WO2024027591A1

WO2024027591A1 - Multi-cancer methylation detection kit and use thereof

Info

Publication number: WO2024027591A1
Application number: PCT/CN2023/109837
Authority: WO
Inventors: 李冰思; 许佳悦; 邱福俊; 汉雨生; 张之宏
Original assignee: 广州燃石医学检验所有限公司
Priority date: 2022-08-01
Filing date: 2023-07-28
Publication date: 2024-02-08
Also published as: CN117535404A

Abstract

Provided are a multi-cancer methylation detection kit and use thereof. Specifically, a biomarker combination for evaluating the correlation between a sample to be detected and a tumorigenesis risk and/or a tumor tissue source is provided, wherein the reference gene version for differentially methylated regions (DMRs) is the hg19 version. Also provided is use of a reagent of the biomarker combination in preparing a kit for determining a tumorigenesis risk and/or evaluating the tumor tissue source of a sample. The present invention is suitable for risk assessment and tissue tracing of various cancers, and has the advantages of cost efficiency and high accuracy.

Description

A multi-cancer methylation detection kit and its application

Technical field

This application relates to the field of biomedicine, specifically to a multi-cancer methylation detection kit and its application.

Background technique

DNA methylation is known to play an important role in the regulation of gene expression. Abnormal DNA methylation marks have been reported during the development and progression of various diseases, including cancer. As a high-resolution, high-throughput technology, DNA methylation sequencing is increasingly recognized for its role in cancer screening, diagnosis, and monitoring. Whole genome bisulfite sequencing (WGBS) is the gold standard for methylation sequencing, but it has become difficult to apply in clinical applications due to the severe damage to DNA during processing and the high cost of sequencing. More importantly, most regions of the human genome are not active in the development of cancer, and cancer-related mutations are often concentrated in certain specific regions, such as CpG islands, which provides a lot of opportunities for targeted sequencing. Good opportunity.

Nonetheless, the discovery and screening of cancer-related differentially methylated regions (DMRs) is challenging because population heterogeneity, including disease, age and other states, can bring about non-specific changes in methylation profiles, so cancer assessment During the establishment of the DOC model, these non-cancer but abnormal signals need to be processed. Finally, for the application of detection of multiple cancer types, the establishment of the tissue traceability TOO model is of great auxiliary significance for tracing the possible source organs of cancer mutations, determining downstream diagnosis and treatment paths, and saving medical costs.

Contents of the invention

This application establishes a low-cost, high-precision method that uses DNA or RNA oligonucleotide sequences to capture methylation variation regions in a variety of different cancers, as well as specific methylation characteristic regions in various organs. And judge the presence of tumor components (ctDNA) in blood cell-free DNA (cfDNA), and evaluate its tissue origin.

On the one hand, the present application provides a biomarker combination for assessing the correlation between a test sample and the risk of tumor formation, characterized in that the biomarker combination includes any of at least 10 differential alphas shown in Table 1A. Kylation region DMR, where the reference gene version involved in the DMR in the table is the hg19 version.

On the one hand, the present application provides a biomarker combination for evaluating the correlation between a test sample and the origin of a tumor tissue, characterized in that the biomarker combination includes any of at least 10 differential A shown in Table 1B. Kylation region DMR, where the reference gene version involved in the DMR in the table is the hg19 version.

On the one hand, the present application provides a biomarker combination for assessing the correlation between a test sample and the risk of tumor formation and/or the origin of tumor tissue, characterized in that the biomarker combination includes what is shown in Table 1C Any at least 10 differentially methylated region DMRs, wherein the reference gene version involved in the DMRs in the table is the hg19 version.

In one aspect, the present application provides a kit comprising the biomarker combination described in the present application, and optionally a second-generation high-throughput sequencing reagent.

In one aspect, the present application provides the use of a reagent for detecting a combination of biomarkers described in the present application in the preparation of a kit for diagnosing the risk of tumor formation and/or the origin of tumor tissue.

On the one hand, the present application provides a method for evaluating the correlation between a sample to be tested and the risk of tumor formation and/or the source of tumor tissue, the method comprising: detecting the methylation level of a combination of biomarkers in the sample to be tested , the biomarker combination includes the biomarker combination described in this application.

On the one hand, the present application provides a storage medium recording a program that can run the method described in the present application.

In one aspect, the application provides an apparatus comprising a storage medium as described herein, and the apparatus optionally includes a processor coupled to the storage medium, the processor configured to The program stored in the storage medium is executed to implement the method described in this application.

The biomarker combination, kit, method, equipment, storage medium and application of the present application can be suitable for risk assessment and tissue traceability of various cancers, and have the advantages of low cost and high accuracy.

Those skilled in the art will readily appreciate other aspects and advantages of the present application from the detailed description below. Only exemplary embodiments of the present application are shown and described in the following detailed description. As those skilled in the art will realize, the contents of this application enable those skilled in the art to make changes to the specific embodiments disclosed without departing from the spirit and scope of the invention covered by this application. Accordingly, the drawings and descriptions of the present application are illustrative only and not restrictive.

Description of the drawings

The specific features of the invention to which this application relates are set forth in the appended claims. The features and advantages of the invention to which this application relates can be better understood by reference to the exemplary embodiments described in detail below and the accompanying drawings. A brief description of the drawings is as follows:

Figure 1 shows an exemplary situation (a theoretical exemplary display, not used to represent actual sequencing situations).

Figure 2 shows another exemplary situation (a theoretical exemplary display, not used to represent actual sequencing situations).

Figures 3A-3C show another exemplary situation (a theoretical exemplary display, not used to represent actual sequencing situations).

Figure 4 shows that in 5-fold cross-validation, a tissue traceability accuracy of 98% (95% CI: 96-99%) can be achieved.

Figure 5 shows the control results of the weight configuration of the Salmon-DOC model of the present application for confusion-related features.

Figures 6A-6F show that the Salmon-DOC model of the present application can efficiently detect 6 cancer types at different stages in the tumor group model.

Figures 7A-7F show that in the healthy group, the Salmon-DOC model of this application has overcome the previous weakness of methylation false positives increasing with age, and maintained balance among various age groups (the horizontal axis is age, and the vertical axis is model cancer probability score).

Figures 8A-8D show that the traceability accuracy of the Salmon-TOO two-layer model of this application is better than that of the single-layer model in both cross-validation and independent verification.

Figure 9 shows the tissue traceability assessment results based on 103 TOO-related DMR areas.

Detailed ways

The implementation of the invention of the present application will be described below with specific examples. Those familiar with this technology can easily understand other advantages and effects of the invention of the present application from the content disclosed in this specification.

Definition of Terms

In this application, the term "differentially methylated region" (DMR) generally refers to a region of DNA containing one or more differentially methylated sites. For example, a DMR that includes a greater number or frequency of methylated sites under a selected condition of interest, such as a cancer state, may be termed a hypermethylated DMR. For example, a DMR that includes a smaller number or frequency of methylated sites under a selected condition of interest, such as a cancer state, may be termed a hypomethylated DMR.

In this application, the terms "second-generation gene sequencing (NGS)", high-throughput sequencing" or "next-generation sequencing" generally refer to the second-generation high-throughput sequencing technology and higher-throughput sequencing methods developed thereafter. Next-generation sequencing platforms include but are not limited to existing sequencing platforms such as Illumina. With the continuous development of sequencing technology, those skilled in the art can understand that other sequencing methods and devices can also be used for this method. For example, two Next-generation gene sequencing can have the advantages of high sensitivity, high throughput, high sequencing depth, or low cost. According to development history, influence, sequencing principles and technologies, there are mainly the following types: Massively Parallel Signature Sequencing (Massively Parallel Sequencing) Signature Sequencing (MPSS), Polony Sequencing, 454 pyro sequencing, Illumina (Solexa) sequencing, Ion semi conductor sequencing, DNA nano-ball sequencing ), Complete Genomics' DNA nanoarray and combined probe anchored ligation sequencing method, etc. The second-generation gene sequencing can make it possible to conduct a detailed and comprehensive analysis of the transcriptome and genome of a species, so it is also called deep sequencing (deep sequencing). For example, the method of this application can also be applied to first-generation gene sequencing, second-generation gene sequencing, third-generation gene sequencing or single molecule sequencing (SMS).

In this application, the term "sample to be tested" generally refers to the sample to be tested. For example, it can be detected whether one or more gene regions on the sample to be tested are modified.

In this application, the terms "polynucleotide," "nucleotide," "nucleic acid," and "oligonucleotide" are used interchangeably. They represent polymeric forms of nucleotides (deoxyribonucleotides or ribonucleotides) of any length, or analogs thereof. Polynucleotides can have any three-dimensional structure and can perform any function, whether known or unknown. The following are non-limiting examples of polynucleotides: coding or non-coding regions of genes or gene fragments, genetic loci (loci) defined by linkage analysis, exons, introns, messenger RNA (mRNA), transfer RNA (tRNA), ribosomal RNA (rRNA), short interfering RNA (siRNA), short-hairpin RNA (shRNA), microRNA (miRNA), ribozyme, cDNA, recombinant polynucleotide, branched polynucleotide, Plasmids, vectors, isolated DNA of any sequence, isolated RNA of any sequence, nucleic acid probes, primers and adapters. Polynucleotides may include one or more modified nucleotides, such as methylated nucleotides and nucleotide analogs.

In this application, the term "methylation" generally refers to the methylation state possessed by gene fragments, nucleotides or bases thereof in this application. For example, the DNA segment in which the gene in this application is located may have methylation on one or more strands. For example, the DNA fragment in which the gene in this application is located may have methylation at one site or DMR or multiple sites or DMRs.

In this application, the term "human reference genome" generally refers to the human genome that can serve as a reference in gene sequencing. For information on the human reference genome, please refer to UCSC. The human reference genome can have different versions, for example, it can be hg19, GRCH37 or ensemble 75.

In this application, the term "machine learning model" generally refers to a system or collection of program instructions and/or data configured to implement an algorithm, process, or mathematical model. In this application, the algorithm, process or mathematical model can evaluate and provide a desired output based on given inputs. In this application, the parameters of the machine learning model may not be explicitly programmed, and in the traditional sense, the machine learning model may not be explicitly designed to follow specific rules in order to provide the desired results for a given input. output. For example, the use of a machine learning model may mean that the machine learning model and/or the data structure/set of rules that is the machine learning model is trained by a machine learning algorithm.

In this application, the term "comprising" generally means the inclusion of explicitly specified features, but not the exclusion of other elements.

In this application, the term "about" generally refers to a variation within the range of 0.5% to 10% above or below the specified value, such as 0.5%, 1%, 1.5%, 2%, 2.5%, above or below the specified value. 3%, 3.5%, 4%, 4.5%, 5%, 5.5%, 6%, 6.5%, 7%, 7.5%, 8%, 8.5%, 9%, 9.5%, or 10%.

In order to detect six cancer types with high incidence and fatality rates, including lung cancer, intestinal cancer, liver cancer, ovarian cancer, pancreatic cancer, and esophageal cancer, this application uses a combination of public database (TCGA) and internal data mining. A new algorithm compares the methylation variation and spatial location of the genome simultaneously, and selects a total of 2536 variant regions (differentially methylated region, DMR) that are highly related to cancer.

Contents of the invention

In one aspect, the present application provides a biomarker combination for assessing the correlation between a test sample and the risk of tumor formation, the biomarker combination comprising any of at least 10 differentially methylated region DMRs shown in Table 1A , where the reference gene version involved in the DMR in the table is the hg19 version.

For example, the biomarker panel includes the 94 DMRs in Table 1A. For example, the biomarker panel includes approximately 94 DMRs in Table 1A, Any at least about 90 DMR, Any at least about 80 DMR, Any at least about 70 DMR, Any at least about 60 DMR, Any at least about 50 DMR, Any at least about 40 DMR, Any at least about 30 DMR, Any At least about 20 DMR, or optionally at least about 10 DMR.

On the one hand, the present application provides a biomarker combination for evaluating the correlation between a test sample and the origin of a tumor tissue, the biomarker combination comprising any of at least 10 differentially methylated region DMRs shown in Table 1B , where the reference gene version involved in the DMR in the table is the hg19 version.

For example, the biomarker panel includes the 103 DMRs in Table 1B. For example, the biomarker panel includes about 103 DMRs in Table IB, any at least about 100 DMRs, any at least about 90 DMRs, any at least about 80 DMRs, any at least about 70 DMRs, any at least about 60 DMRs DMR, any at least about 50 DMR, any at least about 40 DMR, any at least about 30 DMR, any at least about 20 DMR, or any at least about 10 DMR.

In one aspect, the present application provides a biomarker combination for assessing the correlation between a test sample and the risk of tumor formation and/or the origin of tumor tissue, the biomarker combination comprising at least 10 of any of the components shown in Table 1C DMRs of differentially methylated regions, where the reference gene version involved in the DMRs in the table is the hg19 version.

For example, the biomarker panel includes any of at least 222 DMRs in Table IE. For example, the biomarker panel includes about 222 DMRs, any at least about 220 DMRs, any at least about 210 DMRs, any at least about 200 DMRs, any at least about 150 DMRs, any at least about 100 DMRs in Table 1E DMR, any at least about 90 DMR, any at least about 80 DMR, any at least about 70 DMR, any at least about 60 DMR, any at least about 50 DMR, any at least about 40 DMR, any at least about 30 DMR , any at least about 20 DMR, or any at least about 10 DMR.

For example, the biomarker panel includes the 488 DMRs in Table ID. For example, the biomarker panel includes about 488 DMRs, any at least about 480 DMRs, any at least about 450 DMRs, any at least about 400 DMRs, any at least about 300 DMRs, any at least about 200 DMRs in Table ID DMR, any at least about 150 DMR, any at least about 100 DMR, any at least about 90 DMR, any at least about 80 DMR, any at least about 70 DMR, any at least about 60 DMR, any at least about 50 DMR , any at least about 40 DMR, any at least about 30 DMR, any at least about 20 DMR, or any at least about 10 DMR.

For example, the biomarker panel includes the 860 DMRs in Table 1C. For example, the biomarker panel includes about 860 DMRs, any at least about 850 DMRs, any at least about 800 DMRs, any at least about 700 DMRs, any at least about 600 DMRs, any at least about 500 DMRs in Table 1C DMR, 400 DMR, any at least about 300 DMR, any at least about 200 DMR, any at least about 150 DMR, any at least about 100 DMR, any at least about 90 DMR, any at least about 80 DMR, any at least About 70 DMRs, any at least about 60 DMRs, any at least about 50 DMRs, any at least about 40 DMRs, any at least about 30 DMRs, any at least about 20 DMRs, or any at least about 10 DMRs.

For example, the tumors are derived from homogenous tumors, heterogeneous tumors, hematological cancers, and/or solid tumors. For example, the tumor is from one or more of the following group of cancers: brain cancer, lung cancer, skin cancer, nasopharyngeal cancer, throat cancer, liver cancer, bone cancer, lymphoma, pancreatic cancer, skin cancer, bowel cancer , Rectal cancer, thyroid cancer, bladder cancer, kidney cancer, oral cancer, gastric cancer, solid tumors, ovarian cancer, esophageal cancer, gallbladder cancer, bile duct cancer, breast cancer, cervical cancer, uterine cancer, prostate cancer, head and neck cancer, sarcoma, Extrapulmonary thoracic malignancies, melanoma, and testicular cancer. For example, the tumor includes lung cancer, intestinal cancer, liver cancer, ovarian cancer, pancreatic cancer, and/or esophageal cancer.

In one aspect, the present application provides a kit comprising the biomarker combination described in the present application, and optionally a second-generation high-throughput sequencing reagent. For example, the kit can be used to assess the correlation of the sample to be tested with the risk of tumor formation and/or the origin of the tumor tissue.

In one aspect, the present application provides the use of a reagent for detecting a combination of biomarkers described in the present application in the preparation of a kit for diagnosing the risk of tumor formation and/or the origin of tumor tissue. For example, the tumors are derived from homogenous tumors, heterogeneous tumors, hematological cancers, and/or solid tumors. For example, the tumor is from one or more of the following group of cancers: brain cancer, lung cancer, skin cancer, nasopharyngeal cancer, throat cancer, liver cancer, bone cancer, lymphoma, pancreatic cancer, skin cancer, bowel cancer , Rectal cancer, thyroid cancer, bladder cancer, kidney cancer, oral cancer, gastric cancer, solid tumors, ovarian cancer, esophageal cancer, gallbladder cancer, bile duct cancer, breast cancer, cervical cancer, uterine cancer, prostate cancer, head and neck cancer, sarcoma, Extrapulmonary thoracic malignancies, melanoma, and testicular cancer. For example, the tumor includes lung cancer, intestinal cancer, liver cancer, ovarian cancer, pancreatic cancer, and/or esophageal cancer.

On the one hand, the present application provides a method for evaluating the correlation between a sample to be tested and the risk of tumor formation and/or the tumor tissue source of the sample. The method includes: performing methylation on a combination of biomarkers in the sample to be tested. The biomarker combination includes the biomarker combination described in this application.

For example, the sample is selected from the group consisting of tissue sample, blood sample, saliva, sputum, pleural effusion, lung lavage fluid, peritoneal effusion, peritoneal lavage fluid, enema and cerebrospinal fluid.

On the one hand, the present application provides a storage medium recording a program that can run the method described in the present application. For example, the non-volatile computer-readable storage medium may include a floppy disk, a flexible disk, a hard disk, a solid state storage (SSS) (such as a solid state drive (SSD)), a solid state card (SSC), a solid state module (SSM)), an enterprise class flash drives, tapes, or any other non-transitory magnetic media, etc. Non-volatile computer-readable storage media may also include punched cards, paper tape, cursor pads (or any other physical media having a hole pattern or other optically identifiable markings), compact disk read-only memory (CD-ROM) , Compact Disc Rewritable (CD-RW), Digital Versatile Disc (DVD), Blu-ray Disc (BD) and/or any other non-transitory optical media.

Without intending to be limited by any theory, the following examples are only to illustrate the kits, methods and uses of the present application, and are not intended to limit the scope of the invention of the present application.

Example

Example 1

Exemplary bisulfite-treated second-generation sequencing is performed on the sample, and the obtained sequencing data includes the methylation level and sequencing coverage depth of the methylation site CpG. Optionally, perform noise removal on genomic methylation signal CpG and noise region CHH/CHG sites. Then, the p-value obtained by weighted logistic regression is calculated for the "tumor" (C) and "normal" (N) groups. The explanatory variable of the logistic regression adopts a continuous variable, that is, the methyl group of each CpG point. ization level, the response variable takes a binary output, that is, (0, 1), corresponding to C and N. Weighted logistic regression tests the difference between C and N at each CpG site. The null hypothesis is that the difference between C and N at the CpG site is not statistically significant. The weight is based on each CpG. Determined by the depth of coverage of the site.

DMR division

Based on the methylation level and sequencing coverage depth of the methylation site CpG, determine how each region of the DMR is divided. Specifically, the methylation level and sequencing coverage depth of the methylation site CpG were calculated according to the following formula:

Here d _ij is the effective coverage depth of the j-th site in the i-th sample of group C, M _ij is the methylation level of the j-th site in the i-th sample of group C, and the methyl group of continuous sites in the genome space is The level of similarity is evaluated. The deeper the coverage depth and the larger the value of parameter P, the higher the similarity in methylation levels between adjacent CpG sites in the same group.

For the first CpG site in the region, sample A and sample B each obtained coverage of 500 effective sequences, and sample C obtained coverage of 200 effective sequences. For sample A, the methylation level of this CpG site is 0.2. The methylation level of the second CpG site in sample A is 0. For three samples, the coverage depth parameter value P of the first CpG site in this group was calculated to be 0.617. At this time, β _ij =|0.2-0|*e ^(1-0.617) =0.29. At the same time, considering that the difference in methylation levels between the two adjacent CpG sites is less than 0.25, which is one of the necessary conditions for classifying the two adjacent sites into the same DMR, then the first and second CpG sites in this example will not be classified into the same DMR.

If the above samples are replaced by A, B, and D (sample D has obtained 400 effective sequence coverage at the first CpG site). Similarly, for sample A, the methylation level of this CpG site is 0.2. The methylation level of the second CpG site of sample A is 0. However, due to the increased sequencing coverage depth of sample D in this example, the coverage depth parameter value P of the first CpG site in the group is calculated to be 0.962 for the three samples. At this time, β _ij =|0.2-0|*e ^(1-0.962) =0.21, which is less than the threshold 0.25 for dividing into the same DMR. At this time, according to sample A, the first and second CpG sites in this example have A prerequisite for being classified into the same DMR.

Therefore, introducing the coverage depth of CpG sites through the method of this application can significantly improve the accuracy of DMR region division.

Further optionally, for B _ij within a region, the calculation method is as follows:

Figures 3A-3C show another exemplary situation (a theoretical exemplary display, not used to represent actual sequencing situations). When the DMR region contains 10 CpG sites, the B _ij of all samples are combined and averaged to calculate the score of each DMR.

The calculation steps for the B value in the DMR area shown in Group A are as shown in Table 1 below:

Table 1 Calculation of B values within the DMR area of Group A

The B value is divided into 0.1, that is

Similarly, the B value within the DMR shown in group B is divided into 0.7, i.e., The B value within the DMR shown in Group C is divided into 1.233, i.e.,

The DMR regions screened by this method not only contain cancer mutation information of various cancer types, but also contain tissue-specific features, and have better segmentation effects at regional boundaries.

Example 2

Cancer assessment (DOC) model establishment

At different stages of development of different cancers, the ctDNA content in the blood varies greatly and is easily affected by the experimental batch effect. In addition, methylation variations are related to age, disease, race, etc. If these are not dealt with, they may affect the accuracy of the classification model as confounding variables. This application adopts a model construction method called Salmon, which first quantifies the bias caused by confounding variables (the quantification method can use but is not limited to the Hilbert-Schmidt independence criterion), and then embeds the regularization of the model. Regularization is used to correct the model to increase model accuracy and generalizability.

Algorithm establishment

_Assume _m _samples _, set _the _feature _vector An n-dimensional vector represents the methylation feature of sample i, y _i is the classification label of _xi , y _i ∈{-1,+1}, z _i is some kind of confounding variable of sample i.

Here L _H refers to the Hilbert-Schmidt independence criterion, which is used to measure the degree of independence of variables X and Z. h(y) and h(z) are the kernel functions of Y and Z (Kernel function), P _h(x)h(z) represents the probability distribution of h(y) and h(z), F and G represent the reproducing kernel Hilbert space of X and Z respectively. , can be understood as the domain mapped after nonlinear processing of X and Z, C _h(x)h(z) refers to the correlation coefficient of these two kernel functions, HS is the Hilbert space Space).
‖C _h(y)h(z) ‖ ² =(E _h(x)h(z) -E _h(x) E _h(z) ) ²
=(E _h(x)h(z) ) ² +(E _h(x) E _h(z) ) ² -2E _h(x)h(z) E _h(x) E _h(z)

Use support vector machine (SVM, support vector machine) as the main classifier.
f(x;w,b)=sgn(wTx+b)
sgn(a)＝1(-1)if a≥0(<0)

The classification interface is determined by solving the following objective equation,

st y _i (wTx+b)≥1

For inseparable data, soft-margin support vector machine (soft-margin SVM) introduces a penalty term for training errors.

st y _i (wTx+b)≥1-ξ _i
ξ _i ≥0

Here C controls the balance of minimizing training error and maximizing the classification margin, while ξ _i refers to the degree to which sample _xi violates the equation.

In order to control confounding factors, Salmon adds regular terms to the objective equation solved by SVM. The parameter λ controls the balance between confounding factor errors and maximizing the boundary width during training. The objective equation is:

st y _i (wTx+b)≥1-ξ _i
ξ _i ≥0

Here C and λ control the balance of minimizing training error, minimizing the correlation between confounding variables and explanatory variables, and maximizing the classification interval.

Each data point represents a blood sample used for Salmon-DOC model construction. The horizontal axis is the confunding factor of the corresponding sample, and the vertical axis is the original uncorrected variable coef (Figure A) and the corrected variable coef (Figure B). ). Comparison before and after correction shows that the weight of confusion-related features in Salmon-DOC is controlled.

Review cohort data

This application uses retrospective clinical samples of 6 cancer types, divided into training set and validation set, to evaluate the accuracy of Salmon's binary classifier (cancer vs non-cancer).

Figures 7A-7F show that in the healthy group, the Salmon-DOC model of this application has overcome the weakness of the previous increase in methylation false positives with age. Maintain balance among age groups (the horizontal axis is age and the vertical axis is model cancer probability score).

Example 3

Organizational traceability (TOO) model establishment

First layer TOO model construction

The essence of the TOO model is a multi-classification problem. The probability calculation for each category can be simplified to voting on the pairwise results, and then selecting the result with the most votes. However, for possible clinical applications of the tissue traceability model, it is not enough to generate a classification result. Only by generating the probability of classification can the assembly of the model become possible.

Therefore, the first step of the Salmon-TOO model in this application is to quantify the two-category voting results. This quantification can be proven by probability calculations. If we define a certain data point x and label y, we assume that the pairwise classification probability μ _ij exists, then from the i-th and j-th categories in the training set, we can get a model as long as any new data point x is input , that is, the calculated r _ij can be used as an approximate estimate of μ _ij . The problem can be simplified to estimating the probability of the i-th class using all r _ij :
p _i =P(y=i|x),i=1,…,k

Define r _ij as the estimate of μ _ij , assuming μ _ij +μ _ji =1. For multi-classification problems, a "voting" system is used,
μ _ij ≡P(y＝i|y＝i or j,x)

Define I as the objective equation: I _{x} = 1 if x is true, otherwise false. The probability calculation can be written as:

Second layer TOO model construction

The second layer of the Salmon-TOO model is MLR fitting for different categories.

Assuming that it is necessary to calculate the probability of the source of the tissue, a quantified two-class probability can be obtained based on the first layer, and the value range is (∞,-∞). Since the actual distribution of each pair of binary classification probabilities is inconsistent, the quantified binary classification probabilities can be further used as explanatory variables for logistic regression. The response variables adopt multivariate outputs, corresponding to known tissue sources in the modeling process.

Table 2 Two-category evaluation probabilities corresponding to pairwise organizational categories

As shown in Table 2 above, each column represents a characteristic variable of logistic regression That is, the probability of binary classification evaluation of pairwise organizational categories; each row represents a response variable y ₁ , that is, the organizational category (class).

It is a characteristic variable used to explain the binary classification probability. Assuming that there are J non-continuous reflection variables, the evaluation results are converted into Y _i1 ,..., Y _iJ , and β _j is the feature weight based on each reflection variable.

Since in the Salmon-DOC model, we can get that some cancer types are judged as negative, and some cancer types are judged as positive, so for this judgment, when performing traceability modeling, the tissue category ( class) performed weight correction based on the quasi-maximum likelihood estimation method. Taking binary logistic regression as an example, it can be explained as:

Review cohort data

All data in the retrospective cohort were randomly split into training and validation sets at a ratio of 1:1. First, the traceability evaluation results are obtained through cross-validation on the training set. During this process, the model parameters are continuously optimized and finally locked. Finally, all data in the validation set are evaluated with the locked model for traceability results. In the traceability model training set, there were a total of 300 cases of six cancer types, and the numbers of each cancer type and stage were relatively balanced: 36 cases of lung cancer (the number of cases in stages I to IV were 4/12/5/15 respectively), and 62 cases of intestinal cancer ( The number of cases in stages I to IV was 8/18/18/18 respectively), 74 cases of liver cancer (the number of cases in stages I to IV were 25/14/22/13 respectively), and 48 cases of ovarian cancer (the number of cases in stages I to IV were respectively 1/4/38/5), 40 cases of pancreatic cancer (number of cases in stages I to IV were 3/6/13/18 respectively), and 42 cases of esophageal cancer (number of cases in stages I to IV were 5/10/15 respectively) /12). The traceability model verification set has a total of 224 samples, including: 31 cases of lung cancer (the number of cases in stages I to IV is 4/5/12/10 respectively), and 52 cases of intestinal cancer (the number of cases in stages I to IV is 7/15/13 respectively) /17), 55 cases of liver cancer (number of cases in stages I to IV were 17/11/20/7 respectively), 27 cases of ovarian cancer (number of cases in stages I to IV were 3/4/8/12 respectively), 25 cases of pancreatic cancer There were 34 cases of esophageal cancer (the number of cases in stages I to IV was 4/6/6/9 respectively), and 34 cases of esophageal cancer (the number of cases in stages I to IV were 4/7/8/15 respectively).

Figure 8 shows that the traceability accuracy of the Salmon-TOO two-layer model of this application is better than that of the single-layer model in both cross-validation and independent verification.

Figures 8A and 8B show the traceability evaluation results of the cross-validation of the six cancer types data in the six cancer types training set. Among them, Figure 8A shows the output result after only building the first layer TOO model. The traceability accuracy is 0.87 (260/300). If the suboptimal traceability results are included, the accuracy is 0.93 (279/300); Figure 8B is Based on the first-layer TOO model and the output result of the second-layer MLR model, the traceability accuracy is increased to 0.90 (270/300). If sub-optimal traceability results are included, the accuracy can be further improved to 0.95 (284/284/ 300). Similarly, Figures 8C and 8D show the traceability evaluation results of the independent verification of the six cancer types in the above verification set. Among them, Figure 8C shows the output result after only building the first layer of TOO model. The traceability accuracy is 0.77 (173/224). If the suboptimal traceability results are included, the accuracy is 0.87 (194/224); Figure 8D is After supplementing the output results of the second-layer MLR model based on the first-layer TOO model, the traceability accuracy is improved to 0.84 (187/224). If the sub-optimal traceability results are included, the accuracy can be further improved to 0.89 (199/224 ).

To sum up, the evaluation accuracy of the Salmon-TOO two-layer traceability model of this application is better than that of the single-layer model in both training set cross-validation and independent verification.

Example 4. ADCC activity detection of anti-CCR8 chimeric antibodies

DOC cancer detection model

Table 3A shows the list of 94 DMR regions used in the DOC cancer detection model.

Table 3A List of DMR regions used for DOC cancer detection model

Based on 94 DOC-related DMR regions, 100 healthy human samples and 318 six cancer-positive samples in independent validation set 1 were evaluated, with an overall sensitivity of 80.5% (256/318) and an overall specificity of 95% (95 /100). While maintaining the specificity at 90% level, the specific cancer types and staging sensitivities are shown in Table 3B below:

Table 3B Evaluation results of sensitivity of staging of different cancer types

Then repeat the test, using 50 random areas out of the 94 DOC areas for each test. While maintaining the specificity at the 90% (90/100) level, the sensitivity results of six cancer-positive samples in five repeated tests are shown in Table 3C below:

Table 3C Sensitivity results of six cancer-positive samples in five repeated tests with specificity at 90% level

Example 5. Anti-CCR8 chimeric antibody inhibits the calcium flux signal of target cells caused by CCL1

TOO organizational traceability model

Table 4A shows the list of 103 DMR areas used in the TOO organization traceability model.

Table 4A List of DMR areas used in the TOO organization traceability model

Based on 103 TOO-related DMR areas, 473 six-cancer positive samples in the independent verification set 2 were evaluated for traceability. The accuracy of the first traceability was 63.0% (298/473). If the suboptimal traceability results are included, the accuracy can be improved. to 71.5% (338/473).

Then four rounds of repeated testing were conducted, each time using 50 random ones from the 103 TOO areas. The traceability accuracy results in the four rounds of evaluation are shown in Table 4B below:

Table 4B Traceability accuracy results of four rounds of repeated testing assessments

Example 6

DMR evaluates DOC and TOO simultaneously:

Table 5A shows the list of 860 DMR regions used for the DOC and TOO evaluation models.

Table 5A List of DMR regions for DOC and TOO evaluation models

Table 5B shows the list of 488 DMR regions used for the DOC and TOO evaluation models.

Table 5B List of DMR regions for DOC and TOO evaluation models

Table 5C shows the list of 222 DMR regions used for the DOC and TOO evaluation models.

Table 5C List of DMR regions for DOC and TOO evaluation models

In the independent validation set 3, for 473 negative samples and 473 positive six cancer samples, with the number of markers gradually compressed, the sensitivity and traceability at a unified specificity of 95.1% (450/473) were calculated accuracy. The evaluated tumor detection and tissue traceability results are shown in Tables 5D and 5E below:

Table 5D Tumor detection results in different DMRs

Table 5E Tissue traceability results in different DMRs

The foregoing detailed description is provided by way of explanation and example, and is not intended to limit the scope of the appended claims. Various modifications to the embodiments described herein will be apparent to those of ordinary skill in the art and remain within the scope of the appended claims and their equivalents.

Claims

A biomarker combination for assessing the correlation between a test sample and the risk of tumor formation, characterized in that the biomarker combination includes any of at least 10 differentially methylated region DMRs shown in Table 1A, wherein the The reference gene version involved in the DMR in the table is the hg19 version.
The biomarker combination of claim 1, said biomarker combination comprising any at least 50 DMRs in Table 1A.
The biomarker combination of any one of claims 1-2, said biomarker combination comprising 94 DMRs in Table 1A.
A biomarker combination for assessing the correlation between a test sample and the origin of a tumor tissue, characterized in that the biomarker combination includes any of at least 10 differentially methylated region DMRs shown in Table 1B, wherein the The reference gene version involved in the DMR in the table is the hg19 version.
The biomarker combination of claim 4, comprising any of at least 50 DMRs in Table IB.
The biomarker combination of any one of claims 4-5, said biomarker combination comprising 103 DMRs in Table 1B.
A biomarker combination for assessing the correlation between a test sample and the risk of tumor formation and/or the origin of tumor tissue, characterized in that the biomarker combination includes any of at least 10 differential A shown in Table 1C Kylation region DMR, where the reference gene version involved in the DMR in the table is the hg19 version.
The biomarker combination of claim 7, said biomarker combination comprising any at least 50 DMRs in IE, Table 1D or Table 1C.
The biomarker combination of any one of claims 7-8, said biomarker combination comprising 222 DMRs in Table 1E.
The biomarker combination of any one of claims 7-9, said biomarker combination comprising 488 DMRs in Table 1D.
The biomarker combination of any one of claims 7-10, said biomarker combination comprising 860 DMRs in Table 1C.
The biomarker combination of any one of claims 1-11, the tumor is derived from homogenous tumors (homogenous tumors), heterogeneous tumors, blood cancers and/or solid tumors; preferably, the tumor is derived from One or more cancers from the following groups: brain cancer, lung cancer, skin cancer, nasopharyngeal cancer, throat cancer, liver cancer, bone cancer, lymphoma, pancreatic cancer, skin cancer, bowel cancer, rectal cancer, thyroid cancer, Bladder cancer, kidney cancer, oral cancer, gastric cancer, solid tumors, ovarian cancer, esophageal cancer, gallbladder cancer, bile duct cancer, breast cancer, cervical cancer, uterine cancer, prostate cancer, head and neck cancer, sarcoma, thoracic malignant tumors except lung, melanoma, and testicular cancer.
The biomarker combination of any one of claims 1-12, the tumor comprising lung cancer, intestinal cancer, liver cancer, ovarian cancer, pancreatic cancer, and/or esophageal cancer.
A kit comprising the biomarker combination according to any one of claims 1-13 and optionally a second-generation high-throughput sequencing reagent.
The kit according to claim 14, which is used to evaluate the correlation between the sample to be tested and the risk of tumor formation and/or the origin of tumor tissue.
Use of a reagent for detecting the biomarker combination according to any one of claims 1 to 13 in the preparation of a kit for diagnosing the risk of tumor formation and/or the origin of tumor tissue.
The application of claim 16, wherein the tumor is derived from homogenous tumors, heterogeneous tumors, blood cancers and/or solid tumors; preferably, the tumor is derived from one of the following groups of cancers or more: brain cancer, lung cancer, skin cancer, nasopharyngeal cancer, throat cancer, liver cancer, bone cancer, lymphoma, pancreatic cancer, skin cancer, bowel cancer, rectal cancer, thyroid cancer, bladder cancer, kidney cancer, oral cancer , gastric cancer, solid tumors, ovarian cancer, esophageal cancer, gallbladder cancer, bile duct cancer, breast cancer, cervical cancer, uterine cancer, prostate cancer, head and neck cancer, sarcoma, thoracic malignant tumors other than lung, melanoma, and testicular cancer.
The use according to any one of claims 16-17, wherein the tumor comprises lung cancer, intestinal cancer, liver cancer, ovarian cancer, pancreatic cancer, and/or esophageal cancer.
A method for assessing the correlation between a sample to be tested and the risk of tumor formation and/or the source of tumor tissue, the method comprising: detecting methylation levels of a combination of biomarkers in the sample to be tested, the biomarker combination Comprising the biomarker combination of any one of claims 1-13.
The evaluation method of claim 19, wherein the sample is selected from the group consisting of tissue samples, blood samples, saliva, sputum, pleural effusion, lung lavage fluid, peritoneal effusion, peritoneal lavage fluid, enema and cerebrospinal fluid.
A storage medium recording a program capable of executing the method according to any one of claims 19-20.
An apparatus comprising the storage medium of claim 21 , and optionally comprising a processor coupled to the storage medium, the processor configured to perform data based on the storage medium stored in the storage medium. The program in is executed to implement the method described in any one of claims 19-20.