Colorectal cancer early screening method based on multi-omics sequencing

Abstract

The invention belongs to the technical field of molecular biomedicine, and provides a colorectal cancer early screening method based on multi-omics sequencing, which comprises the following steps: collecting a plasma sample to be detected, collecting sequencing data to be detected, and classifying a colorectal tumor early screening model. The training process of the colorectal tumor early screening model comprises the steps of original sequencing data acquisition, quality control and human reference genome comparison, reference methylation data acquisition, tumor specific region screening, plasma cfDNA sequencing data acquisition, coordinate determination sequencing data acquisition, methylation feature set screening and terminal motif feature screening. And carrying out breakpoint motif feature screening, feature merging and model training. By constructing the colorectal tumor early screening model, the non-invasive colorectal cancer detection method with high cost effectiveness is realized, the economic cost of sequencing is reduced, and the accuracy and sensitivity of detection are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of molecular biomedical technology, and in particular to an early screening method for colorectal cancer based on multi-omics sequencing. Background Art

[0002] Colorectal cancer is one of the most common malignant tumors in the world, with high morbidity and mortality. According to the World Health Organization, colorectal cancer causes millions of deaths each year. Since early colorectal cancer usually lacks obvious symptoms, many patients are already in the advanced stage when diagnosed, which greatly affects the effectiveness of treatment. Therefore, early detection and intervention of colorectal cancer is crucial to improve the survival rate of patients.

[0003] Traditional colorectal cancer screening methods mainly include fecal occult blood test and colonoscopy. Although the fecal occult blood test is easy to operate, its sensitivity and specificity are relatively low, and it is easy to miss or misdiagnose. Although colonoscopy is more accurate, it is less accepted by patients due to its invasiveness and is more expensive. With the development of medical imaging technology, such as CT colonography and magnetic resonance colonography, non-invasive means of colorectal cancer detection are provided. However, these technologies also have problems such as high cost, high equipment requirements, and radiation exposure to patients. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide an early screening method for colorectal cancer based on multi-omics sequencing. By constructing an early screening model for colorectal tumors, a non-invasive and cost-effective colorectal cancer detection method is achieved, the economic cost of sequencing is reduced, and the accuracy and sensitivity of detection are improved.

[0005] To achieve the above object, the present invention provides the following solutions:

[0006] An early screening method for colorectal cancer based on multi-omics sequencing, including:

[0007] Collecting plasma samples to be tested;

[0008] Performing DNA methylation sequencing on the plasma sample to be tested to obtain sequencing data to be tested;

[0009] The sequencing data to be tested is input into the trained colorectal tumor early screening model for classification to obtain the target screening results.

[0010] The training process of the colorectal tumor early screening model includes:

[0011] Collecting colorectal cancer tissue and adjacent normal tissue samples from colorectal cancer patients, extracting genomic DNA and performing DNA methylation sequencing on the colorectal cancer tissue and the adjacent normal tissue samples to obtain original sequencing data;

[0012] The raw sequencing data is sequentially subjected to quality control and compared with the human reference genome to obtain sequence correspondence;

[0013] Dividing the human reference genome into a plurality of CpG island regions, and analyzing the methylation levels of different haploids in each of the CpG island regions to obtain reference methylation data;

[0014] According to the sequence correspondence, the reference methylation data and the original sequencing data are used to perform a P value corrected by the Benjamin and Hochberg multiple test to analyze the difference in methylation levels, and obtain tumor-specific regions;

[0015] Collecting plasma samples from colorectal cancer patients and healthy subjects, and performing cfDNA extraction and methylation sequencing on the plasma samples to obtain plasma cfDNA sequencing data;

[0016] The plasma cfDNA sequencing data is sequentially subjected to quality control and human reference genome alignment to obtain coordinate-determined sequencing data;

[0017] Counting the coordinates to determine the methylation degree of the sequencing data in the tumor-specific region to obtain a methylation feature set;

[0018] Obtaining the coordinates to determine the coordinates of all read segment data of the sequencing data on the reference genome, taking the p bases at the 3' end of the read segment data on the human reference genome as the first base fragment, and calculating the proportion of the first base fragment to obtain the terminal motif feature;

[0019] The q bases upstream and q bases downstream of the 3' end of the read data are taken as the second base fragment, and the proportion of the second base fragment is calculated to obtain the breakpoint motif feature;

[0020] Combining the methylation feature set, the terminal motif feature, and the breakpoint motif feature into an initial feature value;

[0021] The initial characteristic value is used as input data, and the disease condition is used as the output result, and the pre-designed integrated machine learning classifier is trained to obtain the colorectal tumor early screening model.

[0022] Preferably, the haploid includes any one or more of MM, MHL, CHALM, PDR and Entropy.

[0023] Preferably, p is any integer between 4 and 10; and q is any integer between 2 and 5.

[0024] Preferably, the integrated machine learning classifier includes: a first layer and a second layer; the first layer includes: a support vector machine, a random forest, XGBoost and CatBoost; the second layer includes: a logistic regression network.

[0025] The present invention discloses the following technical effects:

[0026] The present invention provides an early screening method for colorectal cancer based on multi-omics sequencing. By constructing an early screening model for colorectal tumors, it solves the defects of the existing technology such as low accuracy, high cost, and radiation, and realizes a non-invasive and cost-effective colorectal cancer detection method. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0028] Figure 1 A schematic diagram of the process of early screening for colorectal cancer based on multi-omics sequencing provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] The purpose of the present invention is to provide an early screening method for colorectal cancer based on multi-omics sequencing. By constructing an early screening model for colorectal tumors, a non-invasive and cost-effective colorectal cancer detection method is achieved, the economic cost of sequencing is reduced, and the accuracy and sensitivity of detection are improved.

[0031] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0032] Figure 1 The schematic diagram of the early screening process of colorectal cancer based on multi-omics sequencing provided in this embodiment is as follows: Figure 1As shown, this embodiment provides a method for early screening of colorectal cancer based on multi-omics sequencing, comprising:

[0033] Step 100: Collecting a plasma sample to be tested;

[0034] Step 200: performing DNA methylation sequencing on the plasma sample to be tested to obtain sequencing data to be tested;

[0035] Step 300: Input the sequencing data to be tested into the trained colorectal cancer early screening model for classification to obtain the target screening results.

[0036] The training process of the colorectal cancer early screening model includes:

[0037] Collect colorectal cancer tissue and adjacent normal tissue samples from colorectal cancer patients, extract genomic DNA and perform DNA methylation sequencing on the colorectal cancer tissue and adjacent normal tissue samples to obtain original sequencing data;

[0038] The original sequencing data is quality controlled and aligned to the human reference genome to obtain the sequence correspondence;

[0039] The human reference genome is divided into several CpG island regions, and the methylation levels of different haploids in each CpG island region are analyzed to obtain reference methylation data;

[0040] According to the sequence correspondence, the P value after Benjaminiand-Hochberg multiple test correction of the reference methylation data and the original sequencing data was used to analyze the difference in methylation degree and obtain tumor-specific regions;

[0041] Collect plasma samples from colorectal cancer patients and healthy people, extract cfDNA and perform methylation sequencing on the plasma samples to obtain plasma cfDNA sequencing data;

[0042] The plasma cfDNA sequencing data were sequentially subjected to quality control and human reference genome alignment to obtain coordinate-determined sequencing data;

[0043] Statistical coordinates are used to determine the methylation degree of sequencing data in tumor-specific regions to obtain a methylation feature set;

[0044] Obtain coordinates to determine the coordinates of all read data of the sequencing data on the reference genome, take the p bases at the 3' end of the read data on the human reference genome as the first base fragment, and calculate the proportion of the first base fragment to obtain the terminal motif feature;

[0045] The q bases upstream and q bases downstream of the 3' end of the read data in the human reference genome are taken as the second base fragment, and the proportion of the second base fragment is calculated to obtain the breakpoint motif feature;

[0046] The methylation feature set, the terminal motif feature, and the breakpoint motif feature are combined into an initial feature value;

[0047] The initial feature values ​​were used as input data and the disease conditions were used as output results. The pre-designed integrated machine learning classifier was trained to obtain an early screening model for colorectal cancer.

[0048] Specifically, healthy human plasma samples were used as negative controls for training the model and were processed in the same way as colorectal cancer plasma samples.

[0049] Optionally, the haploid includes any one or more of MM, MHL, CHALM, PDR and Entropy.

[0050] Preferably, p is any integer between 4 and 10; q is any integer between 2 and 5.

[0051] Specifically, the integrated machine learning classifier includes: a first layer and a second layer; the first layer includes: support vector machine, random forest, XGBoost and CatBoost; the second layer includes: logistic regression network.

[0052] Furthermore, a device for early detection of colorectal cancer comprises the following components:

[0053] Tissue sample sequencing module: processes collected cancer tissues and adjacent normal tissue samples, performs genomic DNA extraction, library construction, and methylation sequencing to generate raw sequencing output data;

[0054] Blood sample sequencing module: processes the collected blood samples, extracts cfDNA, constructs libraries and performs methylation sequencing, and also generates raw sequencing data;

[0055] Methylation feature set acquisition module: performs quality control on the original sequencing data and compares it with the human reference genome to identify the methylation status of cancer-specific regions under different haplotypes and form a methylation feature set;

[0056] Terminal motif feature acquisition module: Analyze the p bases at the 3' end of each sequencing fragment to form a base fragment set, and calculate the distribution ratio of these fragments in the entire sample to determine the terminal motif features;

[0057] Breakpoint motif feature acquisition module: Analyze the q base pairs upstream and downstream of the 3' end of each sequencing fragment to form a base fragment set, and calculate the distribution ratio of these fragments to determine the breakpoint motif features;

[0058] Model building module: The methylation feature set, terminal motif features and breakpoint motif features extracted above are used as input to the first layer of a composite machine learning classification framework, and the cancer diagnosis results are used as the output of the second layer for model training to build an early screening model.

[0059] Optionally, a computer-readable medium stores a software program capable of executing the above-mentioned method for constructing an early screening model for colorectal cancer.

[0060] Specifically, the data sets used in the model construction process are shown in Table 1.

[0061] Table 1

[0062]

[0063]

[0064] Preferably, the tissue sample extraction and sequencing method:

[0065] Tissue samples were collected from the subjects, and genomic DNA was extracted using the genomic DNA extraction kit provided by TIANGEN according to the product instructions. The extracted genomic DNA was subjected to library construction, and RRBS sequencing technology was performed after completion.

[0066] After sequencing is completed, the raw data is quality controlled, high-quality sequencing data is screened out, and compared with the human reference genome for further analysis.

[0067] Further, the extraction and sequencing methods of plasma cfDNA samples:

[0068] A 10ml whole blood sample was collected from the subject using a free DNA collection tube. After separating the plasma, cfDNA was extracted using the plasma DNA extraction kit provided by GENFINE according to the product instructions. The extracted cfDNA library was constructed using the aforementioned patented technology, and RRBS sequencing was performed after the construction was completed. After sequencing was completed, the original data was quality controlled, high-quality sequencing data was screened, and compared with the human reference genome for in-depth analysis.

[0069] Specifically, data processing: The early screening method of this embodiment covers the construction of multiple feature sets, including haploid methylation features, terminal motif features, and breakpoint motif features:

[0070] 1) Identification of haploid methylation feature sets:

[0071] This example identifies multiple haploid states, including MM, MHL, CHALM, PDR, and Entropy, each of which reflects the degree of methylation in the CpG island region through a specific algorithm. By comparing the haploid methylation levels of cancer tissue and adjacent tissue samples, significant differences in the CpG island region are revealed, which can be used as a set of methylation features to distinguish cancer.

[0072] 2) Analysis of terminal motif characteristics:

[0073] In tumor tissue, tumor-specific DNA fragments are generated due to changes in chromatin state and abnormal nuclease activity. Differences in the breakpoint positions of these fragments lead to changes in the proportion of terminal motif sequences. By extracting the 6-base sequence at the 3' end of each read segment in the cfDNA high-throughput sequencing data, we can calculate the distribution characteristics of the terminal motif. Considering the permutations and combinations of the four bases (A, T, C, G), this embodiment can identify up to 4096 different terminal motif distributions.

[0074] 3) Breakpoint motif feature analysis:

[0075] Similar to the terminal motif feature, this feature focuses on the adjacent sequence of the break position, that is, the sequence of 3 base pairs upstream and downstream of the 3' end of each read segment. This sequence selection reveals a different feature distribution from the terminal motif. Considering the permutations and combinations of the four bases (A, T, C, G), this embodiment can identify up to 4096 different breakpoint motif distributions.

[0076] Furthermore, based on feature extraction and training set samples, this embodiment further constructs an integrated machine learning classification model, which consists of two layers: the first layer integrates four different machine learning algorithms, each of which is independently trained on the training set samples to identify the probability of cancer. The second layer integrates the output of the first layer through a logistic regression model to generate the final classification decision. The following are the algorithms involved and their brief descriptions:

[0077] 1) Logistic regression model:

[0078] Logistic regression is a widely used supervised learning model suitable for handling classification problems. It is known for its computational efficiency and intuitiveness by establishing a cost function and iteratively solving the optimal model parameters through optimization methods. This method is suitable for binary classification problems, fast and easy to understand.

[0079] 2) Support Vector Machine:

[0080] Support vector machine is a powerful supervised learning model for classification and regression analysis. It maps data into a high-dimensional space and finds an optimal hyperplane to maximize the interval between different categories. When training the model, the following parameters are set: kernel = "linear", scale = T, probability = TRUE, type = "C-classification"

[0081] 3) Random Forest:

[0082] Random forest is an ensemble learning method that improves prediction accuracy by building multiple decision trees and voting or averaging their predictions. When training the model, the following parameters were set: method = "rf", prox = TRUE, ntree = 500, metric = "Accuracy".

[0083] 4) XGBoost:

[0084] XGBoost is an efficient gradient boosting decision tree framework that improves the performance of base learners through optimization algorithms. It is particularly suitable for processing large-scale data sets. When training the model, the following parameters are set: nrounds = 1000, params = list (booster = "gbtree", objective = "binary:logistic", eval_metric = "logloss")

[0085] 5) CatBoost:

[0086] CatBoost is an advanced gradient boosting decision tree algorithm, specially designed to process categorical features and reduce model bias, improve prediction accuracy through iterative optimization, and is particularly suitable for processing categorical features. When training the model, the following parameters are set: params = list (loss_function = 'Logloss', iterations = 1000)

[0087] Specifically, the process of establishing a colorectal cancer early screening detection model: This embodiment uses a randomization method to divide plasma samples into two parts: a training set and a validation set, with a ratio of 1:1. After quality control, 1863 samples were randomly divided into a training set (931 in number) and a test set (932 in number). The training set is used for model training, and the test set is used for model evaluation. The training set is used to build the model, and the validation set is used to evaluate the performance of the model. This embodiment accurately identifies CpG island regions with specific methylation patterns through in-depth analysis of tumor tissue and corresponding normal tissue samples. These patterns include MM, MHL, CHALM, PDR and Entropy haploid states. Furthermore, this embodiment combines the haploid methylation characteristics in the blood, the motif characteristics at the end of the sequence, and the motif characteristics near the breakpoints to form a comprehensive set of biomarkers.

[0088] Furthermore, in the feature screening stage, this embodiment implemented a P value screening strategy based on Benjaminiand-Hochberg multiple test correction, with a P value less than 0.01 as the threshold to ensure the statistical significance of the screening results. The MM, MHL, Entropy, CHALM, PDR, BPM_raw_4bp and mer6_raw_4bp features screened in this embodiment are as follows:

[0089] Characteristics of MM, a total of 251 CpG islands, see Table 2:

[0090] Table 2

[0091]

[0092]

[0093]

[0094]

[0095] The characteristics of MHL, a total of 255 CpG islands, are shown in Table 3:

[0096] Table 3

[0097]

[0098]

[0099]

[0100]

[0101] Entropy features, a total of 182 CpG islands, see Table 4:

[0102] Table 4

[0103]

[0104]

[0105]

[0106]

[0107] The characteristics of CHALM, a total of 214 CpG islands, are shown in Table 5:

[0108] Table 5

[0109]

[0110]

[0111]

[0112] The characteristics of PDR, a total of 146 CpG islands, are shown in Table 6:

[0113] Table 6

[0114]

[0115]

[0116]

[0117] The features of BPM_raw_4bp total 321 motifs, see Table 7:

[0118] Table 7

[0119]

[0120]

[0121]

[0122] The features of mer6_raw_4bp total 271 motifs, see Table 8:

[0123] Table 8

[0124]

[0125]

[0126] Specifically, after completing the screening and integration of features, this embodiment constructs an early screening model for colorectal cancer containing seven features, which integrate the above-mentioned five haploid states, terminal motifs and breakpoint motif features. Using the data in the training set, this embodiment inputs the samples into four different machine learning algorithms, each of which uses the probability of cancer occurrence as the prediction output, forming 28 basic models, which constitute the first layer of the model. Subsequently, the prediction results of these basic models are further input into the logistic regression model of the second layer to integrate the predictions of each model and optimize the final screening results. The method used in this embodiment is called the stack ensemble learning method, which optimizes the combination of the prediction results of different basic learners through the meta-learning algorithm. The advantage of this method is that it can give full play to the synergistic effect of multiple high-performance models to achieve more accurate classification or regression predictions.

[0127] Further, model evaluation. Reference 9 shows that the model shows high specificity (training set: 97.50%, test set: 97.35%), sensitivity (training set: 91.72%, test set: 93.10%) and accuracy (training set: 95.7%, test set: 96.03%) on both the training set and the test set, indicating that the model has good generalization ability and low overfitting risk.

[0128] Table 9

[0129]

[0130] Referring to Table 10, the sensitivity at different stages is as follows: the sensitivity of the model is low in stage I and stage II, and high in stage III and stage IV. However, even in the early stage of cancer (stage I), the sensitivity is still high (91%).

[0131] Table 10

[0132] I II III IV Training set 91.42％ 91.50％ 91.36％ 100.00％ Test Set 90.48％ 94.29％ 92.42％ 100.00％

[0133] The beneficial effects of the present invention are as follows:

[0134] The present invention realizes a non-invasive and cost-effective colorectal cancer detection method by constructing an early screening model for colorectal tumors, reduces the economic cost of sequencing, and significantly improves the accuracy and sensitivity of detection.

[0135] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0136] The principles and implementation methods of the present invention are described in this article using specific examples. The description of the above embodiments is only used to help understand the method and core idea of ​​the present invention. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as limiting the present invention.

Claims

1. A method for early screening of colorectal cancer based on multi-omics sequencing, characterized in that: include: Collecting plasma samples to be tested; Performing DNA methylation sequencing on the plasma sample to be tested to obtain sequencing data to be tested; The sequencing data to be tested is input into the trained colorectal tumor early screening model for classification to obtain the target screening results. The training process of the colorectal tumor early screening model includes: Collecting colorectal cancer tissue and adjacent normal tissue samples from colorectal cancer patients, extracting genomic DNA and performing DNA methylation sequencing on the colorectal cancer tissue and the adjacent normal tissue samples to obtain original sequencing data; The raw sequencing data is sequentially subjected to quality control and compared with the human reference genome to obtain sequence correspondence; Dividing the human reference genome into a plurality of CpG island regions, and analyzing the methylation levels of different haploids in each of the CpG island regions to obtain reference methylation data; According to the sequence correspondence, the reference methylation data and the original sequencing data are used to perform a P value corrected by the Benjamin and Hochberg multiple test to analyze the difference in methylation levels, and obtain tumor-specific regions; Collecting plasma samples from colorectal cancer patients and healthy subjects, and performing cfDNA extraction and methylation sequencing on the plasma samples to obtain plasma cfDNA sequencing data; The plasma cfDNA sequencing data is sequentially subjected to quality control and human reference genome alignment to obtain coordinate-determined sequencing data; Counting the coordinates to determine the methylation degree of the sequencing data in the tumor-specific region to obtain a methylation feature set; Obtaining the coordinates to determine the coordinates of all read segment data of the sequencing data on the reference genome, taking the p bases at the 3' end of the read segment data on the human reference genome as the first base fragment, and calculating the proportion of the first base fragment to obtain the terminal motif feature; The q bases upstream and q bases downstream of the 3' end of the read data are taken as the second base fragment, and the proportion of the second base fragment is calculated to obtain the breakpoint motif feature; Combining the methylation feature set, the terminal motif feature, and the breakpoint motif feature into an initial feature value; The initial characteristic value is used as input data, and the disease condition is used as the output result, and the pre-designed integrated machine learning classifier is trained to obtain the colorectal tumor early screening model.

2. The method for early screening of colorectal cancer based on multi-omics sequencing according to claim 1, characterized in that: The haploid includes any one or more of MM, MHL, CHALM, PDR and Entropy.

3. The method for early screening of colorectal cancer based on multi-omics sequencing according to claim 1, characterized in that: The p is any integer between 4 and 10; the q is any integer between 2 and 5.

4. The method for early screening of colorectal cancer based on multi-omics sequencing according to claim 1, characterized in that: The integrated machine learning classifier includes: a first layer and a second layer; the first layer includes: a support vector machine, a random forest, XGBoost and CatBoost; the second layer includes: a logistic regression network.

Images