WO2020077552A1 - Tumor prognostic prediction method and system - Google Patents

Abstract

A tumor prognostic prediction method and system. The tumor prognostic prediction method comprises: acquiring feature information about a tumor patient, the feature information at least reflecting gene mutation information about the tumor patient (410); and on the basis of the feature information about the tumor patient, and according to a tumor prognostic prediction model, determining a prognostic prediction result of the tumor patient (420). Said method establishes a tumor prognostic prediction model on the basis of tumor patient data, being able to improve the accuracy of tumor prognostic prediction.

Description

Method and system for predicting tumor prognosis Technical field

This application relates to the medical field, in particular to a method and system for predicting tumor prognosis.

Background technique

Tumors (eg, osteosarcoma, etc.) are the second leading cause of death in the world, and the mortality and morbidity rate of tumors are still increasing. Although the diagnosis and treatment of tumors continue to improve, the mortality of patients is still not effectively controlled. Recurrence and metastasis are the main causes of death of tumor patients. For example, osteosarcoma can metastasize to various tissues and organs such as lung and spinal cord Threatening the life of the patient.

At present, the clinical evaluation of tumors mainly through pathological and imaging morphological changes, to clarify the patient's age, tumor pathology type, surgical stage and residual tumor and other indicators. With the development of molecular biology and molecular epidemiology and other technologies, the screening of tumor-related genes and molecular markers at the molecular level is currently a hot spot in cancer research. Such methods can be used for tumors at the molecular level of tumor cells. The patient provides reference indications for surgery, predicts postoperative recurrence or metastasis, objective indications for radical cure of tumors, and provides targets for anti-metastatic treatment.

Therefore, study the differences in gene expression in tumor formation, development, drug resistance, etc., and analyze the activation and inhibition of genes in tumors, so as to more comprehensively and accurately assess the patient's condition and prognosis, so as to implement individualized treatment for tumor patients It has important significance and is also the focus of those skilled in the art.

Summary of the invention

One of the embodiments of the present application provides a tumor prognosis prediction method, including: obtaining characteristic information of a tumor patient, the characteristic information reflecting at least gene mutation information of the tumor patient; based on the characteristic information of the tumor patient, according to the tumor prognosis The prediction model determines the prognosis prediction result of the tumor patient.

In some embodiments, the gene mutation information includes genes and mutation abundances that have been mutated on DNA, and / or genes related to tumor prognosis prediction on DNA and their mutation abundances.

In some embodiments, the obtaining characteristic information of the tumor patient further includes: obtaining a tissue sample of the tumor patient; extracting DNA of the tissue sample; preparing a library of the DNA; performing gene sequencing according to the library to obtain Sequencing results; analyzing the sequencing results to determine gene mutation information of the tumor patient.

In some embodiments, the characteristic information further includes at least one of the following information of the tumor patient: age, gender, smoking history, years of education, working years, treatment plan, and sample storage time.

In some embodiments, the tumor prognosis prediction model is a support vector machine model or a neural network model.

In some embodiments, the tumor prognosis prediction method further includes: training the initial model using the feature information of multiple tumor patients and their prognosis information to obtain the tumor prognosis prediction model.

In some embodiments, the training of the initial model using the feature information and prognostic information of multiple tumor patients to obtain the tumor prognosis prediction model includes: removing the mutation abundance in the gene mutation information of the multiple tumor patients less than Mutated gene information at a certain threshold.

In some embodiments, the training the initial model using the feature information of multiple tumor patients and the prognostic information to obtain the tumor prognosis prediction model includes: removing redundant gene mutation information from the gene mutation information of the multiple tumor patients .

In some embodiments, the tumor prognosis prediction model is a support vector machine model; the training the initial model using the feature information of multiple tumor patients and its prognosis information to obtain the tumor prognosis prediction model includes: The contribution value of each gene mutation information in the feature information to the support vector machine model to determine at least part of the genes as genes related to tumor prognosis prediction; using the gene mutation information of the tumor prognosis prediction related genes of multiple tumor patients and its prognosis information training institute The initial model obtains the tumor prognosis prediction model.

In some embodiments, the tumor prognosis prediction model is a support vector machine model; the training initial model to obtain the tumor prognosis prediction model further includes: optimizing the support vector machine model using particle swarm optimization or meshing parameter.

In some embodiments, the prognosis prediction results include: disease progression, stable disease, partial remission, and complete remission; or, the prognosis prediction results include: good treatment effect and bad treatment effect.

In some embodiments, the tumor is osteosarcoma.

In some embodiments, the characteristic information at least reflects mutation information of at least one of the following genes in osteosarcoma patients: KMT2C, SOX9, LRP1B, NF-1, PRKDC, FAT1, STAG2, SLIT2, NOTCH1, EPHA7, ATRX, KDM6A APC, RANBP2, RARA.AS1, C11orf30, ROS1, ARID2, TAF1, DICER1, MSH2, MSH6, TP53, KDM5A, JAK2, ALK, RB1, NOTCH2 and RICTOR.

In some embodiments, the gene mutation information of the tumor patient is gene mutation information of the osteosarcoma lesion site.

One of the embodiments of the present application provides a tumor prognosis prediction system, including an acquisition module and a prediction module, wherein the acquisition module is used to acquire characteristic information of a tumor patient, and the characteristic information reflects at least gene mutation information of the tumor patient The prediction module is used to determine the prognosis prediction result of the tumor patient based on the tumor patient's characteristic information and according to the tumor prognosis prediction model.

In some embodiments, the gene mutation information includes genes and mutation abundances that have been mutated on DNA, and / or genes related to tumor prognosis prediction on DNA and their mutation abundances.

In some embodiments, the characteristic information further includes at least one of the following information of the tumor patient: age, gender, smoking history, years of education, years of work, treatment plan, and sample storage time.

In some embodiments, the tumor prognosis prediction model is a support vector machine model or a neural network model.

In some embodiments, the tumor prognosis prediction system further includes a training module for training the initial model to obtain the tumor prognosis prediction model by using feature information of multiple tumor patients and their prognosis information.

In some embodiments, the training module is further configured to remove the mutation gene information whose mutation abundance is less than a set threshold in the gene mutation information of the multiple tumor patients.

In some embodiments, the training module is further used to remove redundant gene mutation information from the gene mutation information of the multiple tumor patients.

In some embodiments, the tumor prognosis prediction model is a support vector machine model; the training module is further configured to: according to the contribution value of each gene mutation information in the feature information of multiple tumor patients to the support vector machine model, determine at least Some genes are genes related to tumor prognosis prediction; the gene mutation information of the tumor prognosis prediction related genes of multiple tumor patients and their prognosis information are used to train the initial model to obtain the tumor prognosis prediction model.

In some embodiments, the tumor prognosis prediction model is a support vector machine model; the training module is also used to optimize the parameters of the support vector machine model using particle swarm optimization or meshing.

In some embodiments, the prognosis prediction results include: disease progression, stable disease, partial remission, and complete remission; or, the prognosis prediction results include: good treatment effect and bad treatment effect.

In some embodiments, the tumor is osteosarcoma.

In some embodiments, the characteristic information at least reflects mutation information of at least one of the following genes in osteosarcoma patients: KMT2C, SOX9, LRP1B, NF-1, PRKDC, FAT1, STAG2, SLIT2, NOTCH1, EPHA7, ATRX, KDM6A, APC, RANBP2, RARA.AS1, C11orf30, ROS1, ARID2, TAF1, DICER1, MSH2, MSH6, TP53, KDM5A, JAK2, ALK, RB1, NOTCH2 and RICTOR.

In some embodiments, the gene mutation information of the tumor patient is gene mutation information of the osteosarcoma lesion site.

One of the embodiments of the present application provides a tumor prognosis prediction device. The device includes at least one processor and at least one memory; the at least one memory is used to store computer instructions; and the at least one processor is used to execute the computer instructions At least part of the instructions to implement the tumor prognosis prediction method.

One embodiment of the present application provides a computer-readable storage medium that stores computer instructions, and when the computer instructions are executed by a processor, implements the tumor prognosis prediction method.

One embodiment of the present application provides a tumor prognosis prediction system, including: at least one computer-readable storage medium, including a set of instructions for tumor prognosis prediction; and at least one processor in communication with the at least one storage medium, When executing the set of instructions, the at least one processor is configured to: obtain characteristic information of a tumor patient, the characteristic information reflects at least gene mutation information of the tumor patient; and based on the characteristic information of the tumor patient, according to The tumor prognosis prediction model determines the prognosis prediction result of the tumor patient.

BRIEF DESCRIPTION

The present application will be further described in terms of exemplary embodiments, which will be described in detail through the drawings. These embodiments are not limiting, and in these embodiments, the same numbers indicate the same structure, where:

1 is a schematic diagram of an application scenario of a tumor prognosis prediction system according to some embodiments of the present application;

2 is a schematic structural diagram of a computing device according to some embodiments of the present application;

3 is a block diagram of a tumor prognosis prediction system according to some embodiments of the present application;

4 is an exemplary flowchart of a tumor prognosis prediction method according to some embodiments of the present application;

5 is an exemplary flowchart for determining gene mutation information of a tumor patient according to some embodiments of the present application;

6 is an exemplary flowchart of obtaining a tumor prognosis prediction model according to training shown in some embodiments of the present application;

7 is a heat map of gene mutation in a patient with osteosarcoma according to an exemplary embodiment of the present application;

8 is a heat map of gene mutation in a patient with osteosarcoma with good therapeutic effect according to an exemplary embodiment of the present application;

9 is a heat map of gene mutation in a patient with osteosarcoma with poor treatment effect according to an exemplary embodiment of the present application; and

10 is a schematic diagram of prediction result verification of a tumor prognosis prediction model according to an exemplary embodiment of the present application.

detailed description

In order to explain the technical solutions of the embodiments of the present application more clearly, the drawings required in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some examples or embodiments of the present application. For those of ordinary skill in the art, the application can be applied to these drawings based on these drawings without creative efforts. Other similar scenarios. Unless obvious from the language environment or otherwise stated, the same reference numerals in the figures represent the same structure or operation.

It should be understood that "system", "device", "unit" and / or "module" used herein is a method for distinguishing different components, elements, parts, parts or assemblies at different levels. However, if other words can achieve the same purpose, the words can be replaced by other expressions.

As shown in this application and claims, unless the context clearly indicates an exception, the terms "a", "an", "an", and / or "the" are not specific to the singular but may include the plural. In general, the terms "include" and "include" only suggest that steps and elements that are clearly identified are included, and these steps and elements do not constitute an exclusive list, and the method or device may also contain other steps or elements.

This application uses a flowchart to illustrate the operations performed by the system according to the embodiments of the application. It should be understood that the preceding or following operations are not necessarily performed accurately in order. Instead, the steps can be processed in reverse order or simultaneously. At the same time, you can also add other operations to these processes, or remove a certain step or several steps from these processes.

FIG. 1 is a schematic diagram of an application scenario of a tumor prognosis prediction system 100 according to some embodiments of the present application. As shown in FIG. 1, the tumor prognosis prediction system 100 may include a server 110, a network 120 and a database 130. In some embodiments, the database 130 can store the patient's basic information, disease history, treatment plan data, and can also store the patient's genetic information, such as the genetic mutation information of the tumor patient 140 at the tumor site, the genetic information of the normal tissue of the tumor patient, and Reference gene information, etc. The patient's biological tissue sample or fluid sample, such as the tissue sample 145 of the tumor patient 140, can be stored in a special storage device for further processing, such as gene sequencing processing. Specifically, the tissue sample 145 may include a tumor tissue sample of the patient or tissue samples of other parts of the patient's body. The server 110 may be used to process and analyze relevant information to generate a prognostic prediction result. In some embodiments, the server 110 may obtain relevant information and / or data from the database 130 (for example, gene mutation information of the tumor patient at the tumor site, basic information of the tumor patient, reference gene data, etc.), or may directly obtain work Relevant information and / or data obtained by processing the tissue sample 145 of the tumor patient 140 by personnel or other equipment and instruments.

The server 110 may be a server or a server group. The server group may be centralized, such as a data center. The server group can also be distributed, such as a distributed system. The server 110 may be local or remote. In some embodiments, the server 110 may be implemented on a cloud platform. For example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an intermediate cloud, a multi-cloud, etc., or any combination thereof. In some embodiments, the server 110 may be implemented on the computing device 200 having at least one component shown in FIG. 2.

In some embodiments, the server 110 may include a processing engine 112. The processing engine 112 can be used to execute instructions (program code) of the server 110. For example, the processing engine 112 can execute an instruction to analyze the characteristic information of the tumor patient 140, and then obtain a tumor prognosis prediction result. The instructions for analyzing the characteristic information of the tumor patient 140 may be stored in a computer-readable storage medium (not shown) in the form of computer instructions. In some embodiments, the processing engine 112 may include one or more sub-processing devices (eg, single-core processing devices or multi-core multi-core processing devices). For example only, the processing engine 112 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), an application specific instruction processor (ASIP), a graphics processor (GPU), a physical processor (PPU), and a digital signal processor ( DSP), field programmable gate array (FPGA), editable logic circuit (PLD), controller, microcontroller unit, reduced instruction set computer (RISC), microprocessor, etc. or any combination of the above.

The network 120 may provide a channel for information exchange. In some embodiments, information can be exchanged between the server 110 and the database 130 through the network 120. For example, the server 110 may receive the reference gene data in the database 130 through the network 120. In some embodiments, information about tumor patients 140 and / or tissue samples 145 may be transmitted to the server 110 and / or database 130 via the network 120. For example, the characteristic information of the tumor patient 140 (such as gene mutation information, basic information, etc.) may be transmitted to the server 110 through the network 120. In some embodiments, the network 120 may be any type of wired or wireless network. For example, the network 120 may include a cable network, a wired network, a fiber optic network, a telecommunications network, an internal network, an internet network, a regional network (LAN), a wide area network (WAN), a wireless regional network (WLAN), and a metropolitan area network (MAN ), Public switched telephone network (PSTN), Bluetooth network, ZigBee network, near field communication (NFC) network, etc. or any combination of the above.

The database 130 may be used to store data and / or instruction sets. In some embodiments, the database 130 may store data obtained from the server 110. In some embodiments, the database 130 may store information and / or instructions for execution or use by the server 110 to perform the exemplary methods described in this application. In some embodiments, the reference gene data may be stored in the database 130. Specifically, the database 130 may store genetic data in various types of genomic databases and / or genetic data that has an impact (or significant impact) on tumorigenesis reported in existing literature. The genome database may include, but is not limited to, COSMIC database, ClinVar database, HGMD database, OMIM database, TCGA database, GeneCards database, and so on. In some embodiments, the database 130 may include mass storage, removable storage, volatile read-write memory, read-only memory (ROM), etc., or any combination thereof. In some embodiments, the database 130 may be implemented on a cloud platform. For example only, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an intermediate cloud, a multi-cloud, etc., or any combination thereof. In some embodiments, the database 130 may be part of the server 110.

In some embodiments, the tumor patient 140 may be a patient with one or more tumor diseases. Among them, tumor diseases may include cancer, sarcoma, benign tumor, etc. or any combination thereof. Specifically, the cancer may include squamous cell carcinoma, adenocarcinoma, undifferentiated carcinoma, and the like. For example, squamous cell carcinoma may include cancers that occur in the skin, esophagus, lungs, cervix, vagina, vulva, penis, and the like. Adenocarcinoma can include cancers that occur in the digestive tract, lungs, uterus, breast, ovary, prostate, thyroid, liver, kidney, pancreas, gallbladder, and other parts. Sarcomas can include, but are not limited to: soft tissue sarcoma, osteosarcoma, malignant fibrous histiocytoma, bilateral sarcoma, rhabdomyosarcoma, lymphosarcoma, synovial sarcoma, leiomyoma, and the like. Benign tumors may include, but are not limited to, hamartomas, benign pancreatic tumors, thyroid adenoma, breast fibroids, uterine tumors, gastrointestinal plain osteomyomas, soft tissue fibroids, synovial tumors, ligament fibroma, and the like. In a specific embodiment of the present application, the tumor patient 140 may be an osteosarcoma patient. In some embodiments, the tumor patient 140 may be a patient whose tumor is at various stages (eg, early, middle, late, etc.). The tumor patient 140 may also be a patient at various stages of treatment (eg, before treatment, during treatment, after treatment, etc.).

In some embodiments, the tissue sample 145 may be used to reflect tumor patient 140 tumor related information. Specifically, the tissue sample 145 may be a biological tissue or fluid sample extracted from a tumor site (such as a target lesion) and / or a non-tumor site (such as a site other than the lesion) of the tumor patient 140. For example, tissue samples may include, but are not limited to: sputum, blood samples, fresh tissue (such as surgical tissue, puncture tissue, etc.), paraffin-embedded tissue, urine, serous cavity fluid (such as ascites, pleural effusion, Pericardial effusion, etc.), or tissues, cells, etc. extracted from the tumor site, or any combination of the above. In some embodiments, the tissue sample 145 may include the tissues and cells of the tumor patient 140 at the tumor site and at sites other than the tumor. In some embodiments, the tissue sample 145 may include only the tissues and cells of the tumor patient 140 at the tumor site.

In some embodiments, the relevant information of the tumor patient 140 and / or the tissue sample 145 may be transmitted to one or more components (such as a server) of the tumor prognosis prediction system 100 by humans (such as staff) or machines (such as equipment, etc.) 110, database 130).

FIG. 2 is a schematic diagram of the architecture of a computing device 200 according to some embodiments of the present application. As shown in FIG. 2, the computing device 200 may include a processor 210, a memory 220, an input / output interface 230 and a communication port 240. The server 110 and / or the database 130 may be implemented on the computing device 200. For example, the processing engine 112 may be implemented on the computing device 200 and configured to perform the functions of the processing engine 112 in this application.

The processor 210 may execute calculation instructions (program code) and perform the functions of the server 110 described in this application. Computing instructions may include programs, objects, components, data structures, processes, modules, and functions (functions refer to specific functions described in this application). For example, the processor 210 may process instructions in the tumor prognosis prediction system 100 to predict the effect of tumor prognosis. In some embodiments, the processor 210 may include a microcontroller, a microprocessor, a reduced instruction set computer (RISC), an application specific integrated circuit (ASIC), an application specific instruction set processor (ASIP), and a central processing unit (CPU) , Graphics processing unit (GPU), physical processing unit (PPU), microcontroller unit, digital signal processor (DSP), field programmable gate array (FPGA), advanced RISC machine (ARM), programmable logic device and capable Any circuit, processor, etc. that performs one or more functions, or any combination thereof. For illustration only, only one processor 210 is depicted in FIG. 2, but it should be noted that this application may include multiple processors.

The memory 220 may store data / information obtained from any component in the tumor prognosis prediction system 100. In some embodiments, the memory 220 may include mass storage, removable memory, volatile read and write memory, read-only memory (ROM), etc., or any combination thereof. Exemplary mass storage may include magnetic disks, optical disks, solid-state drives, and the like. Removable memory can include flash drives, floppy disks, optical disks, memory cards, U disks, compact disks, and mobile hard disks. Volatile read and write memory can include random access memory (RAM). RAM may include dynamic RAM (DRAM), double-rate synchronous dynamic RAM (DDRSDRAM), static RAM (SRAM), thyristor RAM (T-RAM), zero capacitance (Z-RAM), and so on. ROM may include mask ROM (MROM), programmable ROM (PROM), erasable programmable ROM (PEROM), electrically erasable programmable ROM (EEPROM), compact disk ROM (CD-ROM) and digital universal disk ROM Wait.

The input / output interface 230 may be used to input or output signals, data, or information. In some embodiments, the input / output interface 230 may be used for user (eg, tumor patient 140, user of the tumor prognosis prediction system 100, etc.) to contact the server 110. In some embodiments, the user can input the characteristic information of the tumor patient through the input / output interface 230. In some embodiments, the input / output interface 230 may include an input device and an output device. Exemplary input devices may include a keyboard, mouse, touch screen, microphone, etc., or any combination thereof. Exemplary output devices may include display devices, speakers, printers, projectors, etc., or any combination thereof. Exemplary display devices may include a liquid crystal display (LCD), a light-emitting diode (LED) based display, a flat panel display, a curved display, a television device, a cathode ray tube (CRT), etc., or any combination thereof.

The communication port 240 may be connected to the network 120 for data communication. The connection may be a wired connection, a wireless connection, or a combination of both. Wired connections may include cables, fiber optic cables, telephone lines, etc., or any combination thereof. The wireless connection may include Bluetooth, WiFi, WiMax, WLAN, ZigBee, mobile network (e.g., 3G, 4G, or 5G, etc.), etc., or any combination thereof. In some embodiments, the communication port 240 may be a standardized port, such as RS232, RS485, and so on. In some embodiments, the communication port 240 may be a specially designed port.

FIG. 3 is a block diagram of a tumor prognosis prediction system according to some embodiments of the present application. As shown in FIG. 3, the tumor prognosis prediction system may include an acquisition module 310, a prediction module 320, and a training module 330.

The obtaining module 310 can be used to obtain the characteristic information of the tumor patient 140. In some embodiments, the feature information may reflect at least the gene mutation information of the tumor patient. In some embodiments, the characteristic information of the tumor patient 140 may include any combination of one or more of gene mutation information of the tumor patient, basic information of the tumor patient, and the like.

The prediction module 320 may be used to predict the prognosis prediction result of the tumor patient. For example, the prediction module 320 may determine the prognosis prediction result of the tumor patient based on the tumor patient's feature information and according to the tumor prognosis prediction model.

The training module 330 may be used for training to obtain a tumor prognosis prediction model. Specifically, the training module 330 may obtain the characteristic information and prognostic information of multiple tumor patients. The training module 330 may use the feature information and prognostic information of multiple tumor patients to train the initial model to obtain a tumor prognosis prediction model. In some embodiments, the training module 330 can remove the mutation gene information whose mutation abundance is less than a set threshold in the gene mutation information. In some embodiments, the training module 330 may remove redundant gene mutation information from the gene mutation information. In some embodiments, the training module 330 may determine that at least part of the genes are related genes for tumor prognosis prediction according to the contribution value of each gene mutation information in the feature information of multiple tumor patients to the support vector machine model. In some embodiments, the training module 330 may use the gene mutation information of the related genes of the tumor prognosis prediction genes of multiple tumor patients and the prognosis information to train the initial model to obtain the tumor prognosis prediction model. In some embodiments, the training module 330 may also use particle swarm optimization or meshing to optimize the parameters of the support vector machine model.

It should be understood that the system and its modules shown in FIG. 3 can be implemented in various ways. For example, in some embodiments, the system and its modules may be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented with dedicated logic; the software part can be stored in the memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art will understand that the above methods and systems can be implemented using computer-executable instructions and / or contained in processor control code, for example, on a carrier medium such as a magnetic disk, CD, or DVD-ROM, such as a read-only memory (firmware Such codes are provided on programmable memories or data carriers such as optical or electronic signal carriers. The system and its modules of the present application can be implemented by not only hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc. It can also be implemented by, for example, software executed by various types of processors, or by a combination of the above hardware circuits and software (for example, firmware).

It should be noted that the above descriptions of the candidate display and determination system and its modules are for convenience of description only, and cannot limit the application to the scope of the illustrated embodiments. It can be understood that, for those skilled in the art, after understanding the principle of the system, it is possible to arbitrarily combine various modules or form a subsystem to connect with other modules without departing from this principle. For example, in some embodiments, the acquisition module 310, the prediction module 320, and the training module 330 may be different modules in a system, or may be a module that implements the functions of the above two or more modules. For example, the acquisition module 310 and the prediction module 320 may also be a module having both acquisition and prediction functions. For another example, each module may share a storage module, or each module may have its own storage module. Such deformations are within the scope of protection of this application.

FIG. 4 is an exemplary flowchart of a tumor prognosis prediction method according to some embodiments of the present application. As shown in FIG. 4, the tumor prognosis prediction method may include:

Step 410: Obtain characteristic information of a tumor patient, and the characteristic information can at least reflect gene mutation information of the tumor patient. Specifically, step 410 may be performed by the obtaining module 310.

In some embodiments, the characteristic information of the tumor patient 140 may include any combination of one or more of gene mutation information of the tumor patient, basic information of the tumor patient, and the like. In some embodiments, the characteristic information of the tumor patient may only include the genetic mutation information of the tumor patient. Specifically, the gene mutation information of the tumor patient may include genes and mutation abundances that have been mutated on DNA, and / or genes related to tumor prognosis prediction on DNA and their mutation abundances. The basic information of the tumor patient can reflect other information related to the tumor patient except the gene mutation information. For example, the basic information of cancer patients can include the age, sex, smoking history, years of education, working years, sample storage time (such as blood storage time, tumor tissue storage time, and other normal tissue storage time) of the cancer patient, treatment plan Etc., or any combination thereof. In some embodiments, the treatment regimen may include the type of treatment regimen (eg, radiation therapy, chemotherapy, immunotherapy, etc.), duration of treatment, dose of radiation used, dose of medication, name or type of medication, and so on. In some specific embodiments, the gene mutation information of the tumor patient may be the gene mutation information of the tumor patient at the tumor site (such as the target lesion). For example, the gene mutation information of the osteosarcoma patient may be the gene mutation information of the osteosarcoma lesion. In some embodiments, the tumor patient 140 may be a patient at various stages of the tumor (eg, early, middle, late, etc.), and / or at various stages of treatment (eg, before treatment, during treatment, after treatment, etc.). For example, characteristic information of patients with osteosarcoma before treatment (such as chemotherapy) can be obtained for predicting the prognostic effect of treatment, and can further provide reference for the formulation and selection of treatment plans.

In some embodiments, acquiring / determining the gene mutation information of the tumor patient 140 may include: obtaining a tissue sample 145 of the tumor patient 140, extracting the DNA of the tissue sample, preparing a library of the DNA, and performing gene sequencing based on the library to obtain sequencing results , Analyze the sequencing results to determine the genetic mutation information of tumor patients and other steps. For more details on determining mutation information of 140 genes in tumor patients, please refer to FIG. 5 and related descriptions.

Step 420, based on the characteristic information of the tumor patient, according to the tumor prognosis prediction model, determine the prognosis prediction result of the tumor patient. Specifically, this step 420 may be performed by the prediction module 320.

In some embodiments, the characteristic information of the tumor patient may be input into the trained tumor prognosis prediction model to obtain the prognosis prediction result of the tumor patient. In some embodiments, the tumor prognosis prediction model may be a supervised learning model. Specifically, the supervised learning model may include one or a combination of one of support vector machine model, decision tree model, neural network model, nearest neighbor classifier, and so on. For the training process of the tumor prognosis prediction model, please refer to FIG. 6 and related descriptions.

In some embodiments, the prognostic prediction result may be the prognosis of a period of time (eg, 5 years) after treatment. For example, the prognostic prediction results can be divided into four categories according to changes in target lesions: PD (progressive disease), stable disease (SD), partial response (PR), partial response (PR), and complete response (CR). . Specifically, PD can mean that the sum of the largest diameters of target lesions increases by 20% or more, or new lesions appear (such as new lesions due to tumor metastasis); SD can mean that the sum of the largest diameters of target lesions does not reach PR, or increases Not reaching PD; PR can refer to the reduction of the sum of the maximum diameter of the target lesions by 30% or more, which should be maintained for at least 4 weeks; CR can refer to the disappearance of all target lesions, no new lesions appear, and the normal tumor markers should be maintained for at least 4 weeks. In some embodiments, the prognostic prediction result may include two types: good treatment effect and bad treatment effect. Specifically, whether the treatment effect is good or not can be determined according to clinical standards. For example, if the tumor patient relapses within 5 years after treatment, it means that the treatment effect is not good, and if the tumor patient does not relapse within 5 years after treatment, it means that the treatment effect is good. For another example, PD and SD can be classified as having a poor therapeutic effect, and PR and CR can be classified as having a good therapeutic effect. For another example, if the survival time of the patient exceeds 5 years after the first treatment, it means that the treatment effect is good; if the survival time of the patient after the first treatment is less than 5 years, it means that the treatment effect is not good.

In some alternative embodiments, the prognostic prediction results may also be classified into other categories, which are not limited in the embodiments of the present application. For example, the prognosis prediction results can be divided into three categories: good treatment effect, general treatment effect and poor treatment effect. In some embodiments, the prognostic prediction result may also be the prediction value of a specific indicator. For example, the prognostic prediction results may include, but are not limited to, disease remission rate, disease recurrence rate, disease recurrence within a few years, disease survival rate, survival time, recent mortality, long-term mortality, hospital mortality, out-of-hospital mortality, surgical death Rate etc.

It should be noted that the above description of the process 400 is only for illustration and explanation, and does not limit the scope of application of the present application. For those skilled in the art, various modifications and changes can be made to the process 400 under the guidance of this application. However, these amendments and changes are still within the scope of this application.

FIG. 5 is an exemplary flowchart for determining gene mutation information of a tumor patient according to some embodiments of the present application. Specifically, the steps shown in FIG. 5 may be performed by staff (such as doctors, laboratory personnel, operators, etc.) and / or instruments (such as detectors, analyzers, etc.). As shown in FIG. 5, the process of determining gene mutation information of a tumor patient may include:

Step 510: Obtain a tissue sample of a tumor patient.

In some embodiments, the tissue sample 145 can be used to reflect tumor-related information. Specifically, the tissue sample 145 may be a biological tissue or fluid sample extracted from a tumor site (such as a target lesion) and / or a non-tumor site (such as a site other than the lesion) of the tumor patient 140. For example, tissue samples may include, but are not limited to: sputum, blood samples, fresh tissue (such as surgical tissue, puncture tissue, etc.), paraffin-embedded tissue, urine, serous cavity fluid (such as ascites, pleural effusion, Pericardial effusion, etc.), or tissues, cells, etc. extracted from the tumor site, or any combination of the above. In some embodiments, the tissue sample 145 may include the tissues and cells of the tumor patient 140 at the tumor site or at a site other than the tumor. In some embodiments, the tissue sample 145 may include only the tissues and cells of the tumor patient 140 at the tumor site. In some embodiments, inclusion criteria may be established for the tissue sample 145. For example, the requirements for collecting tissue samples can be formulated. The requirements are surgical tissue, fresh tissue, puncture tissue, 10% neutral formalin, and paraffin-embedded tissue. For another example, the paraffin white film can be 10 (5 micrometer) or 5 (10 micrometer) white films, and to ensure that the sliced tissue contains a sufficient proportion of tumor cells (such as> 70%), the same HE stain can be added Film (or email to inform the number of tumor cells after HE staining of the specimen sent for examination). For another example, for surgical tissue or puncture tissue, it may be required that the size of the sample collected is> 0.3 cm ³ , and quickly placed in the EP tube. For another example, the sample transportation standard can be formulated: the paraffin white tablets can be sent for inspection at room temperature within 2 weeks after cutting, such as using an EP tube, and the mouth of the tube is sealed with a sealing film to prevent leakage during transportation and needs to be sent The pathology number of the test sample is written on the application form. For another example, criteria for screening tissue samples can be formulated, such as sample rejection criteria: tissues other than 10% neutral formalin fixatives, sample information submitted for inspection does not match the application form, tissue autolysis or degeneration, etc.

In step 520, the DNA of the tissue sample is extracted.

In some embodiments, the method of extracting the DNA of the tissue sample may include cetyltrimethylammonium bromide method (CTAB method), glass bead method, ultrasonic method, grinding method, freeze-thaw method, guanidine isothiocyanate Method, alkaline lysis method, enzymatic method, etc. or any combination of the above. In some embodiments, any known method may also be used to extract the DNA of the tissue sample, which is not limited in the embodiments of the present application.

Step 530: Prepare the DNA library.

In some embodiments, the library preparation process may include some or all of the steps of DNA disruption, end repair, magnetic bead fragment screening, end tailing, adaptor ligation, PCR enrichment, hybrid sequencing library, and the like. In addition, any known method can also be used to prepare a DNA library of tissue samples, which is not limited in the embodiments of the present application.

Step 540: Perform gene sequencing according to the library to obtain sequencing results.

In some embodiments, the prepared library may be subjected to gene sequencing to obtain sequencing data. Among them, the gene sequencing technology may be a high-throughput sequencing technology. High-throughput sequencing technology ("Next-generation" sequencing technology, NGS) may include: single-molecule real-time sequencing (Pacific Bio), ion semiconductor (Ion Torrent sequencing), pyrophosphate sequencing (454), sequencing by synthesis (Illumina) , Sequencing by connection (SOLiD sequencing), chain termination method (Sanger sequencing) and any combination of one or more. In addition, any known method can also be used for gene sequencing, which is not limited in the embodiments of the present application.

Step 550: Analyze the sequencing results to determine the genetic mutation information of the tumor patient.

In some embodiments, the acquired sequencing data can be analyzed to obtain gene mutation information of the tumor patient (including the gene and mutation abundance on the DNA, and / or genes related to tumor prognosis prediction on the DNA, Mutation site mutation abundance, gene mutation abundance, etc.). In some embodiments, the gene mutation abundance may be the cumulative sum of the mutation abundances at the positions where the non-synonymous single nucleotide variation (Single Nucleotide Variation, SNV) in the statistical sequencing result is greater than a certain set value. The set value may be 0.05%, 0.1%, 0.2%, 1%, 2%, 3% or 10%, and so on. Mutation site abundance can refer to the proportion of a base mutation. Specifically, mutation abundance at the mutation site = number of mutant reads / (number of mutant reads + number of wild-type reads), where reads represents a short sequence of sequencing fragments. For example, the mutant gene KMT2C of a patient obtained by sequencing has 5 mutation sites. The mutation abundances of the 5 mutation sites are: 1%, 3%, 4%, 6%, 8%, and the threshold is set to 2% . Then the mutation abundance of the mutant gene KMT2C is the cumulative sum of the mutation abundances of 4 mutation sites greater than 2%. In some embodiments, data analysis may include (1) removing linker sequences in sequencing data; (2) performing quality control and removing low-quality sequencing data (eg, low-quality bases, too short sequencing data, etc.); 3) Compare the processed sequencing data with the reference gene data to identify the mutant gene; (4) remove the normal mutations of the gene (such as polymorphic mutation, synonymous mutation, etc.); (5) obtain the tumor patient ’s Gene mutation information and some or all of the above steps. In some embodiments, the reference gene data may be normal gene data (for example, gene data in normal cells of a non-tumor site of a tumor patient, gene data of a non-tumor patient, etc.), gene data of a corresponding tumor disease (for example, each tumor Prognosis prediction related genes) and so on. 93 patients were sequenced by the sequencing method of the present application, and the statistical coverage of the target area was 98.2% -99.6%, with an average value of 99.41%; the average sequencing depth of the target area was 462.7-1252.89, with an average value of 705.51; the target area was captured The efficiency ranges from 75.6% to 84.6%, with an average value of 80.01%. In some embodiments, the reference gene data may be stored in the database 130, and may be retrieved from the database 130 in use. In some embodiments, any known method can also be used to determine the mutation abundance of a gene. For example, second-generation sequencing, BEAMING, PARE and other technologies.

After sequencing, it was found that different mutant genes were distributed differently in different patient samples. 7-9 are heat maps of gene mutations in osteosarcoma patients according to some embodiments of the present application; wherein, FIG. 7 are heat maps of gene mutations in all osteosarcoma patients according to exemplary embodiments of the application; The heat map of gene mutation in patients with osteosarcoma with good treatment effect according to the exemplary embodiment of the present application; FIG. 9 is the heat map of gene mutation in patients with osteosarcoma with poor treatment effect according to the exemplary embodiment of the present application.

In this embodiment, the corresponding tissues and cells can be extracted from the target lesion (osteosarcoma lesion site) of osteosarcoma patients (93 samples of osteosarcoma patients as shown in FIG. 7), and the genes of osteosarcoma patients can be determined therefrom Mutation information. Specifically, the gene mutation information of the osteosarcoma patient can be determined through the foregoing process steps of determining the gene mutation information of the tumor patient.

In this embodiment, the mutation status (eg, gene mutation abundance) of 315 genes in the sample (according to the genes reported in the literature that have a more significant effect on cancer) is mainly detected. In some alternative embodiments, the number of genes detected may increase or decrease as appropriate. Figure 7-9 lists the gene mutation heat maps of the top 29 mutation ratios in all patients with osteosarcoma, patients with good prognosis and patients with poor prognosis. Among them, the left ordinate of Figure 7-9 represents a certain mutation The ratio of mutations in genes in 93 samples. The right ordinate represents the mutant gene and the abscissa represents the sample. Specifically, in this embodiment, the mutation gene information with a higher proportion of gene mutations in the sample (part of the mutation gene information shown in Figures 7-9) includes: Lysine N-methyltransferase 2C (KMT2C), SRY- box 9 (SOX9), LDL receptor related protein 1B (LRP1B), Neurofibromatosis type I (NF-1), protein Kinase (PRKDC), FAT typical cadherin 1 (FAT1), slit Guidance ligand 2 (SLIT2), Notch1, EPHreceptor A7 (EPHA7), ATRX, Lysine demethylase 6A (KDM6A), APC, RAN binding protein 2 (RANBP2), ROS proto-oncogene 1 (ROS1), EMSY (C11orf30), AT-rich interactive domain-containing protein 2 (ARID2), RARA antisense RNA 1 (RARA.AS1), TATA-box binding protein protein associated 1 factor (TAF1), mutS homolog 2 (MSH2), mutS homolog 6 (MSH6), Tumor protein p53 (TP53), dicer 1 (DICER1), lysine demethylase 5A (KDM5A), Januskinase 2 (JAK2), ALK receptors tyrosinekinase (ALK), RB transcriptional corepressor 1 (RB1), NOTCH2, RPTOR independent dependent companion of MTOR complex 2 ( RICTOR), stromal antibiotic 2 (STAG2), polybromo 1 (PBRM1), melanogenesis associated with transcription factor (MITF), cytochrome P450 family 2 subfamily C member 8 (CYP2C8), phosphatidylinositol-4, 5-bisphosphate 3-kinase catalytic sub3 CA (alpha) , Phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit beta (PIK3CB), B-Raf proto-oncogene (BRAF), MET proto-oncogene, receptor tyrosine Kinase (MET), heat shock protein 90 alpha family family class A member A1 HSP90AA1), membrane associated guanylate skin, WW and PDZ domain containing 2 (MAGI2), mitogen-activated proteinkinasekinasekinase1 (MAP3K1), hepatocytegrowth factor (HGF), E1A bindingprotein protein300 (EP300), AKTineser 3 (AKT3), ASXL transcriptional regulator 1 (ASXL1), ATM serine / threoninekinase (ATM), AXIN1, AXL receptor, tyrosine skinase (AXL), BLM Recq like Helicase (BLM), BRCA2DNA repair associated (BRCA2), cell 73 (CDC73), cy clin dependentkinase12 (CDK12), CREB binding protein (CREBBP), catenin alpha1 (CTNNA1), CYLDlysine 63deubiquitinase (CYLD), EPHreceptorA3 (EPHA3), EPHreceptorB1 (EPHB1), erb-b2recept ERBB3), erb-b2receptor tyrosinekinase 4 (ERBB4), ERBB receptor feedback inhibitor 1 (ERRFI1), FA complementation group A (FANCA), FA complementation group D2 (FANCD2), FAT typical cadherin 1 (FAT1), far1 protein 1 (FUBP1), GATA binding protein 1 (GATA1), GATA binding protein 2 (GATA2), interleukin 7 receptor (IL7R), Januskinase 1 (JAK1), lysine acetyltransferase 6A (KAT6A), LOC101929829, LOC115110like, leucinez regulator 1 (LZTR1), mitogen-activated protein, kinase 2 (MAP2K2), MDM4, mediator complex subunit 12 (MED12), mutL homolog 1 (MLH1), MYC proto-oncogene (MYC), MYCN proto-oncogene (MYCN), NFKBinhibitoralpha (NFKBIA), PARK2, phosphatidylinositol-4,5- bisphosphate 3-kinase catalytic subunit gamma (PIK3CG), phosphoinositide-3-kinase regulatory subunit 2 (PIK3R2), proteinkinase ciota (PRKCI), patched 1 (PTCH1), ret proto-oncogene (RET), SET domain name containing 2 SETD2), SMAD family member 4 (SMAD4), SMARCA4, spence family transcriptional (repressor) (SPEN), spectrin “alpha, erythrocytic 1 (SPTA1), signal“ transducer ”and“ activator ”of“ transcription ”3 (STAT3), transforming“ growth ”factor 2 (GFT2) , TSC complex subunit 1 (TSC1) and other genes.

In addition, after sequencing 315 genes of each patient, it can also be found that the mutation abundances of different mutant genes in the patient samples are different, as shown in Table 1.

Table 1 lists the high-abundance mutant gene information corresponding to each patient (only 10 patients with good prognosis and 10 patients with poor prognosis are shown as examples).

It should be noted that the above description of the process of determining the gene mutation information of a tumor patient is only for illustration and explanation, and does not limit the scope of application of the present application. For those skilled in the art, under the guidance of the present application, any cancer patient gene mutation information obtained by other technical means can be used to achieve the technical purpose of predicting the patient's prognosis.

6 is an exemplary flowchart of obtaining a tumor prognosis prediction model according to training shown in some embodiments of the present application. Specifically, the process shown in FIG. 6 (such as step 610, step 620, etc.) may be performed by the training module 330. As shown in FIG. 6, an exemplary process of training to obtain a tumor prognosis prediction model may include:

Step 610: Acquire feature information and prognostic information of multiple tumor patients.

In some embodiments, the characteristic information of multiple tumor patients may include: any combination of one or more of gene mutation information of tumor patients, basic information of tumor patients, and the like. Specifically, the gene mutation information of multiple tumor patients may include genes and mutation abundances of mutations in the DNA of each tumor patient. In some specific embodiments, the genetic mutation information of the multiple tumor patients may be the genetic mutation information of the tumor patient at the tumor site (such as a target lesion). For a specific method for determining the gene mutation information of the multiple tumor patients, please refer to the flow for determining the gene mutation information of the tumor patient described in FIG. 5. The basic information of the tumor patient can reflect other information related to the tumor patient except the gene mutation information. For example, the basic information of the cancer patient may include the age, gender, smoking history, years of education, working years, treatment plan, sample storage time, medication type, etc. of the cancer patient, or any combination thereof.

In some embodiments, the prognostic information of multiple tumor patients can be divided into disease progression (PD, progressive disease), stable disease (SD, stable disease), partial response (PR, partial response), and complete response according to the changes of target lesions (CR, complete, response) Four categories. For another example, the prognosis may include two types: good treatment effect and bad treatment effect. In some embodiments, the prognosis may also be the value of a specific indicator. For example, the prognosis may include but is not limited to disease remission rate, disease recurrence rate, disease recurrence within a few years, disease survival rate, survival time, recent mortality, long-term mortality, hospital mortality, out-of-hospital mortality, surgical mortality Wait. In some embodiments, the prognosis situation described herein may correspond to the prognosis prediction result determined in step 420.

Step 620: Using the feature information and prognostic information of multiple tumor patients, train an initial model to obtain a tumor prognosis prediction model. In some embodiments, the tumor prognosis prediction model may be a supervised learning model. Specifically, the supervised learning model may include one or a combination of one of support vector machine model, decision tree model, neural network model, nearest neighbor classifier, and so on. In this embodiment, the support vector machine model will be used as an example to illustrate the training process of the tumor prognosis prediction model.

In some embodiments, initial model parameters (eg, parameter c (cost), parameter g (gamma), etc.) may be set to establish an initial support vector machine model. It can also use the meshing method to search for the optimal model parameters (eg, parameter c (cost), parameter g (gamma), etc.) based on the feature information and prognostic information of multiple tumor patients to update and optimize the model. In some embodiments, the kernel function of the support vector machine model (such as linear kernel function, polynomial kernel function, Gaussian (RBF) kernel function, sigmoid kernel function) can be selected, and based on the characteristic information of multiple tumor patients and their prognosis Information training obtains the support vector machine model. In some embodiments, the optimal model parameters can also be found by combining the grid division method and the verification method. For example, the model parameters (eg, parameter c (cost), parameter g (gamma), etc.) are adjusted by the meshing method, the model after the parameter adjustment is verified, and the optimal model parameter is selected according to the verification result.

In still other embodiments, the particle swarm optimization algorithm may be used to optimize the parameters of the support vector machine model. Specifically, you can first initialize the parameters of the particle swarm optimization algorithm, and then use the particle swarm optimization algorithm to find the best parameters to update the model (eg, paired parameters c, g, etc.), and use the best parameters as optimization After the model parameters. Among them, the particle swarm optimization algorithm may include but is not limited to a basic particle swarm optimization algorithm, an adaptive mutation particle swarm optimization algorithm, and the like. The parameters of the particle swarm optimization algorithm can include local search capability parameters, global search capability parameters, elastic coefficients for speed updates, maximum number of evolutions, maximum number of populations, number of cross-validation folds, change range of parameter C, change range of parameter g, etc. , Or any combination thereof. In some embodiments, the parameters of the particle swarm optimization algorithm can be initialized manually or non-manually.

In other embodiments, grid search and particle swarm optimization algorithms can also be used in combination to optimize the parameters of the support vector machine model. For example, you can first use grid search to optimize the parameters of the support vector machine model, and then use particle swarm optimization to optimize it again.

In order to improve the accuracy of the model or improve the training efficiency, the feature information of multiple tumor patients can be further screened, and the filtered feature information can be used for model training.

In some embodiments, the mutation gene information whose mutation abundance is less than a set threshold can be removed from the gene mutation information of the multiple tumor patients. The gene mutation abundance can be the cumulative sum of the mutation abundances of multiple different mutation sites in the gene, and the threshold of the gene mutation abundance at the mutation site can be artificially set (such as 0.05%, 0.1%, 0.2%, 1%, 2%, 3%, etc.), remove the mutation gene information whose mutation abundance is less than the set threshold. For example, some mutation sites with abundances of mutations less than a certain value (such as 0.05%, 0.1%, 0.2%, etc.) may not be counted in their gene mutation abundances.

In some embodiments, redundant gene mutation information in the gene mutation information of the multiple tumor patients may be removed. Specifically, in the gene mutation information, there may be two or more genes, and the correlation between them is relatively high. In some embodiments, when the mutations of the two genes are the same or similar, or the expressions of the mutation abundances of the two genes are similar, the two genes are considered to be highly correlated. For such highly correlated genes, one or more of them may be considered redundant genes. By removing redundant gene mutation information (for example, keeping only one gene in highly correlated genes), the gene dimension can be effectively reduced without affecting the training effect of the model.

In some embodiments, at least part of the genes may be related genes for tumor prognosis prediction according to the contribution value of the mutation information of each gene in the feature information of multiple tumor patients to the support vector machine model.

In some embodiments, the mutation information of each gene in the characteristic information of multiple tumor patients may be further screened. Specifically, the recursive feature elimination method can be used to screen the mutation information of each gene in the feature information of multiple tumor patients. Taking the prediction accuracy of the model as the evaluation standard, the mutation information of each gene in the characteristic information of multiple tumor patients is selectively eliminated to obtain multiple training sets, and a model is trained on each training set. Based on the prediction accuracy The gene mutation information eliminated during the training of each model is sorted by contribution value. It can be understood that the eliminated gene mutation information corresponding to the model with lower prediction accuracy is greater than the eliminated gene mutation information corresponding to the model with higher prediction accuracy. Finally, the mutation information of each gene can be screened according to the contribution value to obtain at least part of genes as genes related to tumor prognosis prediction. In some embodiments, the random forest algorithm can also be used to screen the mutation information of each gene in the characteristic information of multiple tumor patients. Specifically, (1) First build a decision tree: you can define that there are P trees in the forest (such as 20, 40, etc.); you can use the bootstrap sampling method to extract multiple sample sets from 93 samples as each decision tree. Training sample set, repeating P round sampling can get the training sample set of each decision tree. Each round of sampling can sample 93 times from 93 samples with replacement sampling to get the training set of a decision tree; At each node of the tree, assuming a total of 315 feature variables, m feature variables are randomly selected from it, and a feature is selected from the m feature variables for branch growth. The pruning operation is not performed during the growth process, and the best is calculated. (2) Combine the trained P decision trees to get a random forest. According to the P decision trees, multiple tumor patients can be predicted, and the final prediction result is the output of the random forest by weighting or voting. In the process of training each decision tree, it can be calculated how much each tree reduces the impurity of each feature. For a decision tree forest, it is possible to calculate how much impurity each feature has reduced on average, and use the average impurity that it has reduced as the evaluation criterion of how much the contribution value is. For example, the mutation information of the gene with the most reduced impurity can be used as the feature with the largest contribution value, and so on, to determine the contribution value of different mutant genes to the model (as shown in Table 2), so as to screen out at least some genes for tumor prognosis prediction Related genes. For example, n mutant genes with the largest contribution to the tumor prognosis prediction model (such as 20, 29, 40, 100, etc.) can be selected from the mutant genes that have a significant impact on tumorigenesis as tumor prognosis prediction correlation gene.

Table 2 List of contribution values of different mutant genes to the model

In some embodiments, the tumor prognosis prediction model obtained by training can be verified. For example, for the support vector machine model, you can use the cross-validation method to verify the model effect. Specifically, the cross-validation method may include: Hold-Out Method, K-fold Cross-Validation (K-CV) and Leave-One-Out Cross-Validation, LOO-CV). Taking LOO-CV as an example, the training sample can be divided into the total number of samples (for example, 93), one of which is used as a verification sample, and the remaining 92 are used as training samples to input the initial support vector machine model for training, and the cross-validation process is repeated 93 times, 93 verification results were obtained, and the 93 verification results were combined to determine the final verification result of the tumor prognosis prediction model obtained by training. Further, the receiver operating characteristic curve (ROC curve) can be drawn according to the verification result and has been visually represented (as shown in FIG. 10). As shown in FIG. 10, the points on the ROC curve represent the sensitivity and specificity of the osteosarcoma prognosis prediction model under different truncation conditions (such as prognostic effect classification criteria). The upper left corner of the ROC curve is close to the upper left corner, which can reflect the higher prediction accuracy of the osteosarcoma prognosis prediction model obtained in this example; the area under the ROC curve is 0.988, very close to 1, which can reflect The osteosarcoma prognosis prediction model obtained in this example has a good classification effect; in addition, the osteosarcoma prognosis prediction model of the present application has higher sensitivity average value (0.95) and specificity average value (0.97) under different truncation conditions.

In this example, 6 additional patients with osteosarcoma were selected (4 of which are known to have a poor prognostic effect and the other 2 have a good prognostic effect). Obtain the genetic mutation information of the osteosarcoma lesion site, and based on this information, determine the prognostic prediction results of the 6 osteosarcoma patients according to the osteosarcoma prognosis prediction model trained in this embodiment (as shown in Table 3, where the predicted values The threshold is set to 0.5, less than 0.5 is a good prognosis, and greater than 0.5 is a poor prognosis), the obtained prediction results are completely consistent with the known prognostic effect.

Table 3 Comparison of prediction results of osteosarcoma prognosis prediction model and actual prognosis effect

Sample name Predictive value Predicted performance Actual prognosis effect Patient 1 0.335717 Good prognosis Good prognosis Patient 2 0.44896 Good prognosis Good prognosis Patient 3 0.67417 Poor prognosis Poor prognosis Patient 4 0.735268 Poor prognosis Poor prognosis Patient 5 0.756405 Poor prognosis Poor prognosis Patient 6 0.930926 Poor prognosis Poor prognosis

It should be noted that the above description of the process 600 is only for illustration and explanation, and does not limit the scope of application of the present application. For those skilled in the art, various modifications and changes can be made to the process 600 under the guidance of this application. However, these amendments and changes are still within the scope of this application.

The possible benefits brought by the embodiments of the present application include, but are not limited to: (1) the prognostic effect of tumor patients based on gene mutation information can be realized; (2) the accuracy of tumor prognosis prediction is improved; (3) the tumor prognosis prediction process is implemented Convenient; (4) Provide reference for the formulation and selection of treatment plan. It should be noted that different embodiments may have different beneficial effects. In different embodiments, the possible beneficial effects may be any one or a combination of the above, or any other beneficial effects that may be obtained.

The basic concept has been described above. Obviously, for those skilled in the art, the above detailed disclosure is only an example, and does not constitute a limitation on the present application. Although it is not explicitly stated here, those skilled in the art may make various modifications, improvements, and amendments to this application. Such modifications, improvements and amendments are suggested in this application, so such modifications, improvements and amendments still belong to the spirit and scope of the exemplary embodiments of this application.

Meanwhile, this application uses specific words to describe the embodiments of this application. For example, "one embodiment", "one embodiment", and / or "some embodiments" mean a certain feature, structure, or characteristic related to at least one embodiment of the present application. Therefore, it should be emphasized and noted that "one embodiment" or "one embodiment" or "an alternative embodiment" mentioned twice or more at different positions in this specification does not necessarily refer to the same embodiment . In addition, certain features, structures, or characteristics in one or more embodiments of the present application may be combined as appropriate.

In addition, those skilled in the art can understand that various aspects of this application can be illustrated and described through several patentable categories or situations, including any new and useful processes, machines, products, or combinations of materials, or Any new and useful improvements. Correspondingly, various aspects of the present application can be completely executed by hardware, can be completely executed by software (including firmware, resident software, microcode, etc.), or can be executed by a combination of hardware and software. The above hardware or software may be called "data blocks", "modules", "engines", "units", "components" or "systems". In addition, various aspects of this application may appear as a computer product located in one or more computer-readable media, the product including computer-readable program code.

The computer storage medium may contain a propagated data signal containing a computer program code, for example, on baseband or as part of a carrier wave. The propagated signal may have multiple manifestations, including electromagnetic form, optical form, etc., or a suitable combination form. The computer storage medium may be any computer-readable medium except the computer-readable storage medium, and the medium may be connected to an instruction execution system, apparatus, or device to communicate, propagate, or transmit a program for use. Program code located on a computer storage medium may be propagated through any suitable medium, including radio, cable, fiber optic cable, RF, or similar media, or any combination of the foregoing.

The computer program code required for the operation of each part of this application can be written in any one or more programming languages, including object-oriented programming languages such as Java, Scala, Smalltalk, Eiffel, JADE, Emerald, C ++, C #, VB.NET, Python Etc., conventional programming languages such as C, Visual Basic, Fortran 2003, Perl, COBOL 2002, PHP, ABAP, dynamic programming languages such as Python, Ruby and Groovy, or other programming languages. The program code may run entirely on the user's computer, or as an independent software package on the user's computer, or partly on the user's computer, partly on a remote computer, or entirely on the remote computer or server. In the latter case, the remote computer can be connected to the user's computer through any network, such as a local area network (LAN) or a wide area network (WAN), or connected to an external computer (eg, via the Internet), or in a cloud computing environment, or as a service Use as software as a service (SaaS).

In addition, unless explicitly stated in the claims, the order of processing elements and sequences, the use of alphanumeric characters, or the use of other names in the present application are not intended to limit the order of the processes and methods of the present application. Although the above disclosure discusses some currently considered useful embodiments of the invention through various examples, it should be understood that such details are for illustrative purposes only, and the appended claims are not limited to the disclosed embodiments. The requirement is to cover all amendments and equivalent combinations that conform to the essence and scope of the embodiments of the present application. For example, although the system components described above can be implemented by hardware devices, they can also be implemented by software solutions only, such as installing the described system on an existing server or mobile device.

For the same reason, it should be noted that, in order to simplify the expression disclosed in this application and thereby help to understand one or more embodiments of the invention, in the foregoing description of the embodiments of this application, various features are sometimes merged into one embodiment, In the drawings or its description. However, this disclosure method does not mean that the object of this application requires more features than those mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiments disclosed above.

Some embodiments use numbers describing the number of components and attributes. It should be understood that such numbers used in embodiment descriptions use the modifiers "about", "approximately", or "generally" in some examples. Grooming. Unless otherwise stated, "approximately", "approximately" or "substantially" indicates that the figures allow a variation of ± 20%. Correspondingly, in some embodiments, the numerical parameters used in the specification and claims are approximate values, and the approximate value may be changed according to the characteristics required by individual embodiments. In some embodiments, the numerical parameters should consider the specified significant digits and adopt the method of general digit retention. Although the numerical fields and parameters used to confirm the breadth of the ranges in some embodiments of the present application are approximate values, in specific embodiments, the setting of such numerical values is as accurate as possible within the feasible range.

For each patent, patent application, patent application publication, and other materials cited in this application, such as articles, books, specifications, publications, documents, etc., the entire contents are hereby incorporated by reference into this application. Except for application history documents that are inconsistent with or conflict with the content of this application, documents that have the widest scope of claims in this application (current or later appended to this application) are also excluded. It should be noted that if there is any inconsistency or conflict between the descriptions, definitions, and / or terms used in the accompanying materials of this application and the content described in this application, the descriptions, definitions, and / or terms used in this application shall prevail .

Finally, it should be understood that the embodiments described in this application are only used to illustrate the principles of the embodiments of this application. Other variations may also fall within the scope of this application. Therefore, as an example and not a limitation, the alternative configuration of the embodiments of the present application can be regarded as consistent with the teachings of the present application. Accordingly, the embodiments of the present application are not limited to the embodiments explicitly introduced and described in the present application.

Claims (30)

A tumor prognosis prediction method, which is characterized by including: Obtaining characteristic information of a tumor patient, the characteristic information at least reflects gene mutation information of the tumor patient; Based on the characteristic information of the tumor patient, according to the tumor prognosis prediction model, the prognosis prediction result of the tumor patient is determined. The method for predicting tumor prognosis according to claim 1, characterized in that the gene mutation information includes genes and mutation abundances that are mutated in DNA, and / or genes related to tumor prognosis prediction and mutations in DNA . The tumor prognosis prediction method according to claim 1, wherein the acquiring characteristic information of the tumor patient further comprises: Obtaining a tissue sample of the tumor patient; Extract the DNA of the tissue sample; Preparing the DNA library; Perform gene sequencing according to the library to obtain sequencing results; The sequencing result is analyzed to determine gene mutation information of the tumor patient. The method for predicting tumor prognosis according to claim 1, wherein the characteristic information further includes at least one of the following information of the tumor patient: age, gender, smoking history, years of education, years of work, treatment plan And sample retention time. The tumor prognosis prediction method according to claim 1, wherein the tumor prognosis prediction model is a support vector machine model or a neural network model. The method for predicting tumor prognosis according to claim 1, further comprising: The initial model is trained by using the characteristic information of multiple tumor patients and their prognostic information to obtain the tumor prognosis prediction model. The method for predicting tumor prognosis according to claim 6, wherein the training of the initial model using the feature information of multiple tumor patients and the prognostic information thereof to obtain the tumor prognosis prediction model includes: Remove the mutation gene information whose mutation abundance is less than a set threshold in the gene mutation information of the multiple tumor patients. The method for predicting tumor prognosis according to claim 6, wherein the training of the initial model using the feature information of multiple tumor patients and the prognostic information thereof to obtain the tumor prognosis prediction model includes: Remove redundant gene mutation information from the gene mutation information of the multiple tumor patients. The method for predicting tumor prognosis according to claim 6, wherein: The tumor prognosis prediction model is a support vector machine model; The training of the initial model using the characteristic information of multiple tumor patients and the prognostic information to obtain the tumor prognosis prediction model includes: According to the contribution value of each gene mutation information in the feature information of multiple tumor patients to the support vector machine model, determine that at least some genes are related genes for tumor prognosis prediction; The tumor prognosis prediction model is obtained by training the initial model using the gene mutation information of the tumor prognosis prediction related genes of multiple tumor patients and the prognosis information thereof. The method for predicting tumor prognosis according to claim 6, wherein: The tumor prognosis prediction model is a support vector machine model; The training initial model to obtain the tumor prognosis prediction model further includes: optimizing the parameters of the support vector machine model using particle swarm optimization or meshing. The method for predicting tumor prognosis according to claim 1, wherein: The prognostic prediction results include: disease progression, stable disease, partial remission and complete remission; or, The prognosis prediction results include: good treatment effect and bad treatment effect. The method for predicting tumor prognosis according to claims 1-11, wherein the tumor is osteosarcoma. The method for predicting tumor prognosis according to claim 12, wherein the characteristic information at least reflects mutation information of at least one of the following genes in patients with osteosarcoma: KMT2C, SOX9, LRP1B, NF-1, PRKDC, FAT1, STAG2, SLIT2, NOTCH1, EPHA7, ATRX, KDM6A, APC, RANBP2, RARA.AS1, C11orf30, ROS1, ARID2, TAF1, DICER1, MSH2, MSH6, TP53, KDM5A, JAK2, ALK, RB1, NOTCH2 and RICTOR. The method for predicting tumor prognosis according to claim 12, wherein the gene mutation information of the tumor patient is gene mutation information of the osteosarcoma lesion site. A tumor prognosis prediction system, characterized in that it includes an acquisition module and a prediction module, wherein, The obtaining module is used to obtain characteristic information of a tumor patient, the characteristic information at least reflects gene mutation information of the tumor patient; The prediction module is used to determine the prognosis prediction result of the tumor patient according to the tumor prognosis prediction model based on the characteristic information of the tumor patient. The tumor prognosis prediction system according to claim 15, wherein the gene mutation information includes genes and mutation abundances that are mutated in DNA, and / or genes and mutation abundances related to tumor prognosis prediction in DNA . The tumor prognosis prediction system according to claim 15, wherein the characteristic information further includes at least one of the following information of the tumor patient: age, gender, smoking history, years of education, working years, treatment plan And sample retention time. The tumor prognosis prediction system according to claim 15, wherein the tumor prognosis prediction model is a support vector machine model or a neural network model. The tumor prognosis prediction system according to claim 15, further comprising a training module for training the initial model to obtain the tumor prognosis prediction model by using feature information of multiple tumor patients and their prognosis information. The tumor prognosis prediction system according to claim 19, wherein the training module is further used to remove the mutation gene information whose mutation abundance is less than a certain set threshold in the gene mutation information of the multiple tumor patients. The tumor prognosis prediction system according to claim 19, wherein the training module is further configured to remove redundant gene mutation information from the gene mutation information of the multiple tumor patients. The tumor prognosis prediction system according to claim 19, characterized in that The tumor prognosis prediction model is a support vector machine model; The training module is also used to: According to the contribution value of each gene mutation information in the feature information of multiple tumor patients to the support vector machine model, determine that at least some genes are related genes for tumor prognosis prediction; The tumor prognosis prediction model is obtained by training the initial model using the gene mutation information of the tumor prognosis prediction related genes of multiple tumor patients and the prognosis information thereof. The tumor prognosis prediction system according to claim 19, characterized in that The tumor prognosis prediction model is a support vector machine model; The training module is also used to optimize the parameters of the support vector machine model using particle swarm optimization or meshing. The tumor prognosis prediction system according to claim 15, wherein: The prognostic prediction results include: disease progression, stable disease, partial remission and complete remission; or, The prognosis prediction results include: good treatment effect and bad treatment effect. The tumor prognosis prediction system according to claims 15-24, wherein the tumor is osteosarcoma. The tumor prognosis prediction system of claim 25, wherein the characteristic information reflects at least one mutation information of at least one of the following genes in osteosarcoma patients: KMT2C, SOX9, LRP1B, NF-1, PRKDC, FAT1, STAG2, SLIT2, NOTCH1, EPHA7, ATRX, KDM6A, APC, RANBP2, RARA.AS1, C11orf30, ROS1, ARID2, TAF1, DICER1, MSH2, MSH6, TP53, KDM5A, JAK2, ALK, RB1, NOTCH2 and RICTOR. The tumor prognosis prediction system according to claim 26, wherein the tumor patient gene mutation information is gene mutation information of the osteosarcoma lesion site. A tumor prognosis prediction device, characterized in that the device includes at least one processor and at least one memory; The at least one memory is used to store computer instructions; The at least one processor is used to execute at least part of the computer instructions to implement the tumor prognosis prediction method according to any one of claims 1 to 11. A computer-readable storage medium, characterized in that the storage medium stores computer instructions, and when the computer instructions are executed by a processor, the tumor prognosis prediction method according to any one of claims 1 to 11 is implemented. A tumor prognosis prediction system, including: At least one computer-readable storage medium, including a set of instructions for tumor prognosis prediction; and At least one processor in communication with the at least one storage medium, when executing the set of instructions, the at least one processor is configured to: Acquiring characteristic information of a tumor patient, the characteristic information reflecting at least the gene mutation information of the tumor patient; and Based on the characteristic information of the tumor patient, according to the tumor prognosis prediction model, the prognosis prediction result of the tumor patient is determined.

Images