Large language model construction method and system based on machine learning transcriptome annotation

Abstract

The invention belongs to the technical field of experimental animals, and particularly relates to a machine learning transcriptome annotation-based large language model construction method, which comprises the following steps of: collecting culture data of a transcriptome annotation target disease sample; acquiring orthotopic transplantation animal data of the M transplanted orthotopic transplantation animals in the first transplanted time period; analyzing according to culture data of the transcriptome annotation target disease sample, orthotopic transplantation animal data of the M transplanted orthotopic transplantation animals and first transcriptome annotation target disease differentiation parameters, and respectively selecting complication immunosuppression transcriptome annotation target diseases from the M transplanted orthotopic transplantation animals, obtaining immune deep culture model data; and analyzing the cell activity of the to-be-transplanted animal by utilizing the deep model, and after the cell activity meets a first condition, selecting a transplantation position of the to-be-transplanted animal, so as to realize the construction of the orthotopic transplantation animal model of the transcriptome annotation target disease cell by utilizing a stereotactic technology.

Description

Technical Field

[0001] The present invention belongs to the field of experimental animal technology, relates to the establishment of an in situ transplantation model of a target disease with transcriptome annotation, and specifically relates to a method and system for constructing a large language model based on machine learning transcriptome annotation. Background Art

[0002] The prognosis after surgical resection of the tumor varies greatly. Clinically, it is necessary to accurately evaluate the patient's postoperative prognosis and stratify the risk of recurrence or death to guide the clinic to formulate personalized treatment and follow-up plans and assist clinical decision-making. Accurate preoperative diagnosis, evaluation and prognosis prediction are the basis for obtaining accurate treatment decisions. However, clinical practice has found that after the tumor is completely removed, even at the same stage, the prognosis of patients varies greatly: therefore, accurate prognosis evaluation methods are needed in clinical practice.

[0003] At present, the in vitro co-culture model has achieved excellent invention results in the fields of diagnosis and differential diagnosis, grading and staging, prognosis prediction and efficacy evaluation of transcriptome annotation target diseases, and has formed a relatively mature technical architecture. The integration of imaging genomics and pathology genomics technologies to build a model for predicting the pathological grading and prognosis of early surgically resectable transcriptome annotation target diseases can not only reflect the heterogeneity of tumors in a multi-dimensional, real-time and non-invasive manner, but also provide an important basis for clinicians to formulate precise diagnosis and treatment plans. However, no comprehensive prediction scheme for postoperative risks of transcriptome annotation target diseases has been found in the prior art, which poses a challenge to the recovery of transcriptome annotation target diseases after surgery and the surgical planning of other similar patients.

[0004] Therefore, the present invention intends to construct an orthotopic transplantation animal model based on transcriptome annotation of target disease patients and clinical indicators, explore new biological indicators and markers for accurate prediction of postoperative prognosis of transcriptome annotation target disease, screen the population with advantages in surgical treatment, and assist clinical decision-making. Summary of the invention

[0005] To achieve the above objectives, this application provides the following technical solutions: According to a first aspect of the present invention, the present invention claims a method for constructing a large language model based on machine learning transcriptome annotation, comprising: Collecting culture data of transcriptome-annotated target disease samples; wherein the transcriptome-annotated target disease samples include transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease samples, and the transcriptome-annotated target disease data include attribution data, activity data, and immunosuppression data of the transcriptome-annotated target disease, the activity data includes that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample with complications or that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample without complications, and the immunosuppression data includes immunosuppressants, duration of action of immunosuppressants, and inactivation cycle of the transcriptome-annotated target disease; Collecting in situ transplanted animal data of M transplanted in situ transplanted animals within the first transplanted time period; wherein the in situ transplanted animal data of each transplanted in situ transplanted animal includes the in situ transplanted animal creation time and the in situ transplanted animal inactivation cycle, wherein the in situ transplanted animal creation time is the duration of occurrence of rejection of the immunosuppressive transcriptome annotation target disease of the transplanted in situ transplanted animal, and the in situ transplanted animal inactivation cycle is the duration of occurrence of reduced activity of the immunosuppressive transcriptome annotation target disease after the transplanted in situ transplanted animal returns to the first timestamp; According to the culture data of the transcriptome annotation target disease sample, the orthotopic transplant animal data of the M orthotopic transplant animals, and the first transcriptome annotation target disease differentiation parameter, the M orthotopic transplant animals that have been transplanted are selected to suffer from complications of immunosuppression transcriptome annotation target diseases, and the immune deep culture model data is obtained; Among them, the immune deep culture model data is used to evaluate the transplantation success rate of the transcriptome-annotated target disease within the first transplanted time period; the first transcriptome-annotated target disease differentiation parameters include the shortest operation execution time, the upper limit of immunosuppressants and the upper limit of the duration of action of immunosuppressants.

[0006] Furthermore, the M transplanted orthotopic transplanted animals are M transplanted orthotopic transplanted animals arranged in chronological order according to the creation time of the orthotopic transplanted animals; Among them, according to the culture data of the transcriptome annotation target disease sample, the in situ transplanted animal data of the M transplanted orthotopic transplanted animals, and the first transcriptome annotation target disease differentiation parameter, the M transplanted orthotopic transplanted animals are selected to suffer from complications of immunosuppression transcriptome annotation target diseases, and the immune deep culture model data is obtained, including: Selecting a p-th transplanted orthotopic transplanted animal from the M transplanted orthotopic transplanted animals in the order of creation time of the orthotopic transplanted animals, and collecting orthotopic transplanted animal data of the p-th transplanted orthotopic transplanted animal; Acquiring current data of target disease samples annotated by the transcriptome; Based on the in situ transplanted animal data of the pth transplanted in situ transplanted animal, the current data of the transcriptome annotated target disease sample and the first transcriptome annotated target disease differentiation parameters, the immunosuppressive transcriptome annotated target disease of the pth transplanted in situ transplanted animal is collected.

[0007] Further, the collecting of the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal based on the orthotopic transplanted animal data of the pth transplanted orthotopic transplanted animal, the current data of the transcriptome annotation target disease sample, and the first transcriptome annotation target disease differentiation parameter comprises: When i is not equal to 1, based on the current data of the transcriptome annotated target disease sample, the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample currently suffering from complications are collected; According to the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample currently suffering from complications, the first target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected; wherein the first target transcriptome annotated target disease set is a collection of the first target transcriptome annotated target diseases; When the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is greater than zero, a first target transcriptome annotation target disease in the first target transcriptome annotation target disease set is selected as the immunosuppressive transcriptome annotation target disease of the pth transplanted in situ transplanted animal.

[0008] Further, the first target transcriptome annotation target disease set of the p-th transplanted orthotopic transplanted animal is collected according to the transcriptome annotation target disease data of each transcriptome annotation target disease in the transcriptome annotation target disease sample currently suffering from complications, including: According to the transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease sample with current complications, the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the current inactivation cycle of the transcriptome-annotated target disease are collected for each transcriptome-annotated target disease; wherein the current cumulative immunosuppressant is the cumulative immunosuppressant of the transcriptome-annotated target disease up to the most recent immunosuppression; the current cumulative duration of action of the immunosuppressant is the time interval between the creation time of the in situ transplanted animal of the first immunosuppression of the transcriptome-annotated target disease and the inactivation cycle of the in situ transplanted animal of the most recent immunosuppression; the inactivation cycle of the current transcriptome-annotated target disease is taken from the inactivation cycle of the in situ transplanted animal of the most recent immunosuppression of the transcriptome-annotated target disease; The first target transcriptome annotation target disease set is determined based on the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the inactivation cycle of the current transcriptome annotation target disease for each transcriptome annotation target disease, the in situ transplant animal creation time and the in situ transplant animal inactivation cycle of the pth transplanted in situ transplant animal, and the shortest surgery execution time, the upper limit of the immunosuppressant and the upper limit of the duration of action of the immunosuppressant in the first transcriptome annotation target disease differentiation parameters.

[0009] Further, when the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is greater than zero, selecting a first target transcriptome annotation target disease in the first target transcriptome annotation target disease set as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal comprises: Calculate the matching correlation score of each first target transcriptome annotation target disease according to the current cumulative immunosuppressant action duration, the current cumulative immunosuppressant and the current remaining immunosuppressant for each first target transcriptome annotation target disease in the first target transcriptome annotation target disease set; wherein the current remaining immunosuppressant is equal to the difference between the upper limit of the immunosuppressant and the current cumulative immunosuppressant; The first target transcriptome annotated target disease with the highest matching correlation score in the first target transcriptome annotated target disease set is determined as the immunosuppressive transcriptome annotated target disease of the pth orthotopically transplanted animal.

[0010] Further, the first transcriptome annotation target disease differentiation parameter includes an upper limit of the number of transcriptome annotation target diseases; The step of collecting the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal based on the orthotopic transplanted animal data of the pth transplanted orthotopic transplanted animal, the current data of the transcriptome annotation target disease sample, and the first transcriptome annotation target disease differentiation parameter, further comprises: When the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is not greater than zero, collecting the number of currently suffering complications of the transcriptome annotation target diseases according to the current data of the transcriptome annotation target disease samples, and determining whether the number of currently suffering complications of the transcriptome annotation target diseases is less than the upper limit of the number of transcriptome annotation target diseases; When the number of the currently suffering complications transcriptome annotated target diseases is not less than the upper limit of the transcriptome annotated target diseases, the second target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected according to the transcriptome annotated target disease data of each transcriptome annotated target disease in the currently suffering complications transcriptome annotated target disease sample; wherein the second target transcriptome annotated target disease set is a collection of second target transcriptome annotated target diseases; When the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, one of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is selected as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal.

[0011] Furthermore, when the number of the currently suffering complications transcriptome annotated target diseases is not less than the upper limit of the transcriptome annotated target diseases, the second target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected according to the transcriptome annotated target disease data of each transcriptome annotated target disease in the currently suffering complications transcriptome annotated target disease sample, including: According to the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample with current complications, the current transcriptome annotated target disease inactivation cycle, the current accumulated immunosuppressant and the current accumulated immunosuppressant action duration of each transcriptome annotated target disease are collected; The second target transcriptome annotation target disease set is determined based on the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the inactivation cycle of the current transcriptome annotation target disease for each transcriptome annotation target disease, the in situ transplant animal creation time and the in situ transplant animal inactivation cycle of the pth transplanted in situ transplant animal, and the first transcriptome annotation target disease differentiation parameters.

[0012] Further, when the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, selecting one of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal includes: When the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, based on the current transcriptome annotation target disease inactivation cycle of each second target transcriptome annotation target disease in the second target transcriptome annotation target diseases, the second target transcriptome annotation target disease with the earliest inactivation cycle of the current transcriptome annotation target disease is determined as the immunosuppressive transcriptome annotation target disease of the pth transplanted in situ transplanted animal.

[0013] According to a second aspect of the present invention, the present invention claims a system for constructing a large language model based on machine learning transcriptome annotation, comprising: A first acquisition module is used to acquire culture data of transcriptome-annotated target disease samples; wherein the transcriptome-annotated target disease samples include transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease samples, and the transcriptome-annotated target disease data include attribution data, activity data and immunosuppression data of the transcriptome-annotated target disease, the activity data includes that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample with complications or that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample without complications, and the immunosuppression data includes immunosuppressants, duration of action of immunosuppressants and inactivation cycle of the transcriptome-annotated target disease; The second acquisition module is used to collect the in situ transplanted animal data of M transplanted in situ transplanted animals within the first transplanted time period; wherein the in situ transplanted animal data of each transplanted in situ transplanted animal include the in situ transplanted animal creation time and the in situ transplanted animal inactivation cycle, the in situ transplanted animal creation time is the duration of occurrence of the immunosuppressive transcriptome annotation target disease of the transplanted in situ transplanted animal being rejected, and the in situ transplanted animal inactivation cycle is the duration of occurrence of the immunosuppressive transcriptome annotation target disease completing the transplanted in situ transplanted animal returning to the first timestamp activity reduction; The analysis and result obtaining module is used to analyze the culture data of the transcriptome-annotated target disease samples, the in situ transplanted animal data of the M transplanted in situ transplanted animals, and the first transcriptome-annotated target disease differentiation parameters, and select the M transplanted in situ transplanted animals with complications and immunosuppression transcriptome-annotated target diseases respectively; and obtain immune deep culture model data; wherein, the immune deep culture model data is used to evaluate the transplantation success rate of the transcriptome-annotated target disease within the first transplanted time period; the first transcriptome-annotated target disease differentiation parameters include the shortest operation execution time, the upper limit of immunosuppressants, and the upper limit of the duration of action of immunosuppressants.

[0014] Compared with the prior art, the present invention has the following technical effects: the present invention collects culture data of transcriptome-annotated target disease samples; collects in situ transplanted animal data of M transplanted in situ transplanted animals within a first transplanted time period; analyzes the culture data of the transcriptome-annotated target disease samples, the in situ transplanted animal data of the M transplanted in situ transplanted animals, and the first transcriptome-annotated target disease differentiation parameters, and selects the M transplanted in situ transplanted animals to suffer from complications of immunosuppression transcriptome-annotated target diseases, and obtains immune deep culture model data; after the transcriptome-annotated target disease cells are separated by in vitro culture, their cell activity is analyzed using a deep model, and after the cell activity meets the first condition, the transplantation position of the animal to be transplanted is selected, so as to use stereotactic technology to realize the construction of an in situ transplanted animal model of the transcriptome-annotated target disease cells. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A workflow diagram of a method for constructing a large language model based on machine learning transcriptome annotation as claimed in an embodiment of the present application; Figure 2 This is a structural module diagram of a large language model construction system based on machine learning transcriptome annotation as claimed in an embodiment of the present application. DETAILED DESCRIPTION

[0016] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0017] According to the first embodiment of the present invention, the present invention claims a method for constructing a large language model based on machine learning transcriptome annotation, referring to Figure 1 ,include: Collecting culture data of transcriptome-annotated target disease samples; wherein the transcriptome-annotated target disease samples include transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease samples, and the transcriptome-annotated target disease data include attribution data, activity data, and immunosuppression data of the transcriptome-annotated target disease, the activity data includes that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample with complications or that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample without complications, and the immunosuppression data includes immunosuppressants, duration of action of immunosuppressants, and inactivation cycle of the transcriptome-annotated target disease; Collecting in situ transplanted animal data of M transplanted in situ transplanted animals within the first transplanted time period; wherein the in situ transplanted animal data of each transplanted in situ transplanted animal includes the in situ transplanted animal creation time and the in situ transplanted animal inactivation cycle, wherein the in situ transplanted animal creation time is the duration of occurrence of rejection of the immunosuppressive transcriptome annotation target disease of the transplanted in situ transplanted animal, and the in situ transplanted animal inactivation cycle is the duration of occurrence of reduced activity of the immunosuppressive transcriptome annotation target disease after the transplanted in situ transplanted animal returns to the first timestamp; According to the culture data of the transcriptome annotation target disease sample, the orthotopic transplant animal data of the M orthotopic transplant animals, and the first transcriptome annotation target disease differentiation parameter, the M orthotopic transplant animals that have been transplanted are selected to suffer from complications of immunosuppression transcriptome annotation target diseases, and the immune deep culture model data is obtained; Among them, the immune deep culture model data is used to evaluate the transplantation success rate of the transcriptome-annotated target disease within the first transplanted time period; the first transcriptome-annotated target disease differentiation parameters include the shortest operation execution time, the upper limit of immunosuppressants and the upper limit of the duration of action of immunosuppressants.

[0018] The embodiments of the present invention utilize brain slices that retain intact brain tissue cells, design reasonable culture and inoculation strategies, and construct an in vitro glioma co-culture model. The in vitro glioma co-culture model can effectively simulate the in vivo glioma environment, fill the gap in the transition between glioma cell models and animal models, and verify the pharmacological effects of targeted drugs on gliomas while testing the toxic and side effects of targeted drugs on normal cells around gliomas, providing a powerful tool for the precise screening of targeted glioma drugs.

[0019] Furthermore, the M transplanted orthotopic transplanted animals are M transplanted orthotopic transplanted animals arranged in chronological order according to the creation time of the orthotopic transplanted animals; Among them, according to the culture data of the transcriptome annotation target disease sample, the in situ transplanted animal data of the M transplanted orthotopic transplanted animals, and the first transcriptome annotation target disease differentiation parameter, the M transplanted orthotopic transplanted animals are selected to suffer from complications of immunosuppression transcriptome annotation target diseases, and the immune deep culture model data is obtained, including: Selecting a p-th transplanted orthotopic transplanted animal from the M transplanted orthotopic transplanted animals in the order of creation time of the orthotopic transplanted animals, and collecting orthotopic transplanted animal data of the p-th transplanted orthotopic transplanted animal; Acquiring current data of target disease samples annotated by the transcriptome; Based on the in situ transplanted animal data of the pth transplanted in situ transplanted animal, the current data of the transcriptome annotated target disease sample and the first transcriptome annotated target disease differentiation parameters, the immunosuppressive transcriptome annotated target disease of the pth transplanted in situ transplanted animal is collected.

[0020] Further, the collecting of the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal based on the orthotopic transplanted animal data of the pth transplanted orthotopic transplanted animal, the current data of the transcriptome annotation target disease sample, and the first transcriptome annotation target disease differentiation parameter comprises: When i is not equal to 1, based on the current data of the transcriptome annotated target disease sample, the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample currently suffering from complications are collected; According to the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample currently suffering from complications, the first target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected; wherein the first target transcriptome annotated target disease set is a collection of the first target transcriptome annotated target diseases; When the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is greater than zero, a first target transcriptome annotation target disease in the first target transcriptome annotation target disease set is selected as the immunosuppressive transcriptome annotation target disease of the pth transplanted in situ transplanted animal.

[0021] Further, the first target transcriptome annotation target disease set of the p-th transplanted orthotopic transplanted animal is collected according to the transcriptome annotation target disease data of each transcriptome annotation target disease in the transcriptome annotation target disease sample currently suffering from complications, including: According to the transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease sample with current complications, the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the current inactivation cycle of the transcriptome-annotated target disease are collected for each transcriptome-annotated target disease; wherein the current cumulative immunosuppressant is the cumulative immunosuppressant of the transcriptome-annotated target disease up to the most recent immunosuppression; the current cumulative duration of action of the immunosuppressant is the time interval between the creation time of the in situ transplanted animal of the first immunosuppression of the transcriptome-annotated target disease and the inactivation cycle of the in situ transplanted animal of the most recent immunosuppression; the inactivation cycle of the current transcriptome-annotated target disease is taken from the inactivation cycle of the in situ transplanted animal of the most recent immunosuppression of the transcriptome-annotated target disease; The first target transcriptome annotation target disease set is determined based on the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the inactivation cycle of the current transcriptome annotation target disease for each transcriptome annotation target disease, the in situ transplant animal creation time and the in situ transplant animal inactivation cycle of the pth transplanted in situ transplant animal, and the shortest surgery execution time, the upper limit of the immunosuppressant and the upper limit of the duration of action of the immunosuppressant in the first transcriptome annotation target disease differentiation parameters.

[0022] Among them, in this embodiment, by transplanting the tumor tissue formed in the body of the immunodeficient nude mouse, not only can the structure and microenvironment of the tumor tissue be better maintained, but also it is conducive to the growth of the tumor, and effectively improves the success rate of xenotransplantation. At the same time, directly transplanting dispersed human tumor cells is easily rejected one by one by the remaining immune system in the pig (treated with immunosuppressants); while the transplanted tumor tissue has formed a complete structure and complex cell components, and it is difficult for the immune system in the pig (treated with immunosuppressants) to reject the complete tumor tissue; at the same time, further in vivo screening is used to obtain glioma cells that are adapted to growth in pigs, which is conducive to the survival of tumor cells in pigs and improves the success rate of tumor xenotransplantation.

[0023] Further, when the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is greater than zero, selecting a first target transcriptome annotation target disease in the first target transcriptome annotation target disease set as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal comprises: Calculate the matching correlation score of each first target transcriptome annotation target disease according to the current cumulative immunosuppressant action duration, the current cumulative immunosuppressant and the current remaining immunosuppressant for each first target transcriptome annotation target disease in the first target transcriptome annotation target disease set; wherein the current remaining immunosuppressant is equal to the difference between the upper limit of the immunosuppressant and the current cumulative immunosuppressant; The first target transcriptome annotated target disease with the highest matching correlation score in the first target transcriptome annotated target disease set is determined as the immunosuppressive transcriptome annotated target disease of the pth orthotopically transplanted animal.

[0024] Further, the first transcriptome annotation target disease differentiation parameter includes an upper limit of the number of transcriptome annotation target diseases; The step of collecting the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal based on the orthotopic transplanted animal data of the pth transplanted orthotopic transplanted animal, the current data of the transcriptome annotation target disease sample, and the first transcriptome annotation target disease differentiation parameter, further comprises: When the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is not greater than zero, collecting the number of currently suffering complications of the transcriptome annotation target diseases according to the current data of the transcriptome annotation target disease samples, and determining whether the number of currently suffering complications of the transcriptome annotation target diseases is less than the upper limit of the number of transcriptome annotation target diseases; When the number of the currently suffering complications transcriptome annotated target diseases is not less than the upper limit of the transcriptome annotated target diseases, the second target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected according to the transcriptome annotated target disease data of each transcriptome annotated target disease in the currently suffering complications transcriptome annotated target disease sample; wherein the second target transcriptome annotated target disease set is a collection of second target transcriptome annotated target diseases; When the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, one of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is selected as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal.

[0025] Furthermore, when the number of the currently suffering complications transcriptome annotated target diseases is not less than the upper limit of the transcriptome annotated target diseases, the second target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected according to the transcriptome annotated target disease data of each transcriptome annotated target disease in the currently suffering complications transcriptome annotated target disease sample, including: According to the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample with current complications, the current transcriptome annotated target disease inactivation cycle, the current accumulated immunosuppressant and the current accumulated immunosuppressant action duration of each transcriptome annotated target disease are collected; The second target transcriptome annotation target disease set is determined based on the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the inactivation cycle of the current transcriptome annotation target disease for each transcriptome annotation target disease, the in situ transplant animal creation time and the in situ transplant animal inactivation cycle of the pth transplanted in situ transplant animal, and the first transcriptome annotation target disease differentiation parameters.

[0026] Further, when the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, selecting one of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal includes: When the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, based on the current transcriptome annotation target disease inactivation cycle of each second target transcriptome annotation target disease in the second target transcriptome annotation target diseases, the second target transcriptome annotation target disease with the earliest inactivation cycle of the current transcriptome annotation target disease is determined as the immunosuppressive transcriptome annotation target disease of the pth transplanted in situ transplanted animal.

[0027] According to the second embodiment of the present invention, the present invention claims a large language model construction system based on machine learning transcriptome annotation, referring to Figure 2 ,include: A first acquisition module is used to acquire culture data of transcriptome-annotated target disease samples; wherein the transcriptome-annotated target disease samples include transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease samples, and the transcriptome-annotated target disease data include attribution data, activity data and immunosuppression data of the transcriptome-annotated target disease, the activity data includes that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample with complications or that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample without complications, and the immunosuppression data includes immunosuppressants, duration of action of immunosuppressants and inactivation cycle of the transcriptome-annotated target disease; The second acquisition module is used to collect the in situ transplanted animal data of M transplanted in situ transplanted animals within the first transplanted time period; wherein the in situ transplanted animal data of each transplanted in situ transplanted animal include the in situ transplanted animal creation time and the in situ transplanted animal inactivation cycle, the in situ transplanted animal creation time is the duration of occurrence of the immunosuppressive transcriptome annotation target disease of the transplanted in situ transplanted animal being rejected, and the in situ transplanted animal inactivation cycle is the duration of occurrence of the immunosuppressive transcriptome annotation target disease completing the transplanted in situ transplanted animal returning to the first timestamp activity reduction; The analysis and result obtaining module is used to analyze the culture data of the transcriptome-annotated target disease samples, the in situ transplanted animal data of the M transplanted in situ transplanted animals, and the first transcriptome-annotated target disease differentiation parameters, and select the M transplanted in situ transplanted animals with complications and immunosuppression transcriptome-annotated target diseases respectively; and obtain immune deep culture model data; wherein, the immune deep culture model data is used to evaluate the transplantation success rate of the transcriptome-annotated target disease within the first transplanted time period; the first transcriptome-annotated target disease differentiation parameters include the shortest operation execution time, the upper limit of immunosuppressants, and the upper limit of the duration of action of immunosuppressants.

[0028] The specific implementation methods of the invention are described in detail above, but they are only examples, and the present application is not limited to the specific implementation methods described above. For those skilled in the art, any equivalent modification or substitution of the invention is also within the scope of the present application, and therefore, the equalization, modification, and improvement made without departing from the spirit and principle of the present application should be included in the scope of the present application.

Claims

1. A method for constructing a large language model based on machine learning transcriptome annotation, characterized in that: include: Collecting culture data of transcriptome-annotated target disease samples; wherein the transcriptome-annotated target disease samples include transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease samples, and the transcriptome-annotated target disease data include attribution data, activity data, and immunosuppression data of the transcriptome-annotated target disease, the activity data includes that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample with complications or that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample without complications, and the immunosuppression data includes immunosuppressants, duration of action of immunosuppressants, and inactivation cycle of the transcriptome-annotated target disease; Collecting in situ transplanted animal data of M transplanted in situ transplanted animals within the first transplanted time period; wherein the in situ transplanted animal data of each transplanted in situ transplanted animal includes the in situ transplanted animal creation time and the in situ transplanted animal inactivation cycle, wherein the in situ transplanted animal creation time is the duration of occurrence of rejection of the immunosuppressive transcriptome annotation target disease of the transplanted in situ transplanted animal, and the in situ transplanted animal inactivation cycle is the duration of occurrence of reduced activity of the immunosuppressive transcriptome annotation target disease after the transplanted in situ transplanted animal returns to the first timestamp; According to the culture data of the transcriptome annotation target disease sample, the orthotopic transplant animal data of the M orthotopic transplant animals, and the first transcriptome annotation target disease differentiation parameter, the M orthotopic transplant animals that have been transplanted are selected to suffer from complications of immunosuppression transcriptome annotation target diseases, and the immune deep culture model data is obtained; Among them, the immune deep culture model data is used to evaluate the transplantation success rate of the transcriptome-annotated target disease within the first transplanted time period; the first transcriptome-annotated target disease differentiation parameters include the shortest operation execution time, the upper limit of immunosuppressants and the upper limit of the duration of action of immunosuppressants.

2. A method for constructing a large language model based on machine learning transcriptome annotation according to claim 1, characterized in that: The M transplanted orthotopic transplanted animals are M transplanted orthotopic transplanted animals arranged in chronological order according to the creation time of the orthotopic transplanted animals; Among them, according to the culture data of the transcriptome annotation target disease sample, the in situ transplanted animal data of the M transplanted orthotopic transplanted animals, and the first transcriptome annotation target disease differentiation parameter, the M transplanted orthotopic transplanted animals are selected to suffer from complications of immunosuppression transcriptome annotation target diseases, and the immune deep culture model data is obtained, including: Selecting a p-th transplanted orthotopic transplanted animal from the M transplanted orthotopic transplanted animals in the order of creation time of the orthotopic transplanted animals, and collecting orthotopic transplanted animal data of the p-th transplanted orthotopic transplanted animal; Acquiring current data of the transcriptome annotation target disease samples; Based on the in situ transplanted animal data of the pth transplanted in situ transplanted animal, the current data of the transcriptome annotated target disease sample and the first transcriptome annotated target disease differentiation parameters, the immunosuppressive transcriptome annotated target disease of the pth transplanted in situ transplanted animal is collected.

3. A method for constructing a large language model based on machine learning transcriptome annotation according to claim 2, characterized in that: The step of collecting the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal according to the orthotopic transplanted animal data of the pth transplanted orthotopic transplanted animal, the current data of the transcriptome annotation target disease sample, and the first transcriptome annotation target disease differentiation parameter comprises: When i is not equal to 1, based on the current data of the transcriptome annotated target disease sample, the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample currently suffering from complications are collected; According to the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample of the current complication, the first target transcriptome annotated target disease set of the p-th transplanted in situ transplanted animal is collected; wherein, the first target transcriptome annotated target disease set is a collection of first target transcriptome annotated target diseases; and when the number of the first target transcriptome annotated target diseases in the first target transcriptome annotated target disease set is greater than zero, a first target transcriptome annotated target disease is selected from the first target transcriptome annotated target disease set as the immunosuppressive transcriptome annotated target disease of the p-th transplanted in situ transplanted animal.

4. The method for constructing a large language model based on machine learning transcriptome annotation according to claim 3, characterized in that: The step of collecting the first target transcriptome annotation target disease set of the p-th transplanted orthotopic transplanted animal according to the transcriptome annotation target disease data of each transcriptome annotation target disease in the transcriptome annotation target disease sample currently suffering from complications comprises: According to the transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease sample with current complications, the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the current inactivation cycle of the transcriptome-annotated target disease are collected for each transcriptome-annotated target disease; wherein the current cumulative immunosuppressant is the cumulative immunosuppressant of the transcriptome-annotated target disease up to the most recent immunosuppression; the current cumulative duration of action of the immunosuppressant is the time interval between the creation time of the in situ transplanted animal of the first immunosuppression of the transcriptome-annotated target disease and the inactivation cycle of the in situ transplanted animal of the most recent immunosuppression; the inactivation cycle of the current transcriptome-annotated target disease is taken from the inactivation cycle of the in situ transplanted animal of the most recent immunosuppression of the transcriptome-annotated target disease; The first target transcriptome annotation target disease set is determined based on the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the inactivation cycle of the current transcriptome annotation target disease for each transcriptome annotation target disease, the in situ transplant animal creation time and the in situ transplant animal inactivation cycle of the pth transplanted in situ transplant animal, and the shortest surgery execution time, the upper limit of the immunosuppressant and the upper limit of the duration of action of the immunosuppressant in the first transcriptome annotation target disease differentiation parameters.

5. The method for constructing a large language model based on machine learning transcriptome annotation according to claim 4, characterized in that: When the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is greater than zero, selecting a first target transcriptome annotation target disease in the first target transcriptome annotation target disease set as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal comprises: Calculate the matching correlation score of each first target transcriptome annotation target disease according to the current cumulative immunosuppressant action duration, the current cumulative immunosuppressant and the current remaining immunosuppressant for each first target transcriptome annotation target disease in the first target transcriptome annotation target disease set; wherein the current remaining immunosuppressant is equal to the difference between the upper limit of the immunosuppressant and the current cumulative immunosuppressant; The first target transcriptome annotated target disease with the highest matching correlation score in the first target transcriptome annotated target disease set is determined as the immunosuppressive transcriptome annotated target disease of the pth orthotopically transplanted animal.

6. The method for constructing a large language model based on machine learning transcriptome annotation according to claim 4, characterized in that: The first transcriptome annotation target disease differentiation parameter includes an upper limit of the number of transcriptome annotation target diseases; The step of collecting the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal based on the orthotopic transplanted animal data of the pth transplanted orthotopic transplanted animal, the current data of the transcriptome annotation target disease sample, and the first transcriptome annotation target disease differentiation parameter, further comprises: When the number of the first target transcriptome annotation target diseases in the first target transcriptome annotation target disease set is not greater than zero, collecting the number of currently suffering complications of the transcriptome annotation target diseases according to the current data of the transcriptome annotation target disease samples, and determining whether the number of currently suffering complications of the transcriptome annotation target diseases is less than the upper limit of the number of transcriptome annotation target diseases; When the number of the currently suffering complications transcriptome annotated target diseases is not less than the upper limit of the transcriptome annotated target diseases, the second target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected according to the transcriptome annotated target disease data of each transcriptome annotated target disease in the currently suffering complications transcriptome annotated target disease sample; wherein the second target transcriptome annotated target disease set is a collection of second target transcriptome annotated target diseases; When the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, one of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is selected as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal.

7. The method for constructing a large language model based on machine learning transcriptome annotation according to claim 6, characterized in that: When the number of the currently suffering complications transcriptome annotated target diseases is not less than the upper limit of the transcriptome annotated target diseases, according to the transcriptome annotated target disease data of each transcriptome annotated target disease in the currently suffering complications transcriptome annotated target disease sample, the second target transcriptome annotated target disease set of the p-th transplanted orthotopic transplanted animal is collected, including: According to the transcriptome annotated target disease data of each transcriptome annotated target disease in the transcriptome annotated target disease sample with current complications, the current transcriptome annotated target disease inactivation cycle, the current accumulated immunosuppressant and the current accumulated immunosuppressant action duration of each transcriptome annotated target disease are collected; The second target transcriptome annotation target disease set is determined based on the current cumulative immunosuppressant, the current cumulative duration of action of the immunosuppressant and the inactivation cycle of the current transcriptome annotation target disease for each transcriptome annotation target disease, the in situ transplant animal creation time and the in situ transplant animal inactivation cycle of the pth transplanted in situ transplant animal, and the first transcriptome annotation target disease differentiation parameters.

8. The method for constructing a large language model based on machine learning transcriptome annotation according to claim 7, characterized in that: When the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, selecting one of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set as the immunosuppressive transcriptome annotation target disease of the pth transplanted orthotopic transplanted animal, comprising: When the number of the second target transcriptome annotation target diseases in the second target transcriptome annotation target disease set is greater than zero, based on the current transcriptome annotation target disease inactivation cycle of each second target transcriptome annotation target disease in the second target transcriptome annotation target diseases, the second target transcriptome annotation target disease with the earliest inactivation cycle of the current transcriptome annotation target disease is determined as the immunosuppressive transcriptome annotation target disease of the pth transplanted in situ transplanted animal.

9. A large language model construction system based on machine learning transcriptome annotation, characterized in that: include: A first acquisition module is used to acquire culture data of transcriptome-annotated target disease samples; wherein the transcriptome-annotated target disease samples include transcriptome-annotated target disease data of each transcriptome-annotated target disease in the transcriptome-annotated target disease samples, and the transcriptome-annotated target disease data include attribution data, activity data and immunosuppression data of the transcriptome-annotated target disease, the activity data includes that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample with complications or that the transcriptome-annotated target disease belongs to a transcriptome-annotated target disease sample without complications, and the immunosuppression data includes immunosuppressants, duration of action of immunosuppressants and inactivation cycle of the transcriptome-annotated target disease; The second acquisition module is used to collect the in situ transplanted animal data of M transplanted in situ transplanted animals within the first transplanted time period; wherein the in situ transplanted animal data of each transplanted in situ transplanted animal include the in situ transplanted animal creation time and the in situ transplanted animal inactivation cycle, the in situ transplanted animal creation time is the duration of occurrence of the immunosuppressive transcriptome annotation target disease of the transplanted in situ transplanted animal being rejected, and the in situ transplanted animal inactivation cycle is the duration of occurrence of the immunosuppressive transcriptome annotation target disease completing the transplanted in situ transplanted animal returning to the first timestamp activity reduction; The analysis and result obtaining module is used to analyze the culture data of the transcriptome-annotated target disease samples, the in situ transplanted animal data of the M transplanted in situ transplanted animals, and the first transcriptome-annotated target disease differentiation parameters, and select the M transplanted in situ transplanted animals with complications and immunosuppression transcriptome-annotated target diseases respectively; and obtain immune deep culture model data; wherein, the immune deep culture model data is used to evaluate the transplantation success rate of the transcriptome-annotated target disease within the first transplanted time period; the first transcriptome-annotated target disease differentiation parameters include the shortest operation execution time, the upper limit of immunosuppressants, and the upper limit of the duration of action of immunosuppressants.

Images