112), the end face of the glass observation door (11) is provided with a handle (113), the upper right side of the glass observation door (11) is provided with a display screen (12), the display screen (12) is integrated with the cabinet of the box (1), the lower part of the display screen (12) is provided with a control button (13), the end face of the box (1) is provided with a cabinet door (14), and the lower end of the box (1) is provided with casters (15).

2. The multimodal in vivo genetic co-transformation system according to claim 1, characterized in that: The box (1) integrates a gene introduction module, a cell hybridization device, a radiation treatment unit, a high-throughput phenotypic acquisition system, and a data analysis platform. The gene introduction module is used to introduce exogenous genes into target plant cells through in vivo transfection technology. The cell hybridization device is configured to achieve protoplast fusion of different species or varieties in a sterile environment. The radiation treatment unit uses a γ-ray or X-ray generator to induce mutagenesis in plant materials. The high-throughput phenotypic acquisition system integrates spectral imaging and multispectral sensors to monitor plant physiological traits in real time. The data analysis platform includes a genome database and a machine learning algorithm module, which is used to analyze phenotypic-genotypic associations and optimize breeding strategies.

3. The multimodal in vivo genetic co-transformation system according to claim 2, characterized in that: The radiation treatment unit is equipped with a pulsed light control synchronization device, which controls the intermittent irradiation intensity and period of gamma rays or X-rays through a preset program to achieve precise dose adjustment and avoid excessive mutagenesis that could lead to plant death.

4. The multimodal in vivo genetic co-transformation system according to claim 2, characterized in that: The data analysis platform integrates an AI-driven phenotypic-genotype association prediction model, uses deep learning algorithms to reconstruct spectral imaging data in three dimensions, automatically identifies mutant phenotypic characteristics and matches corresponding gene loci, and accelerates the efficiency of target trait screening.

5. The multimodal in vivo genetic co-transformation system according to claim 2, characterized in that: The live transfection module is equipped with a gene delivery system mediated by magnetic nanobeads, which uses a magnetic field to precisely control the exogenous gene vector to target and enter the plant cell nucleus, significantly improving gene editing efficiency and accuracy.

6. The multimodal in vivo genetic co-transformation system according to claim 2, characterized in that: The cell hybridization device incorporates a sterile gas circulation filtration system and a negative pressure isolation chamber design to prevent cross-contamination and support the parallel processing of germplasm resources with different safety levels.

7. The multimodal in vivo genetic co-transformation system according to claim 2, characterized in that: The high-throughput phenotyping system is equipped with a multi-dimensional mechanical sensor array to monitor plant mechanical traits such as lodging resistance and disease resistance in real time, and to construct a multimodal phenotyping database by combining spectral information.

8. The multimodal in vivo genetic co-transformation system according to claim 7, characterized in that: The multimodal table database adopts a distributed cloud storage architecture, is implemented based on a NoSQL database, supports standardized storage of JSON-LD data format, and provides a RESTful API interface for the integration, querying and sharing of multi-source table data. The API interface allows external systems to perform data read and write operations via the HTTP protocol, realizing interoperability with public databases.

9. The multimodal in vivo genetic co-transformation system according to claim 4, characterized in that: The training data for the AI-driven phenotype-genotype association prediction model comes from publicly available plant genome and phenotype databases, including but not limited to NCBI, Ensembl Plants, or public plant phenomics platforms. This data is updated regularly through standardized data interfaces to ensure the transparency and reproducibility of the model training data.

Images