Transplantable organ intelligent computing system

Abstract

The invention relates to a portable organ intelligent computing system. Aiming at an organoid intelligent medical target, a transcription interaction mode jointly regulated and controlled by multiple transcription factors and cis-acting elements is set as input, biological intelligent calculation is performed through the organoid, and output is obtained through a nucleic acid sequencing technology, a pathological detection technology and a space group imaging technology. And analyzing the key gene regulation and control mechanism problem of the organoid intelligent medicine, and assisting the realization of the organoid intelligent medicine.

Description

Technical Field

[0001] The present invention relates to the field of biomedical engineering, and in particular to a transplantable organoid computing system. Background Art

[0002] Thymus transplantation is a common clinical treatment for patients with complex organ failure. However, the organ shortage crisis remains elusive. Transplantable organoids offer a viable alternative. Compared to native organs, thymic organoids exhibit significant differences in multiple dimensions, including tissue structure and immune function. Their genetic regulation mechanisms are complex and unclear, and they do not yet meet the goals of transplantable organoid-based intelligent medicine.

[0003] Aiming for the goal of transplantable organoid-based intelligent medicine, at the genetic level, the same gene sequence can exhibit significant differences in macroscopic and microscopic phenotypes. Internationally cutting-edge research indicates that, compared to the traditional assumption that a single transcription factor independently regulates gene expression, the timing, location, and quantity of interactions between cis-acting elements and multiple transcription factors is the core mechanism of gene expression variation and a key issue in achieving transplantable organoid-based intelligent medicine.

[0004] The present invention targets the open and complex giant system of transplantable organoids, sets the transcriptional interaction pattern jointly regulated by multiple transcription factors and cis-acting elements as input, constructs transplantable organoids, collects spatial omics high-dimensional imaging data, develops intelligent computing methods, completes organoid screening and optimization, realizes the expected goals of transplantable organoid intelligent medicine, and promotes the progress of translational medicine.

[0005] In view of this, the present invention is proposed. Summary of the Invention

[0006] To address the problem that single-gene studies are insufficient to analyze the macro- and micro-phenotypic differences of organoids with the same gene sequence, this paper proposes a transplantable organoid intelligent computing system. Based on 1,649 human transcription factors in a public database, by traversing the transcriptional regulatory patterns of 2-3 transcription factors combined with different gene cis-acting elements, a transcriptional interaction pattern input database is established. This is then input into transplanted organoids for bio-intelligent computing. This system analyzes the association between various transcriptional patterns in organoids and immune cell functions, pathology, and / or therapeutic mechanisms, as well as the interaction network and mechanism between transcription factors and cis-acting elements. This system can guide the optimization of organoid structure and function, and realize intelligent organoid medicine for patients.

[0007] In a first aspect, the present invention provides a transplantable organoid intelligent computing system, the system comprising:

[0008] a. inputting a database of transcriptional interaction patterns, which includes information on transcriptional interaction patterns between cis-acting elements of immune cell genes and transcription factors, and determining the transcriptional interaction information introduced into the organoid based on one or more transcriptional interaction patterns in the input database;

[0009] b. an organoid intelligent computing output database, which includes organoid-related information after organoid transfer to a subject for treatment and / or subject disease information;

[0010] c. An intelligent analysis module, which uses the transcriptional interaction information introduced by the organoids as input data and the association information of the organoids and / or the disease information of the subject as output data, and reversely solves the correspondence between the input data and output data through artificial intelligence.

[0011] In some specific embodiments, the transcriptional interaction pattern input database is established by a method comprising the following steps: traversing and screening the interactions between all cis-acting elements of various immune cells and transcription factors, excluding structurally inaccessible and non-immunomodulatory interaction patterns based on virtual cell technology and protein-nucleic acid interaction prediction for different cell types, performing data dimensionality reduction, and constructing a transcriptional interaction pattern input library.

[0012] In some specific embodiments, the introduction of transcriptional interaction information includes regulation of transcription factor genes or expression, thereby adjusting transcriptional interactions. In some specific embodiments, the regulation of transcription factor genes or expression is the knockout of transcription factors. The regulation is performed by one or more of the CRISPR-Cas system, the APOBEC3-based gene editing system, and the tRNA editing system. Preferably, knockout is performed by the aforementioned three systems to establish a comprehensive, multi-level transcription factor-cis-acting element interaction pattern. In some specific embodiments, the knocked-out transcription factors are selected from STAT1, STAT3, CEBPB, and IKZF3.

[0013] In some specific embodiments, the organoid association information and / or the subject's disease information includes one or more items selected from the following groups: functional information of the organoid, organoid development information, spatial omics information and pathological information of immune organs, and disease progression information. Specifically, the organoid association information and / or the subject's disease information includes immune cell information measured by flow cytometry, urine protein detection information, blood antinuclear antibody detection information, organ (organoid, native organ, spleen, kidney, lung, cardiovascular and other important organ tissues) spatial transcriptomics information, spatial proteomics information, gene sequencing information, transcriptome sequencing information, chromatin immunoprecipitation sequencing information, pathological staining information, organoid occurrence and progression information, and disease progression information.

[0014] In some specific embodiments, the system further includes an analysis module, which is used to: screen therapeutic organoids based on the pathological information output by the organoid intelligent computing, or to analyze the mechanism of action in the transcriptional interaction patterns of different cells in the organoid based on the spatial transcriptomics and / or spatial proteomics information output by the organoid intelligent computing, and to clarify the relationship between cell type, gene regulation, and organoid therapy. The screening and / or analysis includes: for the sequence-image-pathological information multimodal data in the input and output database, respectively using the extraction of uniquely encoded specific transcriptional pattern vector feature information, spatial omics image data features, sequencing sequence and other discrete data features, and learning them separately through multiple feature extraction algorithms; fusing multimodal joint clustering of multimodal data, constructing a classification framework based on a neural network with elastic back propagation, and outputting the spatiotemporal cis-acting element sequence expression changes, transcription factor interactions, and cell type functions and mechanisms during disease progression, organoid development and treatment.

[0015] In some specific embodiments, the organoids are thymic organoids and spleen organoids.

[0016] In a second aspect, the present invention further proposes an organoid intelligent computing method, the method comprising the following steps:

[0017] 1) Construct an input library of organ-transcriptional interaction patterns;

[0018] 2) Preparation of organoid tissue suspension;

[0019] 3) Inputting the transcriptional interaction patterns into the organoid according to the input library described in step 1);

[0020] 4) Use a syringe to transplant the pre-set input organoid suspension into the recipient's quadriceps muscle or lymph node for development and treatment, and perform in vivo intelligent computing of the organoid;

[0021] 5) Collecting organoid intelligent calculation results; Optionally, the method further comprises:

[0022] 6) analyzing the organoid intelligent computational results, screening therapeutic organoids based on the computational results, revealing the spatiotemporal interaction patterns between transcription factors and cis-acting elements, organoid gene regulation patterns, and the association between organoid transplantation therapeutic progress, and / or establishing a transcription factor and cis-acting element interaction network; optionally, the method further comprises:

[0023] 7) Constructing therapeutic transplantable organoids: Based on the screened therapeutic organoids and their cellular composition, gene regulatory mechanisms, and transcription factor interaction networks, construct therapeutic organoids identical to those obtained by screening according to steps 2)-3).

[0024] In some embodiments, the organoids are thymic organoids and spleen organoids.

[0025] In some specific embodiments, step 1) includes: traversing and screening the interactions between cis-acting elements of immune cell genes and transcription factors, performing data dimensionality reduction based on virtual cell technology and protein-nucleic acid interaction prediction, and constructing a transcription interaction pattern input library.

[0026] In some specific embodiments, step 3) includes: regulating the transcription factor gene or its expression selected from the transcription interaction pattern input library of step 1). In some specific embodiments, the regulation of the transcription factor gene or its expression is the knockout of the transcription factor. The regulation includes knocking out through one or more of the CRISPR-Cas system, the APOBEC3-based gene editing system, and the tRNA editing system. Preferably, knocking out through the above three systems can prevent the transcription or action pattern from establishing a comprehensive, multi-level transcription factor-cis-acting element interaction pattern at the gene level and the transcription level. In some specific embodiments, the knocked-out transcription factor is selected from STAT1, STAT3, CEBPB, and IKZF3.

[0027] In some specific embodiments, step 4) includes: using a 23G or 0.34 mm inner diameter needle and syringe to transplant the thymic organoid suspension into the quadriceps muscles of lupus mice on both sides for organoid transplantation therapy and intelligent computing.

[0028] In some specific embodiments, the collection of in vivo intelligent calculation results of organoids in step 5) includes: regular detection of immune cells by flow cytometry, detection of their transcriptome by transcriptome sequencing, and detection of cis-acting elements bound by transcription factors by chromatin immunoprecipitation sequencing; urine protein detection, blood antinuclear antibody detection, and regular detection of disease progression and organoid development progress; using transcriptional interaction information as input data, obtaining organoid functional level, organoid development progress and disease progression as output data.

[0029] In some specific embodiments, collecting in vivo organoid intelligent computation results in step 5) includes obtaining pathological structural staining, spatial transcriptomic imaging, and spatial proteomic imaging of the subject's native organs, organoids, and / or blood vessels. The artificial intelligence data model is trained using transcriptional interaction information as input data and histopathological staining images, spatial transcriptomic images, and spatial proteomic images as output data.

[0030] In some specific embodiments, step 6) includes extracting one-hot encoded specific transcriptional pattern vector feature information, spatial omics image data feature information, and discrete pathology data feature information from the input and output sequence-image-pathology multimodal data, encoding the multimodal data features into a unified representation, and learning them separately using multiple algorithms. Constructing a multimodal organoid computational database: Constructing a multimodal data input module. The input organoid sequence library, discrete sequences in the output organoid function, and pathological data are one-hot encoded to extract sequence attention features; the output spatial transcriptomics images and pathological staining image data are processed to extract image attention features; multimodal data are aligned through symmetry and average product correction, multimodal data fusion is performed, and a multimodal typing framework is constructed; the local dependencies of base sequences are captured using one-hot encoded sequence features; spatial co-localization patterns of chromatin open regions and gene expression are identified through spatial transcriptomics images; the correspondence between several transcription factor binding and spatial organizational structures is constructed in combination with pathological structural information; multimodal joint clustering of multimodal data is integrated, and a typing framework is constructed based on a neural network with elastic backpropagation to output changes in spatiotemporal sequence expression, transcription factor interactions, and cell type functions and mechanisms during disease progression, organoid development, and treatment; the cell composition of therapeutic organoids and the gene regulatory interaction network of various immune cells therein are calculated.

[0031] In some specific embodiments, the intelligent analysis module is as follows: Figure 3 As shown, optionally, it includes: extracting pattern information and pathological information features based on transcriptional interaction patterns, pathological information and organoid developmental progress, extracting image features based on spatial omics images and pathological staining images, and extracting sequence features based on transcription factor sequences and cis-acting element sequences, thereby constructing a multimodal data fusion module; and constructing an organoid classification output network through the multimodal data fusion module, thereby outputting the analysis of the immune function-transcriptional regulation mechanism, the screening of therapeutic organoids, and the composition of therapeutic organoid immune cells.

[0032] Compared with the prior art, the present invention has the following significant advantages:

[0033] 1) The present invention uses organ suspensions to develop in vivo to construct transplantable organoids, which have an immune microenvironment and can be stably constructed and developed into organoids, solving the current organ shortage problem and avoiding the structural and functional differences caused by different in vitro cultivation batch conditions and microenvironments of current organoid technology.

[0034] 2) The present invention constructs a series of organoid libraries targeting specific transcription patterns of specific genes in specific immune cells, conducts high-throughput screening of therapeutic organoids, improves the efficiency of therapeutic organoid construction, reveals the gene regulation mechanism in organoids, and guides patient-specific organoid intelligent medicine.

[0035] 3) The present invention constructs an in situ organoid library, whose structure, function and in vivo conditions are superior to those of current in vitro cultured organoids and can be used for disease process simulation, in vitro diagnosis and drug evaluation and screening. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The drawings are only for purposes of illustrating particular embodiments and are not to be considered limiting of the invention.

[0037] Figure 1 This is the overall flow chart of the transplantable organoid intelligent computing system described in the present invention.

[0038] Figure 2 This is a simplified diagram of an embodiment method of the transplantable organoid intelligent computing system described in the present invention.

[0039] Figure 3 Schematic diagram of the organoid intelligent computing result analysis module described in the present invention. DETAILED DESCRIPTION

[0040] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the case of no conflict, the embodiments of the present invention and the features in the embodiments can be combined with each other. In addition, the present invention can also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited by the specific embodiments disclosed below. The materials, reagents, detection methods, etc. used in the following implementation methods, unless otherwise specified, can all be obtained from commercial sources.

[0041] Example 1

[0042] Constructing a transcriptional interaction pattern input library: Targeting key genes of important immune cells in the thymus and spleen (e.g., B cells, T cells, macrophages), extract their cis-acting element sequence library and randomly pair them with two or three transcription factors in the transcription factor library (e.g., STAT1, STAT3, IKZF3, CEBPB), traversing all cis-acting element-transcription interaction combinations. Targeting transplantable thymus and spleen organoids, using virtual cell technology, traverse various immune cells to screen and exclude combinations with no biological regulatory significance. Using public models, calculate the binding strength between gene cis-acting element nucleic acid sequences and transcription factor protein molecules, as well as the binding strength between transcription factor protein molecules, eliminate combinations without binding interactions, perform data dimensionality reduction, and construct a transcriptional interaction pattern input library.

[0043] Example 2

[0044] According to Example 1, a specific transcriptional interaction pattern is introduced into thymic organoid immune cells. As an example, the following uses the transcriptional pattern regulation of thymic macrophages by knocking out the transcription factors STAT1, CEBPB, and IKZF3 as an example to illustrate:

[0045] S1. Primary thymic organoid extraction. Thymus was extracted from wild-type B6 mice and ground into a tissue suspension using a 70-mesh Tyler sieve in phosphate buffered saline. The suspension was then cultured to establish a thymic organoid suspension.

[0046] S2. In vitro gene editing of thymic organoids. Thymic organoid suspensions were delivered to macrophages via LNPs. The macrophage transcription factors STAT1, CEBPB, and IKZF3 were knocked out using the CRISPR-Cas system, APOBEC3 gene editing, and tRNA gene editing. A mixture of cationic lipid C12-200, dioleoylphosphatidylethanolamine, cholesterol, liposomal glycerol-polyethylene glycol, and dioleoylphosphatidic acid was prepared at a molar ratio of 24.5:11.2:32.6:1.8:30. Ethanol was then added at a volume fraction of 17.5% to prepare the liposomal nanoparticles. The thymic tissue suspension was incubated at 37°C in a 5% CO2 incubator with 10% fetal bovine serum and 1% penicillin-streptomycin for in vitro gene editing.

[0047] Example 3

[0048] Disease model mice were transplanted with the thymic organoids described in Example 2. The thymic organoid development, functional level, disease progression, and pathological conditions were monitored during and at the end of implantation, and intelligent calculations were performed. The details are as follows:

[0049] S3. Thymic organoid transplantation and in vivo intelligent computing. Thymic organoid suspension was transplanted into the quadriceps femoris of lupus mice using a 23G or 0.34mm inner diameter needle and syringe for four weeks. Organoid function was monitored weekly during and after transplantation. Circulating immune cells in the blood were regularly tested by flow cytometry, transcriptome sequencing was used to analyze the transcriptome, and cis-acting elements bound by transcription factors were detected by chromatin immunoprecipitation sequencing. Disease progression and organoid development were regularly monitored by urine protein and blood antinuclear antibody testing. Artificial intelligence data models were trained using transcriptional interaction information as input data and organoid function, organoid development, and disease progression as output data.

[0050] S4. Intelligent Computation of Thymic Organoids. After the transplantation period, the pathological status of the recipient is recorded. After cardiac perfusion, samples are collected from the main affected areas, such as the thymus, thymic organoids, kidneys, and cardiovascular vessels. Histopathological images are obtained by staining with hematoxylin-eosin, Sirius red, and oil red. The collected organs are quickly frozen to preserve the transcriptome. The tissues are embedded and frozen, and adjacent sections are retained for subsequent imaging such as pathological structural staining, spatial transcriptomic imaging, and spatial proteomic imaging. Transcriptional interaction information is used as input data, and histopathological staining images, spatial transcriptomic images, and spatial proteomic images are used as output data.

[0051] Among them, spatial transcriptomic imaging is performed on tissue sections, including probe hybridization, rolling circle amplification, multiple rounds of imaging, image registration, image segmentation, and signal recognition and decoding steps to obtain spatial transcriptome information and match it to the pathological staining structure of adjacent sections, detecting the transcriptional pattern and spatial distribution corresponding to the tissue pathological structure after thymic organoid transplantation. The spatial omics steps are as follows:

[0052] 1) Incubate the sections with 50 μL of a 5 nmol / L mixed coding probe in a humidified chamber at 40°C for 36 hours. 2) After incubation, wash with phosphate buffer, add 50 μL of 5 units / μL T4 ligase, and incubate at room temperature for 2 hours. 3) After incubation, wash with phosphate buffer, add 50 μL of 0.2 units / μL nucleic acid amplification enzyme, and incubate at 30°C for 2 hours. 4) After incubation, wash with phosphate buffer and perform 5 rounds of imaging. For each round, slice the sections with phosphate buffer, add 20 μL of fluorescent probe, incubate at 37°C for 3 hours, and then rinse with phosphate buffer. Acquire images using a confocal microscope to obtain multiple rounds of gene expression information.

[0053] Among them, spatial proteomics imaging of slices is based on expansion microscopy technology, including gelation, protein secondary structure digestion, tissue expansion, multiple rounds of imaging and antibody elution, image registration, signal recognition, and quantitative analysis. High-resolution multiple rounds of imaging are performed for transcription factors STAT1, STAT3, CEBPB, IKZF3, and other important cell molecular markers to detect their spatial interactions, disease progression, organoid development, and treatment effects at the protein level within the tissue. The steps are as follows:

[0054] 1) Tissues were incubated with a swelling monomer solution (comprised of dimethylacrylamide, sodium acrylate, acrylamide, N,N-methylenebisacrylamide, and sodium chloride, dissolved in phosphate buffered saline (PBS). Polymerization was initiated by adding ammonium persulfate, tetramethylethylenediamine, 4-hydroxytamoxifen, and methacrolein prior to use) at 4°C for 30 minutes. The cells, along with the incubation solution, were then transferred to a 0.1 mm thick gel chamber. The incubated gel chamber was sealed and transferred to 37°C for overnight incubation to allow the cells to gel. 2) The gelled sample was removed from the chamber and digested in homogenization buffer. The digested sample was washed three times with phosphate buffered saline (PBS) for 15 minutes each on a shaker at room temperature and then stored in phosphate buffered saline containing 0.02% sodium azide at 4°C. 3) The cells were stained with antibodies to STAT1, STAT3, and IKZF3 overnight in a dark refrigerator at 4°C and washed with immunofluorescence detergent. After staining, the samples were washed 3–5 times in water until they were fully expanded and imaged under a confocal microscope to complete spatial proteomics detection.

[0055] Example 4

[0056] Screen therapeutic thymic organoids based on pathological information output by intelligent computational analysis. Perform reverse decoding based on spatial omics information to analyze the mechanism of action in the transcriptional interaction patterns of different cell types in organoids and elucidate the relationship between cell type, gene regulation, and organoid therapy.

[0057] S5. For the input and output sequence-image-pathology multimodal data, extract the one-hot encoded specific transcriptional pattern vector feature information and spatial omics image data features. Discrete pathology data features are learned separately through multiple feature extraction algorithms. Construct a multimodal thymic organoid computational database: Construct a multimodal data input module. One-hot encode the input organoid sequence library, discrete sequence and pathology data in thymic organoid function, and extract sequence attention features; process the output spatial transcriptomics image and pathology staining image data to extract image attention features; align the multimodal data through symmetry and mean product correction, perform multimodal data fusion, and construct a multimodal typing framework. Use one-hot encoded sequence features to capture local dependencies of base sequences; identify spatial co-localization patterns of chromatin open regions and gene expression through spatial transcriptomics images; and combine pathological structural information to construct the correspondence between several transcription factor binding and spatial organizational structures. Multimodal co-clustering, integrating multimodal data, and constructing a classification framework based on a neural network with elastic backpropagation, outputs spatiotemporal expression changes, transcription factor interactions, and cell type functions and mechanisms during disease progression, organoid development, and treatment. The cellular composition of therapeutic organoids and the gene regulatory interaction networks of various immune cells within them are calculated.

[0058] Example 5

[0059] S6) Gene editing is performed according to the gene regulation pattern calculated in step S5, and corresponding therapeutic thymic organoids are constructed based on steps S1-S3 and transplanted into diseased lupus mice for treatment. The functional level of the organoids is tested weekly during treatment and after the end of treatment. The circulating immune cells in the blood are regularly tested by flow cytometry, and the transcriptome is tested by transcriptome sequencing. The disease process and the progression of the organoids are regularly monitored by urine protein detection and blood antinuclear antibody detection. In vivo fluorescence imaging of small animals is performed to monitor the distribution of macrophages in the body.

[0060] Example 6

[0061] Based on Examples 1-5, a thymic organoid intelligent computing system was established, comprising:

[0062] a. inputting a database of transcriptional interaction patterns, which includes information on transcriptional interaction patterns between cis-acting elements of immune cell genes and transcription factors, and determining the transcriptional interaction information introduced into the organoid based on one or more transcriptional interaction patterns in the input database;

[0063] b. an organoid intelligent computational output database, which includes organoid-related information after organoid transfer to a subject for treatment and / or subject disease information;

[0064] c. An intelligent analysis module, which uses the transcriptional interaction information introduced by the organoids as input data and the association information of the organoids and / or the disease information of the subject as output data, and reversely solves the correspondence between the input data and output data through artificial intelligence.

[0065] In some specific embodiments, the transcriptional interaction pattern input database is established by a method comprising the following steps: traversing and screening the interactions between all cis-acting elements of various immune cells and transcription factors, excluding structurally inaccessible and non-immunomodulatory interaction patterns based on virtual cell technology and protein-nucleic acid interaction prediction for different cell types, performing data dimensionality reduction, and constructing a transcriptional interaction pattern input library.

[0066] In some specific embodiments, the introduction of transcriptional interaction information includes regulation of transcription factor genes or expression, thereby adjusting transcriptional interactions. In some specific embodiments, the regulation of transcription factor genes or expression is the knockout of transcription factors. The regulation is performed by one or more of the CRISPR-Cas system, the APOBEC3-based gene editing system, and the tRNA editing system. Preferably, knockout is performed by the aforementioned three systems to establish a comprehensive, multi-level transcription factor-cis-acting element interaction pattern. In some specific embodiments, the knocked-out transcription factors are selected from STAT1, CEBPB, and IKZF3.

[0067] In some specific embodiments, the organoid association information and / or the subject's disease information includes one or more items selected from the following groups: functional information of the organoid, organoid development information, spatial omics information and pathological information of immune organs, and disease progression information. Specifically, the organoid association information and / or the subject's disease information includes immune cell information measured by flow cytometry, urine protein detection information, blood antinuclear antibody detection information, spatial transcriptomics information of organs (organoids, native organs, spleen, kidneys and other important organ tissues), spatial proteomics information, staining information, organoid occurrence and progression information, and disease progression information.

[0068] In some specific embodiments, the system further includes an analysis module, which is used to: screen therapeutic organoids based on the pathological information output by the organoid intelligent computing, or to analyze the mechanism of action in the transcriptional interaction patterns of different cells in the organoid based on the spatial transcriptomics and / or spatial proteomics information output by the organoid intelligent computing, and to clarify the relationship between cell type, gene regulation, and organoid therapy. The screening and / or analysis includes: for the sequence-image-pathological information multimodal data in the input and output database, respectively using the extraction of uniquely encoded specific transcriptional pattern vector feature information, spatial omics image data features, sequencing sequence and other discrete data features, and learning them separately through multiple feature extraction algorithms; fusing multimodal joint clustering of multimodal data, constructing a classification framework based on a neural network with elastic back propagation, and outputting the spatiotemporal cis-acting element sequence expression changes, transcription factor interactions, and cell type functions and mechanisms during disease progression, organoid development and treatment.

[0069] In some specific embodiments, the organoid is a thymic organoid.

[0070] In some specific embodiments, the intelligent analysis module is as follows: Figure 3As shown, optionally, it includes: extracting pattern information and pathological information features based on transcriptional interaction patterns, pathological information and organoid developmental progress, extracting image features based on spatial omics images and pathological staining images, and extracting sequence features based on transcription factor sequences and cis-acting element sequences, thereby constructing a multimodal data fusion module; and constructing an organoid classification output network through the multimodal data fusion module, thereby outputting the analysis of the immune function-transcriptional regulation mechanism, the screening of therapeutic organoids, and the composition of therapeutic organoid immune cells.

[0071] The above descriptions are only specific embodiments of the present invention. These examples are not intended to limit the present invention to the precise forms disclosed. Various choices and improvements made by those skilled in the art based on the present invention are within the scope of protection of the present invention.

Claims

1. A transplantable organoid intelligent computing system, characterized in that: The system comprises: a. inputting a database of transcriptional interaction patterns, which includes information on transcriptional interaction patterns between cis-acting elements of immune cell genes and transcription factors, and determining the transcriptional interaction information introduced into the organoid based on one or more transcriptional interaction patterns in the input database; b. The organoid input database intelligently calculates the output database, which includes organoid-related information after the organoid is transferred to the subject and / or the subject's disease information; c. An intelligent analysis module, which uses the transcriptional interaction information introduced by the organoids as input data and the association information of the organoids and / or the disease information of the subject as output data, and reversely solves the correspondence between the input data and output data through artificial intelligence.

2. The system according to claim 1, wherein: The transcriptional interaction pattern input database is established by a method comprising the following steps: traversing and screening the interactions between cis-acting elements of immune cell genes and transcription factors, performing data dimensionality reduction based on virtual cell technology and protein-nucleic acid interaction prediction, and constructing a transcriptional interaction pattern input library.

3. The system according to claim 1, wherein: The introduction of transcriptional interaction information into the organoid includes editing through one or more of a CRISPR-Cas system, an APOBEC3-based gene editing system, and a tRNA editing system.

4. The system according to claim 1, wherein: The organoid-associated information and / or the subject's disease information includes one or more items selected from the following groups: functional information of the organoid, organoid development information, spatial omics information and pathological information of immune organs, kidneys, lungs, and cardiovascular organs, and disease progression information.

5. The system according to claim 1, wherein: The system also includes an analysis module, which is used to: screen therapeutic organoids based on pathological information output by organoid intelligent computing, or analyze the mechanism of action in the transcriptional interaction patterns of different cells in organoids based on spatial transcriptomics and / or spatial proteomics information output by organoid intelligent computing, and clarify the relationship between cell type, gene regulation, and organoid therapy.

6. A method for intelligent calculation of transplantable organoids, the method comprising the following steps: 1) Construct an input library of organoid transcriptional interaction patterns; 2) Preparation of organoid tissue suspension; 3) Inputting the transcriptional interaction patterns into the organoid according to the input library described in step 1); 4) Transplanting the pre-set input organoid suspension into the subject and performing in vivo intelligent computing of the organoids; 5) After transplantation, performing intelligent computation of the organoid; optionally, the method further comprises: 6) analyzing the organoid intelligent computational results, screening therapeutic organoids based on the computational results, revealing the spatiotemporal interaction patterns between transcription factors and cis-acting elements, organoid gene regulation patterns, and the association between organoid transplantation therapeutic progress, and / or establishing a transcription factor and cis-acting element interaction network; optionally, the method further comprises: 7) Constructing therapeutic transplantable organoids: Based on the screened therapeutic organoids and their cellular composition, gene regulatory mechanisms, and transcription factor interaction networks, construct therapeutic organoids identical to those obtained by screening according to steps 2)-3).

7. The method according to claim 6, characterized in that The step 1) includes: traversing and screening the interactions between cis-acting elements of immune cell genes and transcription factors, performing data dimensionality reduction based on virtual cell technology and protein-nucleic acid interaction prediction, and constructing a transcription interaction pattern input library.

8. The method according to claim 6, characterized in that The step 3) comprises: regulating the transcription factor gene or its expression selected from the transcription interaction pattern input library in step 1).

9. The method according to claim 6, characterized in that The in vivo intelligent calculation results of the organoids in step 4) include: regular detection of circulating immune cells in the blood by flow cytometry, and detection of the transcriptome by transcriptome sequencing; detection of the interaction between transcription factors and cis-acting elements by chromatin immunoprecipitation sequencing; regular detection of disease progression and organoid development progress by pathological testing; using transcriptional interaction information as input data, and obtaining organoid functional level, organoid development progress and disease progression as output data through in vivo organoid calculation.

10. The method according to claim 6, characterized in that The organoid intelligent calculation in step 5) includes: obtaining pathological structural staining, spatial transcriptomic imaging and spatial proteomic imaging of the subject's organs, organoids, affected organs and / or blood vessels, using transcriptional interaction information as input data, and tissue pathology staining images, spatial transcriptomic images and spatial proteomic images as output data, and reversely analyzing the mechanism of action in the transcriptional interaction patterns of different cells in the organoids through an artificial intelligence model to clarify the relationship between cell types, gene regulation and organoid treatment.

Images