CN114373511A - Intestinal cancer model based on 5hmC molecular marker detection and intestinal cancer model construction method - Google Patents
Intestinal cancer model based on 5hmC molecular marker detection and intestinal cancer model construction method Download PDFInfo
- Publication number
- CN114373511A CN114373511A CN202210250395.5A CN202210250395A CN114373511A CN 114373511 A CN114373511 A CN 114373511A CN 202210250395 A CN202210250395 A CN 202210250395A CN 114373511 A CN114373511 A CN 114373511A
- Authority
- CN
- China
- Prior art keywords
- parameters
- unit
- data corresponding
- monitored
- core model
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B25/00—ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
- G16B25/10—Gene or protein expression profiling; Expression-ratio estimation or normalisation
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
- G16B50/00—ICT programming tools or database systems specially adapted for bioinformatics
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
Landscapes
- Health & Medical Sciences (AREA)
- Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Biotechnology (AREA)
- Evolutionary Biology (AREA)
- Databases & Information Systems (AREA)
- Biophysics (AREA)
- Public Health (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Bioethics (AREA)
- Genetics & Genomics (AREA)
- Epidemiology (AREA)
- Pathology (AREA)
- Primary Health Care (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)
Abstract
The application discloses an intestinal cancer model based on 5hmC molecular marker detection, which comprises a core model with parameters, the core model with parameters is an intestinal cancer model based on molecular markers as parameters, and further comprises a case collection unit, a case inspection unit, an error collection unit, a group-based analysis unit and a cyclic modification unit, the output end of the case collection unit is connected with the case inspection unit in an interconnecting way, the output end of the case inspection unit is connected with the error collection unit in an interconnecting way, the output end of the error collection unit is interactively connected with the group-based analysis unit, the group-based analysis unit is interactively connected with the cycle modification unit, the circulation modification unit is in bidirectional interactive connection with a core model with parameters, the core model with parameters is also in bidirectional interactive connection with a case inspection unit, and the application discloses an intestinal cancer model construction method based on 5hmC molecular marker detection.
Description
Technical Field
The application relates to an intestinal cancer model based on 5hmC molecular marker detection and a construction method thereof.
Background
In recent years, research on intestinal cancer models based on molecular marker detection has become a hotspot mainly because molecular marker-based intestinal cancer prediction has many advantages, mainly because the molecular marker-based intestinal cancer prediction has the characteristics of being minimally invasive, sensitive, efficient and good in compliance of monitors, and related prior arts also have many, for example, chinese patent document 201710662851.6 discloses a gene marker, a kit and a pancreatic cancer detection method for detecting pancreatic cancer. Chinese patent document 201710662852.0 discloses a gene marker, a kit and a detection method for detecting benign and malignant liver tumors. Also, chinese patent document 201910026672.2 discloses an apparent modified 5hmC multi-molecular marker and an early colorectal cancer diagnosis model, the core of the technology is to select a marker-based gene marker, which includes: GBX2 FAM8A FAM25B LCE1F FBXL7 DBX1 KRTAP27-1 AL353791.1 CEBPD LTB4R2 RP4-583P15.1 OR5B2 RPRM RNASE INSL5 AURKC IL36A AC017081.1 SPA17 NBPF12 FABP1CST 8. Then standardizing the result by the expression level of the selected gene marker 5-hmC of the monitored person; and inputting the standardized result into a regression model established in advance, and then judging the score of the regression model to be used as a reference for judging whether the monitored person has colorectal cancer. Although a great number of markers participate in the judgment in the technology, the normalization result of the corresponding marker and the parameters participating in the judgment of the regression model cannot be determined most accurately in the establishment of the regression model, and in fact, the normalization result and the parameters of the judgment of the regression model are made based on past experience which cannot completely meet the actual situation, so that an error case may occur in the judgment performed by using the previous regression model, and thus the existing intestinal cancer model based on the detection of the molecular marker cannot give reference to the colorectal cancer most accurately.
Disclosure of Invention
The application provides an intestinal cancer model based on 5hmC molecular marker detection, which comprises a core model with parameters, wherein the core model with parameters is specifically an intestinal cancer model based on molecular markers as parameters, and is characterized by further comprising a case collection unit, a case inspection unit, an error collection unit, a group-based analysis unit and a cyclic modification unit, wherein the output end of the case collection unit is in interactive connection with the case inspection unit, the output end of the case inspection unit is in interactive connection with the error collection unit, the output end of the error collection unit is in interactive connection with the group-based analysis unit, the group-based analysis unit is in interactive connection with the cyclic modification unit, the cyclic modification unit is in bidirectional interactive connection with the core model with parameters, and the core model with parameters is also in bidirectional interactive connection with the case inspection unit;
the case collection unit is used for collecting data corresponding to the gene marker of the monitored person, and the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to all the gene markers of the monitored person;
the case inspection unit is used for inspecting the core model with the parameters based on data corresponding to the gene markers of the monitored persons in a plurality of groups each time and outputting data corresponding to the gene markers of the monitored persons with errors in inspection;
the error collecting unit is used for collecting data corresponding to gene markers of a plurality of groups of monitored persons for detecting errors and transmitting the data to the group-by-group analyzing unit;
the group-based analysis unit is used for grouping parameters in the 'core model with parameters', and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises the step of correcting the parameters in the 'core model with parameters' by training and learning data corresponding to gene markers of a plurality of groups of monitored people with errors; and is also used for transmitting the modified parameters to the cyclic modification unit;
the loop modification unit is used for modifying the parameters of the 'core model with parameters' after acquiring the 'parameters after modification' each time, and is also used for circularly executing the modification.
Preferably, the system further comprises a re-inspection unit and an inspection denoising unit, wherein the re-inspection unit is interactively connected with the output end of the error collection unit, the re-inspection unit is also bidirectionally interactively connected with the case collection unit, the output end of the re-inspection unit is also interactively connected with the inspection denoising unit, the inspection denoising unit is interactively connected with the output end of the error collection unit, the output end of the inspection denoising unit is interactively connected with the group-based analysis unit, the re-inspection unit is used for obtaining data corresponding to gene markers of a plurality of groups of monitored persons with inspection errors from the error collection unit and interacting with the case collection unit to determine data corresponding to gene markers of a plurality of groups of monitored persons with inspection errors, and the inspection denoising unit is used for interacting with the error collection unit and the group-based analysis unit and obtaining data corresponding to the gene markers of the group of monitored persons with inspection errors, from the re-inspection unit, wherein the data corresponding to the group of monitored persons with inspection errors should be eliminated Data corresponding to gene markers of a plurality of groups of monitored persons "" and transmitting an error collection unit to data corresponding to gene markers of a plurality of groups of monitored persons who are removed from the data of the analysis unit and checked for errors "" are detected.
Preferably, the genetic marker of the monitored subject comprises a 5hmC multi-molecular marker based on an apparent modification.
Preferably, the parameters in the core model with parameters include parameters of the weighted values corresponding to the gene markers of the monitored person and parameters of the overall judgment threshold.
Preferably, the core model with parameters is tested based on the data corresponding to the gene markers of several groups of monitored persons, and the data corresponding to the gene markers of several groups of monitored persons with test errors is output, specifically: firstly, using 'parameters of weighted value corresponding to each gene marker of monitored person' and whole judgment threshold value parameter 'corresponding to original core model with parameters as reference to input' numerical value corresponding to each gene marker of monitored person 'in data corresponding to gene markers of several groups of monitored person' into original core model with parameters, and if the judgment result is inconsistent with the actual 'data item of confirmed diagnosis result of monitored person', then the inconsistent 'data corresponding to gene markers of several groups of monitored person' is the 'data corresponding to gene markers of several groups of monitored person who are wrong to be tested'.
Preferably, the training and learning of the data corresponding to the gene markers of the groups of the monitored people who are checked for errors further modifies the parameters in the core model with parameters, specifically calibrates the parameters of the weighting values corresponding to the gene markers of the monitored people and the overall judgment threshold parameters corresponding to the original core model with parameters to be quantitative, the other parameters are variables, then the variables are subjected to a plurality of functional transformations, the transformed variables are used as new parameters, then inputting the numerical values corresponding to all the genetic markers of the monitored person in the data corresponding to the genetic markers of the monitored persons for detecting errors into the core model with parameters after the parameters are changed, if the judgment result is not consistent with the actual record in the 'confirmed diagnosis result data item of the monitored person'; continuously calibrating a plurality of parameters in the parameter of the weighted value corresponding to each gene marker of the monitored person and the integral judgment threshold parameter corresponding to the original core model with the parameters as the fixed quantity and the other parameters as the variables, then carrying out a plurality of function transformations on the variables, and taking the transformed variables as new parameters; until the transformed variables are used as new parameters, inputting the numerical values corresponding to each gene marker of the monitored person in the data corresponding to the gene markers of the monitored persons with the wrong detection into a core model with the parameters after the parameters are changed, and judging that the results are consistent with the records in the actual data items of the confirmed diagnosis results of the monitored persons; the latest parameters are used as parameters needed for modifying the 'core model with parameters'.
Preferably, the plurality of kinds of functional transformations include a forward functional transformation that is a functional transformation that increases the weighted values of the variable corresponding parameters and a reverse functional transformation that decreases the weighted values of the variable corresponding parameters.
The intestinal cancer model construction method based on 5hmC molecular marker detection comprises the following steps: collecting data corresponding to the gene marker of the monitored person, wherein the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to all the gene markers of the monitored person;
checking the core model with the parameters based on data corresponding to the gene markers of the monitored persons in a plurality of groups, and outputting data corresponding to the gene markers of the monitored persons with wrong checking;
collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors, and transmitting the data to a group-based analysis unit;
grouping parameters in the core model with the parameters according to a group analysis unit, and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises the step of modifying the parameters in the core model with the parameters by training and learning data corresponding to gene markers of a plurality of groups of monitored people with errors; the group-based analysis unit also transmits the modified parameters to the cyclic modification unit;
the loop modification unit modifies the parameter in the 'core model with parameter' after acquiring the 'parameter after modification' each time and performs necessary loop modification.
The intestinal cancer model construction method based on 5hmC molecular marker detection further comprises the following steps:
before collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors and transmitting the data to a group-based analysis unit, obtaining data corresponding to gene markers of the plurality of groups of monitored persons with detection errors and determining data corresponding to gene markers of the plurality of groups of monitored persons with detection errors which should be eliminated; data corresponding to gene markers of a plurality of groups of monitored persons with 'wrong' detection errors are eliminated from the data transmitted to the group-based analysis unit.
The method has the advantages that the data corresponding to the gene markers of the monitored person are collected, the core model with the parameters is tested based on the data corresponding to the gene markers of the monitored persons of a plurality of groups, and the data corresponding to the gene markers of the monitored persons with wrong test are output; collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors; grouping parameters in the core model with the parameters according to a group analysis unit, performing primary analysis by taking each group of parameters as variable variables, and specifically, correcting the parameters in the core model with the parameters by training and learning data corresponding to gene markers of a plurality of groups of monitored people with detection errors; the parameters after modification are also transmitted to a cyclic modification unit; and then, parameters in the 'core model with parameters' are modified and necessary cyclic modification is executed, so that the parameters of the 'core model with parameters' are continuously optimized in the process of 'collecting data corresponding to the gene markers of the monitored person', namely in the process of using the 'core model with parameters', and the characterization logic of the 'core model with parameters' is continuously modified through the process of learning wrong cases, so that the accuracy of colorectal cancer judgment can be obviously improved.
Drawings
Fig. 1 is a block diagram of an embodiment of an intestinal cancer model based on 5hmC molecular marker detection according to the present application.
Fig. 2 is a block diagram of another embodiment of the intestinal cancer model based on 5hmC molecular marker detection according to the present application.
The present application is further illustrated below with reference to examples.
Detailed Description
In specific implementation, as shown in fig. 1, the embodiment of the intestinal cancer model based on 5hmC molecular marker detection of the present application includes a core model with parameters, the core model with parameters is an intestinal cancer model based on molecular markers as parameters, and further comprises a case collection unit, a case inspection unit, an error collection unit, a group-based analysis unit and a cyclic modification unit, the output end of the case collection unit is connected with the case inspection unit in an interconnecting way, the output end of the case inspection unit is connected with the error collection unit in an interconnecting way, the output end of the error collection unit is interactively connected with the group-based analysis unit, the group-based analysis unit is interactively connected with the cycle modification unit, the cyclic modification unit is bidirectionally and interactively connected with a core model with parameters, and the core model with parameters is also bidirectionally and interactively connected with the case inspection unit;
the case collection unit is used for collecting data corresponding to the gene marker of the monitored person, and the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to all the gene markers of the monitored person;
the case inspection unit is used for inspecting the core model with the parameters based on data corresponding to the gene markers of the monitored persons in a plurality of groups each time and outputting data corresponding to the gene markers of the monitored persons with errors in inspection;
the error collecting unit is used for collecting data corresponding to gene markers of a plurality of groups of monitored persons for detecting errors and transmitting the data to the group-by-group analyzing unit;
the group-based analysis unit is used for grouping parameters in the 'core model with parameters', and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises the step of correcting the parameters in the 'core model with parameters' by training and learning data corresponding to gene markers of a plurality of groups of monitored people with errors; and is also used for transmitting the modified parameters to the cyclic modification unit;
the loop modification unit is used for performing the ' parameter-carrying core model ' after each acquisition of the ' parameters after modification
Parameter modification and also for performing the above-described modification in a loop.
In the application, the case collection unit, the case inspection unit, the error collection unit, the group analysis unit and the cycle modification unit can be realized in a computer application layer.
In the implementation of the method, the case collecting unit of the intestinal cancer model based on 5hmC molecular marker detection collects data corresponding to the gene markers of the monitored person, wherein the data corresponding to the gene markers of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to all the gene markers of the monitored person;
the case inspection unit inspects the core model with the parameters based on the data corresponding to the gene markers of the monitored persons in a plurality of groups, and outputs the data corresponding to the gene markers of the monitored persons with the inspection errors;
the error collecting unit collects data corresponding to gene markers of a plurality of groups of monitored persons for detecting errors and transmits the data to the group-by-group analysis unit;
the analysis unit groups parameters in the core model with parameters, and analyzes each group of parameters as variable variables once, wherein the analysis specifically comprises that the parameters in the core model with parameters are modified by training and learning data corresponding to gene markers of a plurality of groups of monitored people with detection errors; and is also used for transmitting the modified parameters to the cyclic modification unit;
the cyclic modification unit is used for performing modification on the 'core model with parameters' after acquiring 'parameters after modification' each time
Parameter modification and also for performing the above-described modification in a loop.
Therefore, the application also discloses a method for constructing an intestinal cancer model based on 5hmC molecular marker detection, which comprises the following steps:
collecting data corresponding to the gene marker of the monitored person, wherein the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to all the gene markers of the monitored person;
checking the core model with the parameters based on data corresponding to the gene markers of the monitored persons in a plurality of groups, and outputting data corresponding to the gene markers of the monitored persons with wrong checking;
collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors, and transmitting the data to a group-based analysis unit;
grouping parameters in the core model with the parameters according to a group analysis unit, and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises the step of modifying the parameters in the core model with the parameters by training and learning data corresponding to gene markers of a plurality of groups of monitored people with errors; the group-based analysis unit also transmits the modified parameters to the cyclic modification unit;
the loop modification unit modifies the parameter in the 'core model with parameter' after acquiring the 'parameter after modification' each time and performs necessary loop modification.
It is understood that the timing sequence between the steps in the above embodiments is not fixed, for example, in the implementation, possible embodiments are:
s1, collecting data corresponding to the gene marker of the monitored person, wherein the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to each gene marker of the monitored person;
s2, checking the core model with parameters based on the data corresponding to the gene markers of the monitored persons in a plurality of groups, and outputting the data corresponding to the gene markers of the monitored persons with wrong checking;
s3, collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors and transmitting the data to a group-based analysis unit;
s4 grouping parameters in the 'core model with parameters' according to a group analysis unit, and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises that the parameters in the 'core model with parameters' are modified by training and learning 'data corresponding to gene markers of a plurality of groups of monitored people with errors are checked'; the group-based analysis unit also transmits the modified parameters to the cyclic modification unit;
the S5 loop modification unit modifies the parameter in the "core model with parameters" after each acquisition of the "parameters after modification" and performs necessary loop modification.
For example, in practice, possible embodiments are:
s1 grouping parameters in the 'core model with parameters' according to a group analysis unit, and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises that the parameters in the 'core model with parameters' are modified by training and learning 'data corresponding to gene markers of a plurality of groups of monitored people with errors are checked'; the group-based analysis unit also transmits the modified parameters to the cyclic modification unit;
s2, modifying the parameter in the 'core model with parameter' after the 'modified parameter' is obtained by the loop modification unit each time and executing necessary loop modification;
s3, collecting data corresponding to the gene marker of the monitored person, wherein the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to each gene marker of the monitored person;
s4, checking the core model with parameters based on the data corresponding to the gene markers of the monitored persons in a plurality of groups, and outputting the data corresponding to the gene markers of the monitored persons with wrong checking;
s5 collects data corresponding to gene markers of several groups of monitored subjects who tested for errors and transmits them to a group-wise analysis unit.
For example, in practice, possible embodiments are:
s1, modifying the parameter in the 'core model with parameter' after the 'modified parameter' is obtained by the loop modification unit each time and executing necessary loop modification;
s2, collecting data corresponding to the gene marker of the monitored person, wherein the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to each gene marker of the monitored person;
s3, checking the core model with parameters based on the data corresponding to the gene markers of the monitored persons in a plurality of groups, and outputting the data corresponding to the gene markers of the monitored persons with wrong checking;
s4, collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors and transmitting the data to a group-based analysis unit;
s5 grouping parameters in the 'core model with parameters' according to a group analysis unit, and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises that the parameters in the 'core model with parameters' are modified by training and learning 'data corresponding to gene markers of a plurality of groups of monitored people with errors are checked'; the group-wise analysis unit also transmits the modified parameters to the cyclic modification unit.
In the implementation of the method, data corresponding to the gene markers of the monitored person are collected, a core model with parameters is tested based on the data corresponding to the gene markers of a plurality of groups of monitored persons, and data corresponding to the gene markers of a plurality of groups of monitored persons with wrong testing are output; collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors; grouping parameters in the core model with the parameters according to a group analysis unit, and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises the step of modifying the parameters in the core model with the parameters by training and learning data corresponding to gene markers of a plurality of groups of monitored people with errors; the parameters after modification are also transmitted to a cyclic modification unit; and then, parameters in the 'core model with parameters' are modified and necessary cyclic modification is executed, so that the parameters of the 'core model with parameters' can be continuously optimized in the process of 'collecting data corresponding to the gene markers of the monitored person', namely in the process of using the 'core model with parameters', so that the characterization logic of the 'core model with parameters' is continuously modified through the process of learning wrong cases, and the accuracy of colorectal cancer judgment can be obviously improved.
In a preferred embodiment, the core model with parameters is tested based on data corresponding to several groups of the genetic markers of the monitored persons, and data corresponding to several groups of the genetic markers of the monitored persons with test errors is output, specifically: firstly, using 'parameters of weighted value corresponding to each gene marker of monitored person' and whole judgment threshold value parameter 'corresponding to original core model with parameters as reference to input' numerical value corresponding to each gene marker of monitored person 'in data corresponding to gene markers of several groups of monitored person' into original core model with parameters, and if the judgment result is inconsistent with the actual 'data item of confirmed diagnosis result of monitored person', then the inconsistent 'data corresponding to gene markers of several groups of monitored person' is the 'data corresponding to gene markers of several groups of monitored person who are wrong to be tested'.
In the preferred implementation, the training and learning "data corresponding to gene markers of several groups of monitored people with errors are checked" to further modify parameters in the "core model with parameters", specifically, several parameters in the "parameters with weights corresponding to various gene markers of monitored people and the overall judgment threshold parameters" corresponding to the original core model with parameters are calibrated as quantitative, the other parameters are variables, then the variables are subjected to a plurality of functional transformations, the transformed variables are used as new parameters, then inputting the numerical values corresponding to all the genetic markers of the monitored person in the data corresponding to the genetic markers of the monitored persons for detecting errors into the core model with parameters after the parameters are changed, if the judgment result is not consistent with the actual record in the 'confirmed diagnosis result data item of the monitored person'; continuously calibrating a plurality of parameters in the parameter of the weighted value corresponding to each gene marker of the monitored person and the integral judgment threshold parameter corresponding to the original core model with the parameters as the fixed quantity and the other parameters as the variables, then carrying out a plurality of function transformations on the variables, and taking the transformed variables as new parameters; until the transformed variables are used as new parameters, inputting the numerical values corresponding to each gene marker of the monitored person in the data corresponding to the gene markers of the monitored persons with the wrong detection into a core model with the parameters after the parameters are changed, and judging that the results are consistent with the records in the actual data items of the confirmed diagnosis results of the monitored persons; the latest parameters are used as parameters needed for modifying the 'core model with parameters'.
In a preferred implementation, the plurality of functional transformations include a forward functional transformation and a reverse functional transformation, wherein the forward functional transformation is a functional transformation that increases the weighted values of the variable-corresponding parameters, and wherein the reverse functional transformation is a functional transformation that decreases the weighted values of the variable-corresponding parameters.
In specific implementation, as shown in fig. 2, the embodiment of the intestinal cancer model detected based on the 5hmC molecular marker further includes a re-inspection unit and an inspection denoising unit, the re-inspection unit is interactively connected with the output end of the error collection unit, the re-inspection unit is also bidirectionally interactively connected with the case collection unit, the output end of the re-inspection unit is also interactively connected with the inspection denoising unit, the inspection denoising unit is interactively connected with the output end of the error collection unit, the output end of the inspection denoising unit is interactively connected with the group-by-group analysis unit, the re-inspection unit is used for obtaining "data corresponding to gene markers of several groups of detected persons with inspection errors" from the error collection unit and interacting with the case collection unit to determine that "data corresponding to gene markers of several groups of detected persons with inspection errors" should be eliminated, the inspection denoising unit is used for interacting with the error collection unit and the group-based analysis unit, acquiring data corresponding to gene markers of a plurality of groups of monitored persons from which 'wrong' inspection errors are to be eliminated from the secondary inspection unit, and transmitting the error collection unit to the group-based analysis unit to eliminate data corresponding to the gene markers of the plurality of groups of monitored persons from which 'wrong' inspection errors are to be eliminated.
In practice, both the re-inspection unit and the inspection denoising unit can be realized in a computer application layer.
In a specific implementation, the rechecking unit obtains the data corresponding to the gene markers of the groups of the monitored persons with the wrong detection from the error collecting unit and interacts with the case collecting unit to determine the data corresponding to the gene markers of the groups of the monitored persons with the wrong detection, the detection and denoising unit interacts with the error collecting unit and the group-by-group analyzing unit and obtains the data corresponding to the gene markers of the groups of the monitored persons with the wrong detection from the rechecking unit and transmits the data corresponding to the gene markers of the groups of the monitored persons with the wrong detection to the group-by-group analyzing unit.
Therefore, the intestinal cancer model construction method based on 5hmC molecular marker detection further comprises the following steps: before collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors and transmitting the data to a group-based analysis unit, obtaining data corresponding to gene markers of the plurality of groups of monitored persons with detection errors and determining data corresponding to gene markers of the plurality of groups of monitored persons with detection errors which should be eliminated; data corresponding to gene markers of a plurality of groups of monitored persons with 'wrong' detection errors are eliminated from the data transmitted to the group-based analysis unit.
In a preferred embodiment, the genetic marker of the monitored subject comprises a 5hmC multi-molecular marker based on an apparent modification.
In a preferred implementation, the parameters in the core model with parameters include parameters of weighted values corresponding to each genetic marker of a monitored person and an overall judgment threshold parameter; for example, the parametric core model may adopt a 5hmC multi-molecular marker and colorectal cancer early diagnosis model based on apparent modification disclosed in chinese patent document 201910026672.2 in the prior art, and when the parametric core model adopts a Logistic regression model, the parameters in the parametric core model include a constant term output in a modeling process, a coefficient of a gene marker 5-hmC expression level, and a score judgment threshold of the Logistic regression model.
Claims (9)
1. The intestinal cancer model based on 5hmC molecular marker detection comprises a core model with parameters, wherein the core model with parameters is specifically an intestinal cancer model based on molecular markers as parameters, and is characterized by further comprising a case collection unit, a case inspection unit, an error collection unit, a group-by-group analysis unit and a cyclic modification unit, wherein the output end of the case collection unit is in interactive connection with the case inspection unit, the output end of the case inspection unit is in interactive connection with the error collection unit, the output end of the error collection unit is in interactive connection with the group-by-group analysis unit, the group-by-group analysis unit is in interactive connection with the cyclic modification unit, the cyclic modification unit is in bidirectional interactive connection with the core model with parameters, and the core model with parameters is also in bidirectional interactive connection with the case inspection unit;
the case collection unit is used for collecting data corresponding to the gene marker of the monitored person, and the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to all the gene markers of the monitored person;
the case inspection unit is used for inspecting the core model with the parameters based on data corresponding to the gene markers of the monitored persons in a plurality of groups each time and outputting data corresponding to the gene markers of the monitored persons with errors in inspection;
the error collecting unit is used for collecting data corresponding to gene markers of a plurality of groups of monitored persons for detecting errors and transmitting the data to the group-by-group analyzing unit;
the group-based analysis unit is used for grouping parameters in the 'core model with parameters', and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises the step of correcting the parameters in the 'core model with parameters' by training and learning data corresponding to gene markers of a plurality of groups of monitored people with errors; and is also used for transmitting the modified parameters to the cyclic modification unit;
the loop modification unit is used for modifying the parameters of the 'core model with parameters' after acquiring the 'parameters after modification' each time, and is also used for circularly executing the modification.
2. The intestinal cancer model based on 5hmC molecular marker detection according to claim 1, further comprising a re-examination unit and a detection and denoising unit, wherein the re-examination unit is interactively connected with the output end of the error collection unit, the re-examination unit is also bidirectionally interactively connected with the case collection unit, the output end of the re-examination unit is also interactively connected with the detection and denoising unit, the detection and denoising unit is interactively connected with the output end of the error collection unit, the output end of the detection and denoising unit is interactively connected with the group-by-group analysis unit, the re-examination unit is used for obtaining the data corresponding to the gene markers of the detected errors from the error collection unit and interacting with the case collection unit to determine the data corresponding to the gene markers of the detected errors, the inspection denoising unit is used for interacting with the error collection unit and the group-based analysis unit, acquiring data corresponding to gene markers of a plurality of groups of monitored persons from which 'wrong' inspection errors are to be eliminated from the secondary inspection unit, and transmitting the error collection unit to the group-based analysis unit to eliminate data corresponding to the gene markers of the plurality of groups of monitored persons from which 'wrong' inspection errors are to be eliminated.
3. The model of intestinal cancer detected based on 5hmC molecular markers as claimed in claim 1, wherein the genetic markers of the monitored subject include 5hmC multi-molecular markers based on apparent modification.
4. The intestinal cancer model detected based on 5hmC molecular markers as claimed in claim 1, wherein the parameters in the core model with parameters include weighted parameters and overall judgment threshold parameters corresponding to each genetic marker of the monitored person.
5. The intestinal cancer model based on 5hmC molecular marker detection according to claim 1, wherein the data corresponding to gene markers of several groups of monitored subjects is used for testing a parametric core model, and data corresponding to gene markers of several groups of monitored subjects with testing errors is output, specifically: firstly, using 'parameters of weighted value corresponding to each gene marker of monitored person' and whole judgment threshold value parameter 'corresponding to original core model with parameters as reference to input' numerical value corresponding to each gene marker of monitored person 'in data corresponding to gene markers of several groups of monitored person' into original core model with parameters, and if the judgment result is inconsistent with the actual 'data item of confirmed diagnosis result of monitored person', then the inconsistent 'data corresponding to gene markers of several groups of monitored person' is the 'data corresponding to gene markers of several groups of monitored person who are wrong to be tested'.
6. The intestinal cancer model detected based on 5hmC molecular markers according to claim 1, wherein the parameters in the "parametric core model" are modified by learning the data corresponding to the "genetic markers of several groups of the detected subjects who are detected to be in error" through training, specifically, several parameters in the "weighted value parameters corresponding to each genetic marker of the detected subject, the whole judgment threshold parameter" corresponding to the original parametric core model are specified as fixed quantity, several other parameters are variable, then several function transformations are performed on the variables, the transformed variables are used as new parameters, and then the "numerical values corresponding to each genetic marker of the detected subject" in the data corresponding to the genetic markers of several groups of the detected subjects who are detected to be in error "are input into the parametric core model after the parameters are changed, if the judgment result is not consistent with the actual record in the 'confirmed diagnosis result data item of the monitored person'; continuously calibrating a plurality of parameters in the parameter of the weighted value corresponding to each gene marker of the monitored person and the integral judgment threshold parameter corresponding to the original core model with the parameters as the fixed quantity and the other parameters as the variables, then carrying out a plurality of function transformations on the variables, and taking the transformed variables as new parameters; until the transformed variables are used as new parameters, inputting the numerical values corresponding to each gene marker of the monitored person in the data corresponding to the gene markers of the monitored persons with the wrong detection into a core model with the parameters after the parameters are changed, and judging that the results are consistent with the records in the actual data items of the confirmed diagnosis results of the monitored persons; the latest parameters are used as parameters needed for modifying the 'core model with parameters'.
7. The model of intestinal cancer based on detection of 5hmC molecular markers according to claim 6, wherein the plurality of functional transformations include a forward functional transformation that increases the weighted value of the variable-corresponding parameter and an inverse functional transformation that decreases the weighted value of the variable-corresponding parameter.
8. The intestinal cancer model construction method based on 5hmC molecular marker detection is characterized by comprising the following steps: collecting data corresponding to the gene marker of the monitored person, wherein the data corresponding to the gene marker of the monitored person comprises an ID number data item of the monitored person, a diagnosis result data item of the monitored person and values corresponding to all the gene markers of the monitored person;
checking the core model with the parameters based on data corresponding to the gene markers of the monitored persons in a plurality of groups, and outputting data corresponding to the gene markers of the monitored persons with wrong checking;
collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors, and transmitting the data to a group-based analysis unit;
grouping parameters in the core model with the parameters according to a group analysis unit, and performing primary analysis by taking each group of parameters as variable variables, wherein the analysis specifically comprises the step of modifying the parameters in the core model with the parameters by training and learning data corresponding to gene markers of a plurality of groups of monitored people with errors; the group-based analysis unit also transmits the modified parameters to the cyclic modification unit;
the loop modification unit modifies the parameter in the 'core model with parameter' after acquiring the 'parameter after modification' each time and performs necessary loop modification.
9. The method for constructing a intestinal cancer model based on 5hmC molecular marker detection according to claim 8, further comprising the steps of:
before collecting data corresponding to gene markers of a plurality of groups of monitored persons with detection errors and transmitting the data to a group-based analysis unit, obtaining data corresponding to gene markers of the plurality of groups of monitored persons with detection errors and determining data corresponding to gene markers of the plurality of groups of monitored persons with detection errors which should be eliminated; data corresponding to gene markers of a plurality of groups of monitored persons with 'wrong' detection errors are eliminated from the data transmitted to the group-based analysis unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210250395.5A CN114373511B (en) | 2022-03-15 | 2022-03-15 | Intestinal cancer model based on 5hmC molecular marker detection and intestinal cancer model construction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210250395.5A CN114373511B (en) | 2022-03-15 | 2022-03-15 | Intestinal cancer model based on 5hmC molecular marker detection and intestinal cancer model construction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114373511A true CN114373511A (en) | 2022-04-19 |
| CN114373511B CN114373511B (en) | 2022-08-30 |
Family
ID=81146834
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210250395.5A Active CN114373511B (en) | 2022-03-15 | 2022-03-15 | Intestinal cancer model based on 5hmC molecular marker detection and intestinal cancer model construction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114373511B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004187562A (en) * | 2002-12-10 | 2004-07-08 | Jgs:Kk | Dna microarray data analyzing method, dna microarray data analyzer, program, and recording medium |
| US20090275057A1 (en) * | 2006-03-31 | 2009-11-05 | Linke Steven P | Diagnostic markers predictive of outcomes in colorectal cancer treatment and progression and methods of use thereof |
| US20100075323A1 (en) * | 2008-09-22 | 2010-03-25 | Advpharma, Inc. | Molecular markers for lung and colorectal carcinomas |
| WO2010056993A2 (en) * | 2008-11-14 | 2010-05-20 | Emory University | Prostate cancer biomarkers to predict recurrence and metastatic potential |
| CN103091492A (en) * | 2011-11-04 | 2013-05-08 | 中国科学院上海生命科学研究院 | Diagnostic reagent and kit for cancer |
| CA2947624A1 (en) * | 2014-05-13 | 2015-11-19 | Myriad Genetics, Inc. | Gene signatures for cancer prognosis |
| WO2016134191A1 (en) * | 2015-02-18 | 2016-08-25 | Singular Bio, Inc. | Assays for single molecule detection and use thereof |
| US20170119280A1 (en) * | 2015-10-29 | 2017-05-04 | Invoy Technologies, Llc | Flow regulation device for breath analysis and related method |
| WO2018001295A1 (en) * | 2016-06-30 | 2018-01-04 | 博奥生物集团有限公司 | Molecular marker, reference gene, and application and test kit thereof, and method for constructing testing model |
| CN109504778A (en) * | 2019-01-11 | 2019-03-22 | 复旦大学附属中山医院 | It is a kind of that model is early diagnosed based on the 5hmC polymolecular marker apparently modified and colorectal cancer |
| CN110991536A (en) * | 2019-12-02 | 2020-04-10 | 上海应用技术大学 | Training method of early warning model of primary liver cancer |
| CA3121923A1 (en) * | 2018-12-18 | 2020-06-25 | Wenying Pan | Methods for detecting disease using analysis of rna |
| WO2020223537A1 (en) * | 2019-05-01 | 2020-11-05 | Pact Pharma, Inc. | Compositions and methods for the treatment of cancer using a cdb engineered t cell therapy |
| CN112609015A (en) * | 2021-03-08 | 2021-04-06 | 天津奇云诺德生物医学有限公司 | Microbial marker for predicting colorectal cancer risk and application thereof |
| US20220059229A1 (en) * | 2015-06-02 | 2022-02-24 | Infervision Medical Technology Co., Ltd. | Method and apparatus for analyzing medical treatment data based on deep learning |
-
2022
- 2022-03-15 CN CN202210250395.5A patent/CN114373511B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2004187562A (en) * | 2002-12-10 | 2004-07-08 | Jgs:Kk | Dna microarray data analyzing method, dna microarray data analyzer, program, and recording medium |
| US20090275057A1 (en) * | 2006-03-31 | 2009-11-05 | Linke Steven P | Diagnostic markers predictive of outcomes in colorectal cancer treatment and progression and methods of use thereof |
| US20100075323A1 (en) * | 2008-09-22 | 2010-03-25 | Advpharma, Inc. | Molecular markers for lung and colorectal carcinomas |
| WO2010056993A2 (en) * | 2008-11-14 | 2010-05-20 | Emory University | Prostate cancer biomarkers to predict recurrence and metastatic potential |
| CN103091492A (en) * | 2011-11-04 | 2013-05-08 | 中国科学院上海生命科学研究院 | Diagnostic reagent and kit for cancer |
| CA2947624A1 (en) * | 2014-05-13 | 2015-11-19 | Myriad Genetics, Inc. | Gene signatures for cancer prognosis |
| WO2016134191A1 (en) * | 2015-02-18 | 2016-08-25 | Singular Bio, Inc. | Assays for single molecule detection and use thereof |
| US20220059229A1 (en) * | 2015-06-02 | 2022-02-24 | Infervision Medical Technology Co., Ltd. | Method and apparatus for analyzing medical treatment data based on deep learning |
| US20170119280A1 (en) * | 2015-10-29 | 2017-05-04 | Invoy Technologies, Llc | Flow regulation device for breath analysis and related method |
| WO2018001295A1 (en) * | 2016-06-30 | 2018-01-04 | 博奥生物集团有限公司 | Molecular marker, reference gene, and application and test kit thereof, and method for constructing testing model |
| CA3121923A1 (en) * | 2018-12-18 | 2020-06-25 | Wenying Pan | Methods for detecting disease using analysis of rna |
| CN109504778A (en) * | 2019-01-11 | 2019-03-22 | 复旦大学附属中山医院 | It is a kind of that model is early diagnosed based on the 5hmC polymolecular marker apparently modified and colorectal cancer |
| WO2020223537A1 (en) * | 2019-05-01 | 2020-11-05 | Pact Pharma, Inc. | Compositions and methods for the treatment of cancer using a cdb engineered t cell therapy |
| CN110991536A (en) * | 2019-12-02 | 2020-04-10 | 上海应用技术大学 | Training method of early warning model of primary liver cancer |
| CN112609015A (en) * | 2021-03-08 | 2021-04-06 | 天津奇云诺德生物医学有限公司 | Microbial marker for predicting colorectal cancer risk and application thereof |
Non-Patent Citations (3)
| Title |
|---|
| SAHAR OLSADAT SAJADIAN等: "Induction of active demethylation and 5hmC formation by 5-azacytidine is TET2 dependent and suggests new treatment strategies against hepatocellular carcinoma", 《SAJADIAN ET AL. CLINICAL EPIGENETICS》 * |
| 张新丽等: "长链非编码RNA LINC00152在结肠癌组织中的表达及临床意义", 《临床检验杂志》 * |
| 朱兴国: "内质网应激相关蛋白CHOP/GADD153在结肠癌组织中的表达及意义", 《中国普通外科杂志》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114373511B (en) | 2022-08-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108389201B (en) | Lung nodule benign and malignant classification method based on 3D convolutional neural network and deep learning | |
| CN111402979B (en) | Method and device for detecting consistency of disease description and diagnosis | |
| CN109791546A (en) | Target category characteristic model | |
| CN110352389A (en) | Information processing unit and information processing method | |
| CN102958425A (en) | Similar case history search device and similar case history search method | |
| JP2016200435A (en) | Mass spectrum analysis system, method, and program | |
| CN107203701A (en) | A kind of measuring method of fat thickness, apparatus and system | |
| CN113903082A (en) | Human body gait monitoring algorithm based on dynamic time planning | |
| CN117152152B (en) | Production management system and method for detection kit | |
| CN115205601A (en) | Medical examination result auditing system based on artificial intelligence and knowledge graph | |
| CN109190699A (en) | A kind of more disease joint measurement methods based on multi-task learning | |
| CN109165665A (en) | A kind of category analysis method and system | |
| KR101603308B1 (en) | Biological age calculation model generation method and system thereof, biological age calculation method and system thereof | |
| CN113380396A (en) | Method for evaluating risks of multiple intestinal diseases based on fecal microbial markers and human DNA content and application | |
| CN117522861B (en) | Intelligent monitoring system and method for animal rotator cuff injury | |
| CN114373511B (en) | Intestinal cancer model based on 5hmC molecular marker detection and intestinal cancer model construction method | |
| CN113360611A (en) | AI diagnosis method, device, storage medium and equipment based on inspection result | |
| CN110265140A (en) | Foot deformity detection model, foot deformity detection system and foot deformity detection method | |
| CN112102285B (en) | Bone age detection method based on multi-modal countermeasure training | |
| US8131662B2 (en) | Remote vision testing data collection | |
| CN109492690B (en) | Method for detecting CT image based on convolutional neural network | |
| CN107103134A (en) | Low-speed wireless sensor network testability analysis method based on Bayesian network | |
| CN113239075A (en) | Construction data self-checking method and system | |
| CN112528874B (en) | Human-oriented SAR image multi-target visual recognition capability quantization method, system and computer readable medium | |
| CN116741384B (en) | Bedside care-based severe acute pancreatitis clinical data management method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |