Cancer Prognosis Methods
By detecting the differences in DNA methylation levels in cancer tissues and adjacent tissues, and mathematical modeling is used to use the MD ratio evaluation system to solve the problem of insufficient cancer prognosis accuracy in the prior art, achieving more efficient and accurate tumors Recurrence risk prediction reduces the burden on patients.
Patent Information
- Application Number
- CN202110496484.3
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-05-09
- Filing Date
- 2021-05-07
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2041-05-07
AI Technical Summary
The prior art lacks accuracy in cancer prognosis, especially for early detection and prediction of tumor recurrence. Traditional methods rely on frequent imaging follow-up, resulting in a large and lagging burden on patients.
Using the MD ratio evaluation system based on DNA methylation detection, the difference in methylation levels is detected by high-throughput sequencing of the genomes of cancer tissues and adjacent tissues, and the prognostic analysis is carried out using mathematical modeling to calculate the difference in methylation levels. Information to predict the risk of tumor recurrence in patients.
The system can predict the risk of recurrence in cancer patients more efficiently and accurately, reduce the burden of postoperative follow-up, and provide personalized prognosis management, which has higher accuracy and application prospects compared to traditional methods.
Smart Images

Figure BDA0003054610810000031 
Figure BDA0003054610810000035 
Figure BDA0003054610810000036
Abstract
Description
[0001] This application claims priority to Chinese patent application No. 202010385996.8, filed on May 9, 2020, with invention name “Cancer Prognosis Method”. The contents of the above patent application are hereby cited in their entirety as part of this application. Technical Field
[0002] The present disclosure generally relates to the field of biological detection and diagnosis. Specifically, the present disclosure relates to a method for prognosing a patient based on the methylation difference in the genome of a cancer tissue sample and a paracancerous tissue sample of the patient. The present disclosure also relates to a system and device for prognosing a cancer patient. Technical Background
[0003] Cancer is one of the leading causes of death and disease burden worldwide. Cancer prognosis is a prediction of the possible outcome of an individual's current medical condition, and it is an important tool for improving patient diagnosis and treatment management. Accurate prognosis is crucial for choosing the right cancer treatment method and predicting survival rate. At present, the prognosis of patients is mainly evaluated and predicted based on the clinical stage and pathological analysis of the tumor, assisted by some related molecular characteristics (such as immunohistochemistry, DNA mutations, mRNA or microRNA expression, etc.). However, the above molecular evaluation methods still have great limitations. On the one hand, molecular markers such as 21 genes in breast cancer can accurately predict the risk of postoperative recurrence in patients, bring significant clinical benefits to breast cancer patients, and thus enter the diagnosis and treatment guidelines for breast cancer; on the other hand, molecular markers with high accuracy and clinical application efficacy are still very scarce. Most cancers still require frequent imaging follow-up of patients to monitor the course of the disease in clinical practice, which will cause a certain burden on patients, and there may be a lag in the discovery of tumor progression. Therefore, there is a huge clinical demand for molecular markers that can accurately predict patient prognosis independent of pathological staging and other clinical factors.
[0004] In many studies, it has been proven that DNA methylation variation is closely related to the occurrence of cancer. Compared with gene mutation, DNA methylation variation has the characteristics of wider coverage, higher stability, and earlier occurrence, so it is more suitable for early detection of carcinogenesis. However, methods and strategies for using DNA methylation variation to predict the prognosis of cancer patients are still very lacking. Summary of the invention
[0005] The theory of "regional carcinogenesis" holds that normal tissues will gradually begin to become carcinogenic at the molecular level under the action of a certain mechanism, and this change will first appear in DNA. Based on the theory of "regional carcinogenesis", the inventors have designed a new system that uses DNA methylation detection technology to predict the prognosis and recurrence risk of cancer patients - the malignancy density ratio (MD ratio) assessment system for adjacent tissues.
[0006] The MD ratio evaluation system is based on the following theory: the normal tissues in the patient's body are in the process of transforming from normal cells to cancer cells. By detecting methylation variations, the progress of cancer can be evaluated, thereby predicting the patient's risk of tumor recurrence. Compared with traditional tests, the MD ratio evaluation system only predicts the patient's risk of recurrence by testing tissue samples from the patient, thereby enabling better prognosis management for the patient and avoiding frequent follow-up after surgery. It is simple, efficient, and personalized, and therefore has better application prospects in prognosis management. Moreover, in the analysis of real samples, the system has a higher accuracy in predicting recurrence than mutation detection.
[0007] Accordingly, in one aspect, the present disclosure relates to a method for prognosing a cancer patient, the method comprising:
[0008] a. detecting the methylation level in one or more regions of the genome of the cancer tissue and adjacent tissue from the patient by high-throughput sequencing;
[0009] b. determining the difference in methylation level between the cancer tissue and the adjacent paracancerous tissue; and
[0010] c. using the differential information of the methylation levels to perform prognosis on the patient through a mathematical modeling method,
[0011] The smaller the difference in the methylation level between the cancer tissue and the adjacent adjacent tissue, the worse the prognosis of the patient.
[0012] In some embodiments, the method comprises:
[0013] a. detecting the methylation level in one or more regions of the genome of the cancer tissue and adjacent tissue from the patient by high-throughput sequencing;
[0014] b. determining differentially methylated blocks (DMBs) in the cancer tissue and adjacent paracancerous tissue, comprising:
[0015] b1. For each CpG site in the detected region, the number of methylated reads covering the site is recorded as M, and the number of unmethylated reads is recorded as U;
[0016] b2. Combine adjacent CpG sites and define them as methylation blocks (MBs); add the M of all CpG sites in the i-th MB of the patient, denoted as M i ; All U are added together, denoted as U i ; The total coverage is N i =M i +U i , the methylation level is β i =M i / N i ;
[0017] b3. For the i-th MB, the methylation level in the cancer tissue is recorded as β i (T) =M i (T) / N i (T) , the methylation level in adjacent tissues is recorded as β i (A) =M i (A) / N i (A) , where β i (T) Value and β i (A) MBs with significant differences in values were determined as DMBs;
[0018] c. Prognosing the patient, including:
[0019] c1. Introducing a parameter α indicating the similarity of the methylation levels of the patient's paracancerous tissue and cancer tissue and calculating the patient's α value by the following algorithm:
[0020]
[0021] Where 0<α<1 and the larger the value of α, the higher the similarity. Where f(·), g(·), h(·) represent the conditional distribution of the methylation level of paracancerous tissue, the prior distribution of the methylation level of cancer tissue and the prior distribution of the baseline methylation level, respectively. represents the family of parameters of the prior distribution;
[0022] c2. Use the maximum likelihood estimation algorithm to calculate parameter estimates in the range of 0<α<1 And calculate the variance of parameter estimates through the Fisher information matrix
[0023]
[0024]
[0025] c3. Calculate the p-value of the null hypothesis α=0 for each patient's adjacent tissue.
[0026] Among them, the smaller the p value, the worse the prognosis of the patient.
[0027] In some embodiments, in step b2, combining adjacent CpG sites can be, for example, combining CpG sites with a distance of less than 50 bp, less than 100 bp, less than 150 bp, less than 200 bp, less than 250 bp, less than 300 bp, less than 350 bp, less than 400 bp, less than 450 bp, or less than 500 bp. In some embodiments, in step b2, combining CpG sites with a distance of less than 50 bp or less than 100 bp.
[0028] In some embodiments, in step b3, β i (T) Value and β i (A) Determining a MB with a significant difference in value as a DMB can be, for example, by setting |β i (T) -β i (A) |>σ, and MB is determined as DMB, where σ is a value between 0.05 and 1 and τ is a value greater than 0.1. In some embodiments, σ=0.1 and τ=0.4.
[0029] In some embodiments of the above method, the α value of the patient is calculated in step c1 by the following algorithm:
[0030]
[0031] Among them, l(α) is the log-likelihood function of the observed data, p i (0) ,q i (0) is the Beta-binomial distribution (p i (0) ,q i (0) ) in the shape parameter.
[0032] In some embodiments, one or more regions of the genome are regions of the genome in which methylation variations exist in the population of the cancer patient. For example, it is possible to obtain regions of the genome in which methylation variations are known to exist in the patient population of a specific cancer type through a public database, and to use them in the detection method disclosed herein. For example, methylation chip data of cancer tissues and paracancerous samples in various cancer patient populations can be downloaded from these public databases, and these data can be used to determine the methylation characteristic regions that are closely related to the occurrence of each cancer. It can be determined using, for example, parametric tests (such as t-tests, linear regressions, etc.) and non-parametric tests (such as wilcoxon rank sum tests, etc.). The above-mentioned public databases are, for example, TCGA (The Cancer Genome Atlas) database and GEO (Gene Expression Omnibus) database.
[0033] In some embodiments, one or more regions of the genome cover a region of at least 0.3M (megabases), such as a region of at least 0.3M, 0.4M, at least 0.5M, at least 0.6M, at least 0.7M, at least 0.8M, at least 0.9M, or at least 1.0M.
[0034] In some embodiments, the one or more regions of the genome cover a region of about 0.3M-10.0M of the genome, such as a region of 0.3M-5.0M, 0.3M-4.0M, 0.3M-3.0M, 0.3M-2.0M, 0.3M-1.5M, 0.3M-1.0M, 0.4M-5.0M, 0.4M-4.0M, 0.4M-3.0M, 0.4M-2.0M, The ranges above also include the endpoint values and any subset ranges therebetween.
[0035] In some embodiments, the cancer is a solid tumor. Examples of solid tumors include, but are not limited to, lung cancer (including small cell lung cancer, non-small cell lung cancer, lung adenocarcinoma and lung squamous cell carcinoma), colorectal cancer, liver cancer, ovarian cancer, pancreatic cancer, gallbladder cancer, gastric cancer, esophageal cancer, kidney cancer, melanoma, breast cancer, cervical cancer, endometrial cancer, prostate cancer, bladder cancer, testicular cancer, thyroid cancer, salivary gland cancer, skin cancer, squamous cell carcinoma, neuroblastoma, glioblastoma, retinoblastoma, lymphoma (including Hodgkin's lymphoma and non-Hodgkin's lymphoma), bone cancer, myeloma, basal cell carcinoma, peritoneal cancer, choriocarcinoma, eye cancer, head and neck cancer, laryngeal cancer, oral cancer and rhabdomyosarcoma, etc.
[0036] In some embodiments, the cancer may be selected from lung cancer, colorectal cancer, liver cancer, ovarian cancer, pancreatic cancer, gallbladder cancer, gastric cancer, and esophageal cancer.
[0037] In some embodiments, the cancer is a primary cancer. In other embodiments, the cancer is a secondary or metastatic cancer. The cancer can be at any stage of cancer development, such as early, middle or late stages of cancer development, or the cancer can be at clinical stage I, II, III or IV.
[0038] In some embodiments, the cancer is lung cancer, such as non-small cell lung cancer (NSCLC). In the case where the cancer is lung cancer, the one or more regions of the genome detected may include one or more regions selected from those listed in Table 1.
[0039] Table 1.
[0040]
[0041]
[0042]
[0043]
[0044]
[0045]
[0046]
[0047]
[0048]
[0049] In some embodiments, the cancer is lung cancer, and the one or more regions of the genome detected include at least 100 regions, at least 200 regions, at least 300 regions, at least 400 regions, at least 500 regions, at least 600 regions, at least 700 regions, at least 800 regions, at least 900 regions, or at least 1000 regions selected from the group consisting of those listed in Table 1. In some embodiments, the one or more regions of the genome detected include all regions selected from the group consisting of those listed in Table 1.
[0050] In some embodiments, the cancer patient has undergone prior cancer treatment methods, such as surgery, radiation therapy, chemotherapy, targeted drug therapy, immunotherapy, or a combination thereof.
[0051] In some embodiments of the above methods, the cancer tissue and adjacent paracancerous tissue used may be tissue surgically removed from the patient.
[0052] In some embodiments of the above methods, establishing the baseline methylation level using methylation data from normal tissue (e.g., normal tissue from a healthy donor) comprises:
[0053] For the i-th MB, the methylation data of normal tissue is recorded as M i (0) and N i (0) , it is considered that given N i (0) Under the condition of M i (0) The shape parameter is (p i (0) ,q i (0) )’s Beta-binomial distribution:
[0054] M i (0) |N i (0) ,β i (0) ~Binomial(N i (0) ,β i (0) )
[0055] β i (0) ~Beta(p i (0) ,q i (0) ),
[0056] And use the maximum likelihood algorithm to calculate the parameter pi (0) and q i (0) .
[0057] In some embodiments of the above method, the p-value is calculated in step c3 by a Wald test. In other embodiments, the p-value is calculated by, for example, a likelihood ratio test.
[0058] In some embodiments of the above methods, the method is used to predict the risk of postoperative recurrence and / or survival of the cancer patient. In some embodiments, patients with p<0.05 are identified as having a high risk of recurrence and / or low postoperative survival.
[0059] In a second aspect, the present disclosure relates to a system for prognosing a cancer patient, the system comprising:
[0060] a methylation detection module; and
[0061] Prognostic analysis module,
[0062] The methylation sequencing module is configured to detect the methylation levels in one or more regions of the genome of the cancer tissue and adjacent tissue of the patient by high-throughput sequencing, and the prognosis analysis module is configured to perform prognosis for the patient by the following method:
[0063] a. determining the difference in methylation level between the cancer tissue and the adjacent paracancerous tissue; and
[0064] b. using the differential information of the methylation levels to perform prognosis on the patient through mathematical modeling,
[0065] The smaller the difference in the methylation level between the cancer tissue and the adjacent adjacent tissue, the worse the prognosis of the patient.
[0066] In some embodiments, the system is configured for implementing the prognostic method according to the first aspect of the present disclosure.
[0067] In some embodiments, the prognostic analysis module is configured to perform a prognosis for the patient by:
[0068] a. Determining differentially methylated blocks (DMBs) in the cancer tissue and adjacent paracancerous tissue, comprising:
[0069] a1. For each CpG site in the detected region, the number of methylated reads covering the site is recorded as M, and the number of unmethylated reads is recorded as U;
[0070] a2. Combine adjacent CpG sites and define them as methylation blocks (MBs); add the M of all CpG sites in the i-th MB of the patient, denoted as M i ; All U are added together, denoted as U i ; The total coverage is N i =M i +U i , the methylation level is β i =M i / N i ;
[0071] a3. For the i-th MB, the methylation level in the cancer tissue is recorded as β i (T) =M i (T) / N i (T) , the methylation level in adjacent tissues is recorded as β i (A) =M i (A) / N i (A) , where β i (T) Value and β i (A) MBs with significant differences in values were determined as DMBs;
[0072] b. Prognosing the patient, including:
[0073] b1. Introducing a parameter α indicating the similarity of the methylation levels of the patient's paracancerous tissue and cancer tissue and calculating the patient's α value by the following algorithm:
[0074]
[0075]
[0076] Where 0<α<1 and the larger the value of α is, the higher the similarity is. Where f(·), g(·), h(·) represent the conditional distribution of methylation level of paracancerous tissue, the prior distribution of methylation level of cancer tissue and the prior distribution of baseline methylation level, respectively. represents the family of parameters of the prior distribution;
[0077] b2. Use the maximum likelihood estimation algorithm to calculate parameter estimates in the range of 0<α<1 And calculate the variance of parameter estimates through the Fisher information matrix
[0078]
[0079]
[0080] b3. Calculate the p-value of the null hypothesis α=0 for each patient's adjacent tissue.
[0081] Among them, the smaller the p value, the worse the prognosis of the patient.
[0082] In some embodiments, the prognostic analysis module is configured to combine CpG sites with a distance of less than 50 bp, less than 100 bp, less than 150 bp, less than 200 bp, less than 250 bp, less than 300 bp, less than 350 bp, less than 400 bp, less than 450 bp, or less than 500 bp in step a2. In some embodiments, the prognostic analysis module is configured to combine CpG sites with a distance of less than 50 bp or less than 100 bp in step b2.
[0083] In some embodiments, the prognostic analysis module is configured to convert |β i (T) -β i (A) |>σ, and MB is determined as DMB, where σ is a value between 0.05 and 1 and τ is a value greater than 0.1. In some embodiments, σ=0.1 and τ=0.4.
[0084] In some embodiments, the prognostic analysis module is configured to calculate the patient's α value in step b1 by the following algorithm:
[0085]
[0086]
[0087] Among them, l(α) is the log-likelihood function of the observed data, p i (0) ,q i (0) is the Beta-binomial distribution (p i (0) ,q i (0) ) in the shape parameter.
[0088] In some embodiments, the prognostic analysis module is further configured to establish the baseline methylation level using methylation data from normal tissue, comprising:
[0089] For the i-th MB, the methylation data of normal tissue is recorded as M i(0) and N i (0) , it is considered that given N i (0) Under the condition of M i (0) The shape parameter is (p i (0) ,q i (0) )’s Beta-binomial distribution:
[0090] M i (0) |N i (0) ,β i (0) ~Binomial(N i (0) ,β i (0) )
[0091] β i (0) ~Beta(p i (0) ,q i (0) ),
[0092] And use the maximum likelihood algorithm to calculate the parameter p i (0) and q i (0) .
[0093] In some embodiments of the above system, the system is further configured to calculate the p-value in step b3 by Wald test. In other embodiments, the p-value is calculated by, for example, a likelihood ratio test.
[0094] In some embodiments of the above system, the system is further configured to predict the risk of postoperative recurrence and / or survival of the cancer patient. In some embodiments, patients with p<0.05 are identified as having a high risk of recurrence and / or low postoperative survival.
[0095] In a third aspect, the present disclosure relates to an apparatus for prognosing a cancer patient, comprising:
[0096] memory for storing computer program instructions; and
[0097] a processor for executing computer program instructions,
[0098] Wherein when the computer program instructions are executed by the processor, the device performs the method according to the first aspect of the present disclosure.
[0099] In a fourth aspect, the present disclosure relates to a computer-readable medium storing computer program instructions, wherein when the computer program instructions are executed by a processor, the method according to the first aspect of the present disclosure is implemented. BRIEF DESCRIPTION OF THE DRAWINGS
[0100] Figure 1 A line graph showing the disease-free survival (DFS) results of patients with high and low risk of recurrence predicted by detecting the genomic regions listed in Table 1 in the patient's cancer tissue and adjacent adjacent tissue using the MD ratio method.
[0101] Figure 2 A line graph showing the disease-free survival (DFS) results of patients with high and low risk of recurrence predicted by detecting the genomic regions listed in Table 3 in the patient's cancer tissue and adjacent adjacent tissue using the MD ratio method.
[0102] Figure 3 A line graph showing the disease-free survival (DFS) results of patients with high and low risk of recurrence predicted by detecting the genomic regions listed in Table 4 in the patient's cancer tissue and adjacent adjacent tissue using the MD ratio method. DETAILED DESCRIPTION
[0103] The present invention will be further described below in conjunction with specific embodiments, and the advantages and features of the present invention will become clearer as the description proceeds. However, these embodiments are exemplary only and do not constitute any limitation to the scope of the present invention. It should be understood by those skilled in the art that the details and forms of the technical solution of the present invention may be modified or replaced without departing from the spirit and scope of the present invention, but these modifications and replacements all fall within the scope of protection of the present invention.
[0104] Data preparation
[0105] The main strategy of the MD ratio evaluation system is to sequence the methylation levels of tissue samples to find regions in the genome where the methylation status of cancerous tissues and normal tissues (such as paracancerous tissues) in the patient's body is different. Based on the fact that paracancerous tissues are in the process of transforming from normal cells to cancer cells, the degree of similarity between the methylation level of paracancerous tissues and the methylation level of cancerous tissues in the differential regions is calculated, and then the risk level of canceration of normal cells in the patient's body is inferred.
[0106] The cancer recurrence risk MD ratio assessment system detects the patient's cancer tissue and adjacent tissue samples. Taking lung cancer as an example, in lobectomy and segmentectomy, adjacent tissue is defined as tissue 5 cm outside the resection margin; in wedge resection, adjacent tissue is defined as tissue 3 cm outside the resection margin, and is verified to contain no tumor cells through pathological evaluation of tissue cells; at the same time, the cell types of cancer tissue and adjacent tissue are consistent. In addition, 68 normal tissue samples from healthy donors (age distribution 30-86 years; median age 57.5 years; 35 males and 33 females) were used for the baseline construction process. The above samples were also verified to contain no tumor cells through pathological evaluation and were consistent with the cell types of cancer tissue.
[0107] The sample library was prepared using the brELSATM method (Burning Rock Biotech, Guangzhou, China), which includes the following steps: 1) DNA extraction and purification; 2) sodium bisulfite treatment; 3) single-stranded DNA amplification by DNA polymerase; 4) using customized cancer methylation profile RNA bait to enrich the target region shown in Table 1 (covering approximately 0.9M of the human genome); 5) quantification of the target library by real-time PCR. Finally, sequencing was performed using the NovaSeq 6000 sequencer released by Illumina, with an average sequencing depth of 1000 layers.
[0108] The original output file of sequencing is analyzed using the sequence alignment software BWA-meth and the methylation data statistics software MethylDackel to obtain the methylation detection output file for each sample. It contains the location information of each CpG site in the specific capture area, as well as the methylation information in the read segment covering this site. The number of methylated reads covering this site is recorded as M, and the number of unmethylated reads is recorded as U. Adjacent CpG sites (distance less than 50bp) are combined, and this collection of multiple CpG sites is called a methylation block (MB). Add the M at all sites in the patient's i-th MB and record it as M i , the sum of all U is denoted as U i , the total number of covers is recorded as N i (N i =M i +U i ), the methylation level is recorded as β i (β i Moment estimate
[0109] Differential methylation region screening, observational data modeling and parameter solution
[0110] In order to detect the extent to which the patient's paracancerous tissue is similar to the cancer tissue but different from the normal tissue, the research target was locked on the regions with methylation differences in the genomes of cancer tissue and paracancerous tissue, which were defined as differentially methylated blocks (DMBs).
[0111] To screen for DMB in individual cancer patients, the methylation levels of cancer tissue samples from patients were used. (T) Compared with the methylation level of adjacent tissue samples (A) For the i-th MB, the methylation level in the cancer tissue is recorded as β i (T) =M i (T) / N i (T) , the methylation level in adjacent tissues is recorded as β i (A) =M i (A) / N i (A) , will meet |β i (T) -β i (A) |>0.1, and The MB is determined as the personalized DMB for the patient.
[0112] In order to establish the corresponding statistical model, the normal tissue samples (as described in Example 1) are first used to establish the baseline of methylation levels. For the i-th MB, the methylation data of the normal tissue is recorded as M i (0) and N i (0) , it is considered that given N i (0) Under the condition of M i (0) The shape parameter is (p i (0) ,q i (0) )’s Beta-binomial distribution:
[0113] M i (0) |N i (0) ,β i (0) ~Binomial(N i (0) ,β i (0) )
[0114] β i (0) ~Beta(p i (0) ,q i (0) ),
[0115] where β i (0) The maximum likelihood estimation (MLE) algorithm is used to solve the shape parameter {p i ,q i}, and the maximum likelihood estimate of the parameter is recorded as {p i (0) ,q i (0)}.
[0116] For cancer patients, the methylation sequencing data of their adjacent tissue samples (M i (A) , N i (A) , β i (A) ) obeys a mixed Beta-Binomial distribution, which can be expressed as:
[0117] M i (A) |N i (A) , β i (A) ~Binomial(N i (A) , β i (A) )
[0118] β i (A) i =αβ i (T) +(1-α)β i (0)
[0119] β i (0) ~Beta(p i (0) ,q i (0) )
[0120] where β i (T) Indicates the methylation level of the patient's cancer tissue, which can be estimated using the moment of the cancer tissue sample sequencing data Replace; β i (0) represents the baseline methylation level, which follows the shape parameter (p i (0) ,q i (0) ) is a Beta distribution. α is a proportional parameter on [0, 1], which indicates the similarity between paracancerous tissue and cancer tissue. The closer α is to 1, the closer the methylation level of paracancerous tissue is to cancer tissue, and it can be inferred that the patient's risk of recurrence is higher.
[0121] Based on the above distribution, the log-likelihood function of the observed data can be written as:
[0122]
[0123]
[0124] The parameter α satisfies:
[0125] 0<α<1
[0126] The parameter α is calculated by the maximum likelihood algorithm, and the parameter estimate is calculated using the quasi-Newton algorithm in the range of 0<α<1 And calculate the variance of parameter estimates through the Fisher information matrix
[0127]
[0128]
[0129] The risk of cancer recurrence is inferred by the similarity between adjacent tissue and cancer tissue: Statistical inference is made on the null hypothesis of parameter α: α = 0. Wald test is used to infer the null hypothesis:
[0130]
[0131] Under the chi-square distribution with a degree of freedom of 1, the p-value of α=0 for each patient's paracancerous tissue can be calculated. The smaller the p-value, the more similar the patient's paracancerous tissue is to the cancer tissue, and the greater the risk of cancer recurrence in the future.
[0132] Numerical simulation results of MD ratio for cancer recurrence risk
[0133] Apply the beta-binomial mixture model to generate simulated data with the parameter set to N i =1000, m i =1000, p i (0) =11,q i(0) =383, where p i (0) and q i (0) The value of is derived from the maximum likelihood value of normal tissue samples. α is taken as 0, 0.001, 0.003, 0.01, 0.003, 0.1, and each set of parameters is repeated 50 times. The numerical simulation results are shown in Table 2.
[0134] Table 2.
[0135]
[0136] Among them, Bias means The mean of The mean value, Std represents The standard deviation of the data, Wald represents the median of the Wald test, and Power represents the probability of p-value < 0.05. From the table, we can see that the maximum likelihood algorithm error is very small; the false positive rate = 0.06, and the test power = 1, which is in line with expectations.
[0137] Example 2. Real sample test results of cancer recurrence risk MD ratio
[0138] The MD ratio evaluation system described above was used to test cancer tissue samples and adjacent tissue samples from 39 patients with stage IA non-small cell lung cancer (age distribution 40-82 years; median age 61 years; 24 males and 15 females). The methylation levels of the genomic regions listed in Table 1 were detected in cancer tissues and adjacent tissues. The p-value of the patient was calculated using the above algorithm. Based on the test p-value < 0.05, the test results were divided into two groups: high risk of recurrence and low risk of recurrence, and their disease-free survival (DFS) results are as follows: Figure 1 shown.
[0139] from Figure 1 The results show that high-risk patients relapse faster than low-risk patients, and the difference in DFS between the two groups of patients is statistically significant (p-value = 0.039). The above results prove that the MD ratio evaluation system can be used to accurately predict the risk of recurrence and subsequent survival of cancer patients.
[0140] In order to further verify the effect of MD ratio evaluation system on the prognosis of cancer patients, the methylation levels of the genomic regions listed in Tables 4 and 5 (which include 522 and 532 regions randomly selected from Table 1, covering about 0.47M and about 0.48M regions in the genome, respectively) in cancer tissues and paracancerous tissues were detected. The p-value of the patient was calculated by the same algorithm, and the test results were divided into high risk of recurrence and low risk of recurrence according to the test p-value < 0.05, and the disease-free survival (DFS) results were as follows: Figure 2 and Figure 3 shown.
[0141] Table 3.
[0142]
[0143]
[0144]
[0145]
[0146]
[0147] Table 4.
[0148]
[0149]
[0150]
[0151]
[0152]
[0153] In order to compare the effectiveness of the MD ratio evaluation system with traditional tumor driver gene detection in the prognosis of cancer patients, the correlation between the main mutation types EGFR 19del and EGFR L858R and DFS was compared with the correlation between the MD ratio evaluation system detection results and DFS using the somatic mutation detection method. The results are shown in Table 5. Among them, MD ratio-1 represents the result obtained by detecting the area shown in Table 1 (corresponding to Figure 1 ); MD ratio-2 and MD-ratio-3 respectively represent the results obtained by detecting the areas shown in Table 3 and Table 4 (corresponding to Figure 2 and Figure 3 ).
[0154] Table 5.
[0155] p-value Harzard ratio(HR) EGFR L858R 0.230 0.464 EGFR 19del 0.130 2.442 MD ratio-1 0.039 4.692 MD ratio-2 0.030 3.401 MD ratio-3 0.018 4.453
[0156] The above results indicate that the MD ratio assessment system can more effectively assess patients' risk of cancer recurrence and subsequent survival than somatic testing, and can play a better role in patients' prognosis management and clinical treatment.
Claims
1. Use of a methylation detection module in preparing a system for prognosing a patient with non-small cell lung cancer, wherein the methylation detection module is configured to detect the methylation levels in at least 500 regions selected from Table 1 of the genome of the cancer tissue and adjacent tissue of the patient by high-throughput sequencing, wherein the regions cover the 0.4M-0.5M region of the genome, and the system is configured to perform the following method: a. detecting the methylation level in one or more regions of the genome of the cancer tissue and adjacent tissue from the patient by high-throughput sequencing; b. determining the difference in methylation level between the cancer tissue and the adjacent paracancerous tissue; and c. using the differential information of the methylation levels to perform prognosis on the patient through a mathematical modeling method, wherein the smaller the difference in methylation level between the cancer tissue and the adjacent adjacent tissue, the worse the prognosis of the patient; The method comprises: a. detecting the methylation level in one or more regions of the genome of the cancer tissue and adjacent tissue from the patient by high-throughput sequencing; b. determining differentially methylated blocks (DMBs) in the cancer tissue and adjacent paracancerous tissue, comprising: b1. For each CpG site in the detected region, the number of methylated reads covering the site is recorded as M, and the number of unmethylated reads is recorded as U; b2. Combine adjacent CpG sites and define them as methylation blocks (MBs); add the M of all CpG sites in the i-th MB of the patient, denoted as M i ; All U are added together, denoted as U i ; The total coverage is N i =M i +U i , the methylation level is β i =M i / N i ; b3. For the i-th MB, the methylation level in the cancer tissue is recorded as β i (T) =M i (T) / N i (T) , the methylation level in adjacent tissues is recorded as β i (A) =M i (A) / N i (A) , where β i (T) Value and β i (A) MBs with significant differences in values were determined as DMBs; c. Prognosing the patient, including: c1. Introducing a parameter α indicating the similarity of the methylation levels of the patient's paracancerous tissue and cancer tissue and calculating the patient's α value by the following algorithm: Where 0<α<1 and the larger the value of α, the higher the similarity. Where f(·), g(·), h(·) represent the conditional distribution of the methylation level of paracancerous tissue, the prior distribution of the methylation level of cancer tissue and the prior distribution of the baseline methylation level, respectively. represents the family of parameters of the prior distribution; c2. Use the maximum likelihood estimation algorithm to calculate parameter estimates in the range of 0<α<1 And calculate the variance of parameter estimates through the Fisher information matrix c3. Calculate the p-value of the null hypothesis α=0 for each patient's adjacent tissue. Among them, the smaller the p value, the worse the prognosis of the patient.
2. The method according to claim 1, wherein the method optionally comprises any one or more of the following features: (1) in step b2, combining CpG sites with a distance of less than 50 bp or a distance of less than 100 bp; (2) In step b3, |β i (T) -β i (A) |>σ, and MB is determined as DMB, where σ is a value between 0.05 and 1 and τ is a value greater than 0.1; (3) In step c1, the α value of the patient is calculated by the following algorithm: l(α)=∑ ilogL(α;M i (A) ,N i (A) ,p i (0) ,q i (0) ,β i (T) ) in, l(α) is the log-likelihood function of the observed data, p i (0) ,q i (0) is the Beta-binomial distribution (p i (0) ,q i (0) ) in shape parameters; (4) calculating the p-value by Wald test in step c3; (5) Patients with p<0.05 were identified as having a high risk of recurrence and / or low postoperative survival.
3. The use according to claim 2, wherein in (2), σ=0.1 and τ=0.
4.
4. The method according to any one of claims 1 to 3, wherein the method comprises any one or more of the following features: (1) the one or more regions of the genome are regions of the genome where methylation variation exists in the population of cancer patients; (3) The method is used to predict the risk of postoperative recurrence and / or survival of the cancer patient.
5. The use according to any one of claims 1 to 3, wherein: The cancer tissue and adjacent paracancerous tissue are tissues surgically removed from the patient.
6. The method of any one of claims 1 to 3, wherein the method uses methylation data from normal tissue to establish the baseline methylation level, comprising: For the i-th MB, the methylation data of normal tissue is recorded as M i (0) and N i (0) , it is considered that given N i (0) Under the condition of M i (0) The shape parameter is (p i (0) ,q i (0) )’s Beta-binomial distribution: M0 (0) |N i (0) ,b i (0) ~Binomial(N i (0) ,b i (0) ) β i (0) ~Beta(p i (0) ,q i (0) ), And use the maximum likelihood algorithm to calculate the parameter p i (0) and q i (0) .
7. A system for prognosing a non-small cell lung cancer patient, the system comprising: Methylation detection module; and Prognostic analysis module, The methylation sequencing module is configured to detect the methylation levels in at least 500 regions selected from Table 1 of the genome of the cancer tissue and adjacent tissue of the patient by high-throughput sequencing, wherein the regions cover the 0.4M-0.5M region of the genome, and the prognosis analysis module is configured to perform prognosis for the patient by the following method: a. determining the difference in methylation level between the cancer tissue and the adjacent paracancerous tissue; and b. using the differential information of the methylation levels to perform prognosis on the patient through mathematical modeling, wherein the smaller the difference in methylation level between the cancer tissue and the adjacent adjacent tissue, the worse the prognosis of the patient; The prognostic analysis module is configured to perform prognosis on the patient by the following method: a. Determining differentially methylated blocks (DMBs) in the cancer tissue and adjacent paracancerous tissue, comprising: a1. For each CpG site in the detected region, the number of methylated reads covering the site is recorded as M, and the number of unmethylated reads is recorded as U; a2. Combine adjacent CpG sites and define them as methylation blocks (MBs); add the M of all CpG sites in the i-th MB of the patient, denoted as M i ; All U are added together, denoted as U i ; The total coverage is N i =M i +U i , the methylation level is β i =M i / N i ; a3. For the i-th MB, the methylation level in the cancer tissue is recorded as β i (T) =M i (T) / N i (T) , the methylation level in adjacent tissues is recorded as β i (A) =M i (A) / N i (A) , where β i (T) Value and β i (A) MBs with significant differences in values were determined as DMBs; b. Prognosing the patient, including: b1. Introducing a parameter α indicating the similarity of the methylation levels of the patient's paracancerous tissue and cancer tissue and calculating the patient's α value by the following algorithm: Where 0<α<1 and the larger the value of α, the higher the similarity. Where f(·), g(·), h(·) represent the conditional distribution of the methylation level of paracancerous tissue, the prior distribution of the methylation level of cancer tissue and the prior distribution of the baseline methylation level, respectively. represents the family of parameters of the prior distribution; b2. Use the maximum likelihood estimation algorithm to calculate parameter estimates in the range of 0<α<1 And calculate the variance of parameter estimates through the Fisher information matrix b3. Calculate the p-value of the null hypothesis α=0 for each patient's adjacent tissue. Among them, the smaller the p value, the worse the prognosis of the patient.
8. The system of claim 7, wherein the prognostic analysis module is configured to combine CpG sites with a distance less than 50 bp or a distance less than 100 bp in step a2.
9. The system of claim 7, wherein the prognostic analysis module is configured to convert |β i (T) -β i (A) |>σ, and The MB is determined as DMB, where σ is a value between 0.05 and 1 and τ is a value greater than 0.
1.
10. The system of claim 9, wherein σ=0.1 and τ=0.
4.
11. The system of claim 7, wherein the prognostic analysis module is configured to calculate the α value of the patient in step b1 by the following algorithm: l(α)=ΣlogL(α;M i (A) ,N i (A) ,p i (0) ,q i (0) ,β i (T) ) in, l(α) is the log-likelihood function of the observed data, p i (0) ,q i (0) is the Beta-binomial distribution (p i (0) ,q i (0) ) in the shape parameter.
12. The system of claim 7, wherein the prognostic analysis module is further configured to establish the baseline methylation level using methylation data from normal tissue, comprising: For the i-th MB, the methylation data of normal tissue is recorded as M i (0) and N i (0) , it is considered that given N i (0) Under the condition of M i (0) The shape parameter is (p i (0) ,q i (0) )’s Beta-binomial distribution: M i (0) |N i (0) ,b i (0) ~Binomial(N i (0) ,b i (0) ) β i (0) ~Beta(p i (0) ,q i (0) ), And use the maximum likelihood algorithm to calculate the parameter p i (0) and q i (0) .
13. The system of claim 7, wherein the prognostic analysis module is configured to calculate the p-value by Wald test in step b3.
14. The system according to any one of claims 7 to 13, wherein the system is used to predict the risk of postoperative recurrence and / or survival of the cancer patient.
15. The system of any one of claims 7-13, wherein patients with p<0.05 are identified as having a high risk of recurrence and / or low postoperative survival.
16. A device for prognosing a cancer patient, comprising: memory for storing computer program instructions; and a processor for executing computer program instructions, Wherein when the computer program instructions are executed by the processor, the apparatus performs the method defined in any one of claims 1-6.
17. A computer-readable medium storing computer program instructions, wherein when the computer program instructions are executed by a processor, the method defined in any one of claims 1 to 6 is implemented.
