Method, system and equipment for diagnosing malignant tumor or predicting curative effect of malignant tumor patient based on mature red blood cell DNA characteristics

Abstract

The invention discloses a method, a system and equipment for diagnosing malignant tumors or predicting curative effects of malignant tumor patients based on mature red blood cell DNA (deoxyribonucleic acid) characteristics. Clinical significance of red blood cell DNA of a malignant tumor patient is studied, it is found that RBC DNA features are different between healthy people and the malignant tumor patient, it is further found through research that the RBC DNA concentration in the malignant tumor patient is related to the curative effect of the malignant tumor patient, and on the basis, the invention provides a method for diagnosing malignant tumor based on mature red blood cell DNA features. Meanwhile, the invention provides a method, a system and equipment for predicting the curative effect of a malignant tumor patient based on the red blood cell DNA concentration. The invention provides reference for knowing RBC DNA and exploring the potential of the RBC DNA as a clinical marker.

Description

Technical Field

[0001] The present invention belongs to the field of bioinformatics, and in particular, relates to a method, system and device for diagnosing malignant tumors or predicting the efficacy of malignant tumor patients based on mature red blood cell DNA characteristics. Background Art

[0002] Due to the lack of nuclei in mature red blood cells, red blood cell DNA has long been overlooked. In recent years, it has been discovered that DNA exists in mature human red blood cells. Studies have reported that in infectious diseases, red blood cells can act as "immune sentinels" to absorb pathogen-derived CpG DNA through surface TLR9 proteins, thereby activating macrophage-mediated innate immunity. Other studies have shown that RBCs contain long DNA fragments that can absorb driver mutations from the DNA of cancer cell lines through direct contact. However, circulating tumor cells (CTCs) are scarce, with only about 1 in 10 cells per 10 cell. 7 There is only ~1 CTC in 1 white blood cell. Uptake of DNA by red blood cells only through direct contact is an inefficient method. Considering that tumors can release DNA into plasma, it is unclear whether RBCs can absorb free DNA from tumors and its clinical significance. Therefore, it is important to explore the characteristics of mature red blood cell DNA and its clinical significance in patients with malignant tumors. Summary of the invention

[0003] In order to make up for the deficiencies of the prior art, the purpose of the present invention is to provide a method, system and device for diagnosing malignant tumors or predicting the efficacy of malignant tumor patients based on mature red blood cell DNA characteristics.

[0004] In order to achieve the above object, the present invention adopts the following technical solution:

[0005] The first aspect of the present invention provides a method for diagnosing malignant tumors based on mature red blood cell DNA characteristics.

[0006] Furthermore, the method is performed by a computer and comprises the following steps:

[0007] Acquiring data: acquiring data on DNA characteristics of red blood cells in the sample of the tested subject, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient;

[0008] Processing data: inputting the data of the red blood cell DNA characteristics into a constructed malignant tumor diagnosis model, wherein the malignant tumor diagnosis model predicts whether the subject to be tested has a malignant tumor based on the data of the red blood cell DNA characteristics;

[0009] Output the result.

[0010] Furthermore, the steps of constructing the malignant tumor diagnostic model are as follows:

[0011] Obtain data on red blood cell DNA characteristics, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient; the data on the red blood cell DNA characteristics come from normal healthy people and patients with malignant tumors; and input the data on the red blood cell DNA characteristics into a machine learning algorithm to construct a malignant tumor diagnosis model.

[0012] Preferably, the step of constructing the malignant tumor diagnosis model also includes verifying the model effectiveness, and the method of verifying the model effectiveness includes a ROC curve.

[0013] Furthermore, the malignant tumor diagnostic model obtains results through the following criteria: when the DNA concentration or DNA length variation coefficient is higher than a threshold value or more than 50% of the DNA fragment lengths do not fall within the reference value range, a classification result is obtained that the subject to be tested suffers from a malignant tumor; if the DNA concentration or DNA length variation coefficient is lower than a threshold value or more than 50% of the DNA fragment lengths are within the reference value range, a classification result is obtained that the subject to be tested does not suffer from a malignant tumor.

[0014] In some embodiments of the present invention, the preset threshold is a representative value of a normal sample of a healthy population, including but not limited to a maximum value, a third quartile, and an average value. In some preferred embodiments of the present invention, the population sample includes more than 20 samples, such as 30, 50, 80, 100, 150, 200, 300, 500 or more. In a specific embodiment of the present invention, the preset threshold of the DNA concentration is 113 ng / ml, the preset threshold of the DNA length coefficient of variation is 1.032, and the reference value range of the DNA fragment length is 1 kb to 8 kb.

[0015] Furthermore, the machine learning algorithm includes algorithm models developed using various development tools.

[0016] Furthermore, the development tools include but are not limited to TensorFlow, Scikit Learn, PyTorch, OpenNN, RapidMiner, Azure Machine Learning, Apache Mahout, Shogun, KNIME, VertexAI, H2Oai, Anaconda, Keras, Tableau, Fast.ai, Catalyst, Amazon ML, MLJAR, and Spell.

[0017] Furthermore, the algorithm model includes but is not limited to linear regression model, logistic regression model, Lasso regression model, Ridge regression model, linear discriminant analysis model, nearest neighbor model, decision tree model, perceptron model, neural network model, support vector machine model, naive Bayes model, AdaBoost model, GBDT model, XGBoost model, LightGBM model, CatBoost model, and random forest model.

[0018] Furthermore, the subject to be tested includes humans and / or mammals.

[0019] Furthermore, the malignant tumors include gastric cancer, pancreatic cancer, esophageal cancer, rectal cancer, colon cancer, lung cancer, nasopharyngeal cancer, laryngeal cancer, malignant lymphoma, melanoma, endometrial cancer, cervical cancer, ovarian cancer, bladder cancer, and kidney cancer.

[0020] Preferably, the malignant tumor is lung cancer.

[0021] A second aspect of the present invention provides a system for diagnosing malignant tumors based on mature red blood cell DNA characteristics.

[0022] Further, the system comprises:

[0023] Data acquisition unit: used to acquire data on DNA characteristics of red blood cells in the sample of the subject to be tested, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient;

[0024] A data classification unit is used to classify and predict the malignant tumor diagnosis model obtained by the construction method described in the first aspect of the present invention on the data obtained by the data acquisition unit, and obtain a classification result of whether the subject to be tested suffers from a malignant tumor;

[0025] Result output unit: used to output classification results.

[0026] Furthermore, the malignant tumors include gastric cancer, pancreatic cancer, esophageal cancer, rectal cancer, colon cancer, lung cancer, nasopharyngeal cancer, laryngeal cancer, malignant lymphoma, melanoma, endometrial cancer, cervical cancer, ovarian cancer, bladder cancer, and kidney cancer.

[0027] Preferably, the malignant tumor is lung cancer.

[0028] The third aspect of the present invention provides a method for predicting the therapeutic effect of malignant tumor patients based on red blood cell DNA concentration.

[0029] Furthermore, the method is performed by a computer and comprises the following steps:

[0030] Acquiring data: Acquiring data on the red blood cell DNA concentration in samples of patients with malignant tumors to be tested;

[0031] Processing data: inputting the data of red blood cell DNA concentration into a constructed efficacy prediction model, wherein the efficacy prediction model predicts the efficacy of malignant tumor patients based on the data of red blood cell DNA concentration;

[0032] Output the prediction results.

[0033] Furthermore, the steps for constructing the efficacy prediction model are as follows: obtaining data on red blood cell DNA concentration, the data on red blood cell DNA concentration comes from a population of malignant tumor patients with a high disease control rate after treatment and a population of malignant tumor patients with a poor disease control rate after treatment; inputting the data on red blood cell DNA concentration into a machine learning algorithm to construct a efficacy prediction model.

[0034] Preferably, the step of constructing the efficacy prediction model also includes verifying the model effectiveness, and the method for verifying the model effectiveness includes a ROC curve.

[0035] Furthermore, the efficacy prediction model obtains prediction results through the following criteria: when the red blood cell DNA concentration is higher than a threshold value, a classification result is obtained that the disease control rate of the malignant tumor patient is poor after treatment; when the red blood cell DNA concentration is lower than a threshold value, a classification result is obtained that the disease control rate of the malignant tumor patient is high after treatment.

[0036] Furthermore, the malignant tumors include gastric cancer, pancreatic cancer, esophageal cancer, rectal cancer, colon cancer, lung cancer, nasopharyngeal cancer, laryngeal cancer, malignant lymphoma, melanoma, endometrial cancer, cervical cancer, ovarian cancer, bladder cancer, and kidney cancer.

[0037] Preferably, the malignant tumor is lung cancer.

[0038] A fourth aspect of the present invention provides a system for predicting the therapeutic effect of malignant tumor patients based on red blood cell DNA concentration.

[0039] Further, the system comprises:

[0040] Data acquisition module: used to obtain data on red blood cell DNA concentration in samples of patients with malignant tumors to be tested;

[0041] Data processing module: used for classifying and predicting the data obtained in the data acquisition module through the efficacy prediction model obtained by the construction method described in the third aspect of the present invention, and obtaining the classification results of the efficacy of the malignant tumor patients to be tested;

[0042] Output result module: used to output classification results.

[0043] Furthermore, the malignant tumors include gastric cancer, pancreatic cancer, esophageal cancer, rectal cancer, colon cancer, lung cancer, nasopharyngeal cancer, laryngeal cancer, malignant lymphoma, melanoma, endometrial cancer, cervical cancer, ovarian cancer, bladder cancer, and kidney cancer.

[0044] Preferably, the malignant tumor is lung cancer.

[0045] A fifth aspect of the present invention provides a computer device

[0046] Furthermore, the computer device includes: a memory and a processor, the memory is used to store program instructions; the processor is used to call the program instructions, and when the program instructions are executed, the method for diagnosing malignant tumors described in the first aspect of the present invention or the method for predicting the efficacy of malignant tumor patients described in the third aspect of the present invention is implemented.

[0047] A sixth aspect of the present invention provides a computer-readable storage medium.

[0048] Furthermore, the computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, it implements the method for diagnosing malignant tumors described in the first aspect of the present invention or the method for predicting the efficacy of malignant tumor patients described in the third aspect of the present invention.

[0049] Furthermore, the malignant tumors include gastric cancer, pancreatic cancer, esophageal cancer, rectal cancer, colon cancer, lung cancer, nasopharyngeal cancer, laryngeal cancer, malignant lymphoma, melanoma, endometrial cancer, cervical cancer, ovarian cancer, bladder cancer, and kidney cancer.

[0050] Preferably, the malignant tumor is lung cancer.

[0051] In the present invention, the lung cancer includes small cell lung cancer and non-small cell lung cancer. The small cell lung cancer occurs in the bronchial mucosa or glandular cancer, with a high degree of malignancy, including lymphocyte type, intermediate cell type and mixed type. The non-small cell lung cancer includes squamous cell carcinoma, adenocarcinoma and large cell carcinoma.

[0052] Advantages and beneficial effects of the present invention:

[0053] The present invention studies the clinical significance of red blood cell DNA (RBC DNA) in patients with malignant tumors, and finds that there are differences in RBC DNA characteristics between healthy people and patients with malignant tumors. Further research finds that the RBC DNA concentration in patients with malignant tumors is related to the therapeutic effect of malignant tumor patients. Based on this, the present invention provides a method for diagnosing malignant tumors based on mature red blood cell DNA characteristics, and also provides a method, system and device for predicting the therapeutic effect of malignant tumor patients based on red blood cell DNA concentration. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 A schematic diagram of the process of diagnosing malignant tumors provided by the present invention;

[0055] Figure 2A schematic diagram of the system structure for diagnosing malignant tumors provided by the present invention;

[0056] Figure 3 A schematic diagram of the process of predicting the therapeutic effect of patients with malignant tumors provided by the present invention;

[0057] Figure 4 A schematic diagram of the system structure for predicting the therapeutic effect of patients with malignant tumors provided by the present invention;

[0058] Figure 5 A schematic diagram of the structure of a computer device provided by the present invention;

[0059] Figure 6 The figure shows the difference in RBC DNA fragments between patients with malignant tumors and normal healthy people; among them, Figure A is a confocal image showing Hoechst 33342 (blue) and CD235a (yellow) in red blood cells, scale bar, 7.5μm; Figure B is a gel electrophoresis of red blood cell DNA in patients with malignant tumors; Figure C shows the overall fragment distribution (0-20kb) of red blood cell DNA and white blood cell DNA; Figure D is a box plot of the proportion of short (<1kb), medium (1kb-8kb) and long (>8kb) fragments; Figure E shows the distribution of RBC DNA with ladder marks in short fragments (<1kb) analyzed by ONT, CR represents red blood cell DNA in patients with malignant tumors, and the vertical lines are located at 170, 370, 570, 770 and 970bp; Figure F shows the distribution of red blood cell and white blood cell DNA in short fragments (<1kb); Figure G shows the distribution of RBC and WBC The distribution of DNA in the promoter region (within 3kb of TSS); H shows the distribution of short fragments (<1kb) of red blood cell and white blood cell DNA in the promoter region; I shows the frequency of 6 terminal motifs in different ONT samples; J shows the relative frequency of 4-mer motifs (left) and A / T / C / G bases (upper right) in terminal motif distribution map I, and the lower right is the sequence logo map of distribution map I; K shows the coefficient of variation of fragment lengths of malignant tumors and normal sources;

[0060] Figure 7Figure 4 shows the fragment characteristics and methylation patterns of tumor-derived DNA in red blood cell DNA; Figure A is a schematic diagram of the establishment and sampling of the NCG mouse CDX model; Figure B shows the tumor size (left) and volume (right) after 1 month of growth; Figure C shows the coverage of human DNA in mouse RBC DNA throughout the genome, with NG1 on the left, NG4 in the middle, and NG2 on the right; Figure D shows the proportion of red blood cell DNA of human origin; Figure E shows the distribution of human DNA (left), fragments with CCCA terminal motifs (middle), and fragments with CCNN terminal motifs (right); Figure F shows the proportion of CCNN motifs in different fragment lengths, Anova: one-way analysis of variance; Figure G shows the proportion of CCCA motifs in different fragment lengths; Figure H shows RBC from LLC1 (left), LLC2 (middle), and LLC3 (right) samples Length distribution of human and mouse DNA in DNA; Figure I shows CNVs of short (upper) and long (lower) fragments in CR4 samples, and tumor purity inferred by ichorCNA; Figure J shows the methylation distribution of ONT samples (left) and MDS analysis of methylation patterns in long (>8kb) fragments (right), and the whole genome methylation of ONT is located at the probe site of H450K methylation array; Figure K shows 5mC and 5hmC methylation of long fragments in the promoter region of CR1 samples; Figure L shows the fragment characteristics of somatic mutation-carrying fragments (left) and germline mutation-carrying fragments (right) in RBC DNA of ID29;

[0061] Figure 8 The figure shows the relationship between red blood cell DNA concentration, lysis characteristics and clinical indicators; Figure A is a scatter plot showing the relationship between the CV of fragment lengths of 150-340bp (upward) and 340-400bp (downward) and the inverse of the tumor purity inferred from cfDNA; Figure B is a waterfall plot with a CV of 340-400bp fragment length and efficacy evaluation, PD, progressive disease, PR, partial remission, SD, stable disease; Figure C shows the difference in red blood cell DNA levels in patients with different histological types; Figure D shows patients with malignant tumors at different clinical TNM stages Figure 1 shows the difference in red blood cell DNA levels; Figure E shows the box plot of red blood cell DNA concentrations in different tumor sizes; Figure F shows a scatter plot of the relationship between red blood cell DNA levels and HGB, where HGB is hemoglobin; Figure G shows a scatter plot of the relationship between red blood cell DNA levels and RDW; Figure H is a scatter plot showing the relationship between red blood cell DNA levels and the natural logarithm of CA125 (left) and NSE (right); Figure I shows the relationship between red blood cell DNA levels and efficacy evaluation; Figure J shows the changes in red blood cell DNA concentrations and tumor marker levels in consecutively sampled specimens. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0063] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The serial numbers of the operations, such as 101, 102, etc., are only used to distinguish different operations, and the serial numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0064] 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0065] Figure 1 The present invention provides a flow chart of a method for diagnosing a malignant tumor. Specifically, the method comprises:

[0066] 101: Acquire data, acquire data of DNA characteristics of red blood cells in a sample of a test subject, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient;

[0067] In some embodiments of the present invention, the applicant has found through extensive and in-depth research that there are significant differences in RBC DNA characteristics between healthy people and patients with malignant tumors, including DNA concentration, DNA length variation coefficient and DNA fragment length. This suggests that DNA concentration, DNA length variation coefficient and DNA fragment length in red blood cells can be used as potential markers for diagnosing malignant tumors.

[0068] In some embodiments, the subject can be human or non-human, and can include, for example, animal strains or species used as a "model system" for research purposes. Likewise, the subject can include an adult or adolescent (e.g., a child). In addition, the subject can refer to a mammal (e.g., human or non-human). Examples of mammals include, but are not limited to, any member of the class of mammals: humans, non-human primates (e.g., chimpanzees), and other apes and monkeys; livestock, such as cattle, horses, sheep, goats, pigs; domestic animals, such as rabbits, dogs, and cats; experimental animals include rodents, such as rats, mice, and guinea pigs, etc. Examples of non-mammals include, but are not limited to, birds, fish, etc.

[0069] In the context of the present invention, the term "sample" as used herein refers to a composition obtained from or derived from a patient / subject, which contains cells and / or other molecular entities to be characterized and / or identified based on, for example, physical, biochemical, chemical and / or physiological characteristics. For example, a sample refers to any sample derived from a patient / subject that is expected or known to contain cells and / or molecular entities to be characterized. Samples include, but are not limited to, tissue samples, primary or cultured cells or cell lines, cell cultures, cell supernatants, cell lysates, platelets, serum, plasma, vitreous humor, lymphatic fluid, synovial fluid, follicular fluid, semen, pancreatic juice, amniotic fluid, milk, whole blood, blood-derived cells, urine, cerebrospinal fluid, saliva, sputum, tears, sweat, mucus, tissue culture fluid, tissue extracts, homogenized tissues, cell extracts, and combinations thereof.

[0070] In a specific embodiment of the present invention, the sample includes red blood cells in the blood of healthy people and red blood cells in patients with malignant tumors. Blood samples and clinical data are from the internal malignant tumor cohort of Zhongnan Hospital of Wuhan University. Genetic detection information of the patient's primary lesion comes from previous clinical records. Blood samples from 2 healthy blood donors were collected from the applicant and his colleagues. This study was approved by the Medical Ethics Committee of Zhongnan Hospital of Wuhan University (No.: 2022032K). All individuals provided informed consent.

[0071] In the present invention, the DNA length variation coefficient refers to the stability and diversity evaluation index of genomic DNA during DNA replication. The detection of DNA variation coefficient is usually carried out by stability analysis of genomic DNA. Specific detection methods include PCR amplification, sequencing, etc. The measurement method of DNA fragment length can be measured by base pair technology or molecular biology technology (high-throughput sequencing, PCR, etc.). The determination method of DNA concentration includes measuring the absorbance of DNA at 260nm using an ultraviolet spectrophotometer, and calculating the concentration of DNA according to the absorbance value. Therefore, the data of DNA characteristics can be detected by methods well known in the art.

[0072] In one embodiment of the present invention, we found that the length, concentration and coefficient of variation of red blood cell DNA fragments in lung cancer patients were different from those in the normal population. In order to study the red blood cell DNA of lung cancer patients, we used density gradient centrifugation to separate red blood cells from peripheral blood samples of our internal cohort. Flow cytometry confirmed that >99% of the isolated cells were CD45 negative (white blood cell marker) and CD235a positive (red blood cell marker), which verified the purity of the red blood cells we extracted. Using confocal microscopy, we can see the DNA inside mature human red blood cells ( Figure 6 A), RBC DNA of lung cancer patients is long fragments >10kb ( Figure 6 B). Due to the low sensitivity of gel electrophoresis, the exact size of RBC DNA remains unknown. Oxford Nanopore Technology (ONT) is a new method to characterize whole genome fragment length. We extracted RBC DNA from 2 healthy donors and 6 lung cancer patients for ONT sequencing. Compared with lung cancer patients, the distribution of RBC DNA in healthy subjects was skewed to the right and intersected with that of lung cancer patients at around 8 kb ( Figure 6 C). Compared with healthy donors (<1 kb, median 20.5%; >8 kb, median 16.4%), RBC DNA in lung cancer patients was more distributed in <1 kb (median 29.7%) and >8 kb (median 21.3%; Figure 6 D). We found that the short fragments of RBC DNA in tumor patients were mainly around 170 bp. Interestingly, in the RBC DNA of three malignant tumor patients (CR1, CR4, and CR6), DNA fragments showed periodic characteristics, with a peak at 190-200 bp and extending to 1kb-2kb ( Figure 6 E), which is similar to the apoptotic ladder produced by apoptosis. No cycle characteristics were found in the white blood cell DNA (paired with CR1) and red blood cell DNA of healthy people ( Figure 6 F).

[0073] After annotating the DNA to genomic regions, most DNA samples showed no enrichment or depletion near the transcription start site (TSS), except for CR1 and CR6 ( Figure 6 G). When focusing on short fragments <1 kb, we observed a decrease in coverage in the promoter regions for most DNA, however, an increase in coverage was observed for CR1, CR4, and CR6 samples ( Figure 6H). These samples were associated with apoptotic ladders, suggesting that the ladder features in erythrocyte DNA are more detached from transcriptionally active regions. In addition, the DNA fragmentation process is also affected by DNA nucleases. Different nucleases have different preferences for cutting and generating 5' terminal base sequences. 5'4-mer ends have been used previously to describe the fragmentation pattern of cfDNA. Here, we calculated the frequency of all 256 groups of 4-mer A / T / C / G motifs in RBC DNA. Using the non-negative matrix factorization (NMF) algorithm, we identified 6 clusters of terminal pattern profiles ( Figure 6 I). Profile I, characterized by the presence of an A base at position 1, was more common in RBC DNA from tumor patients than in healthy donors. Figure 6 J). The pattern of profile I is close to the cleavage profile of DNA fragmentation factor subunit beta (DFFB). We also found that profile V is prominent in CR1, CR4, and CR6, which have the characteristics of an apoptotic ladder. Profile V is dominated by T-terminal groups. It is worth noting that the coefficient of variation of RBC DNA fragment length in patients with malignant tumors is significantly higher than that in normal subjects ( Figure 6 K), which indicates that the RBC DNA of tumor patients is more disordered. Taken together, these results indicate that the length and coefficient of variation of RBC DNA fragments in patients with malignant tumors are different from those in the normal population.

[0074] Furthermore, we demonstrated through animal experiments that tumor-derived DNA in RBC DNA is concentrated in short and long fragments rather than medium-length fragments. To characterize the characteristics of tumor DNA in RBC DNA, we constructed a cell line-derived xenograft (CDX) model, implanted human PC9 cell lines in the axilla of NCG mice, and selected three mice with different tumor sizes (NG1, NG2, and NG4) one month later, extracted their RBC DNA for WGS detection (depth 22-26, Figure 7 A and 7B). The NG1 tumor tissue was the largest and showed rib infiltration. RBC DNA matching with humans was considered to be of tumor origin. NG1 DNA had the highest proportion of human origin and NG2 had the lowest proportion, which was consistent with the tumor size ( Figure 7 C and 7D).

[0075] Next, we studied the fragmentation characteristics of red blood cell DNA. The peak of NG1 fragments was between 50 and 150 bp, while the fragments of NG2 and NG4 were longer, with a peak of around 300 bp ( Figure 7E). Previous studies have shown that circulating tumor DNA fragments are concentrated in the range of <150bp. Using ichorCNA, we found that 50-150bp fragments had higher tumor purity compared with all fragments. In addition, we found human DNA mutations in RBC DNA, among which mutations in 50-150bp fragments were the most abundant. The terminal motifs CCCA and CCNN are downregulated in cfDNA from patients with malignant tumors because of insufficient expression of DNASE1L3, the enzyme that produces CCNN cleavage preference in tumors. The pan-cancer TCGA database showed that DNASE1L3 was generally downregulated in tumor tissues compared with normal tissues. We studied the length distribution of CCCA motif fragments and found that they were mainly distributed around 300bp ( Figure 7 E). CCCA and CCNN motifs are least abundant in the range of 50 to 150 bp and most abundant in the range of 200 to 340 bp ( Figure 7 F and 7G), indicating that the ~300 bp NG2 and NG4 enriched fragments are not of tumor origin. In addition, after co-culture with the LLC cell supernatant mentioned in Section II, the distribution of mouse-derived (i.e. LLC-derived) DNA in RBC DNA is shorter than that of human-derived DNA ( Figure 7 H).

[0076] In the ONT sequencing dataset, we performed ichorCNA to infer the tumor purity of short, medium, and long fragments. The tumor purity of short fragments was the highest, the tumor purity of medium fragments was the lowest, and the purity of long fragments was not low ( Figure 7 I). Next, we showed the methylation pattern of long fragments. The methylation level of long fragments of red blood cell DNA in lung cancer patients was lower than that in normal subjects ( Figure 7 J). Multidimensional scaling (MDS) analysis distinguished between red blood cell DNA from tumor patients, red blood cell DNA from normal subjects, and white blood cell DNA, whereas MDS analysis of medium-length fragments could not distinguish between them ( Figure 7 J). This suggests that the long fragments in erythrocyte DNA are tumor-specific. In addition, we found that 5mC of the long fragments of erythrocyte DNA was hypomethylated in the promoter region, while 5hmC was hypermethylated ( Figure 7 K). This feature was also found in short and medium fragments. Short fragments have the most disordered methylation in the promoter region. Erythrocyte DNA fragments with somatic mutations were defined as tumor-derived in high-depth targeted sequencing data. The SVM method was used to determine whether a DNA fragment had somatic mutations. We found that somatic mutations were concentrated in 90-150bp and 340-400bp ( Figure 7 L), and surprisingly, this interval remained consistent across patients. In contrast, the distribution of germline variants was more right-shifted. In summary, we found a preference for tumor-derived fragments smaller than 1 kb or larger than 8 kb in erythrocyte DNA.

[0077] In another embodiment of the present invention, we found that red blood cell DNA levels were associated with tumor burden in lung cancer patients. To facilitate lung cancer management, we performed clinical analysis on RBC DNA in 87 patients, of whom 16 patients underwent 1021 sets of targeted sequencing and 71 patients underwent RBC DNA concentration testing. The targeted sequencing cohort enabled us to study the clinical significance of fragment features. The length coefficient of variation of 340-400bp RBC DNA was positively correlated with tumor purity inferred from cfDNA, while 150-340bp was negatively correlated with tumor purity ( Figure 8 A). In addition, we observed that patients with a low coefficient of variation of fragment length between 340 and 400 bp were more likely to achieve disease control rate (DCR) in the most recent efficacy evaluation before blood collection (100% vs 54.5%, Fisher's test P = 0.12, the critical value was the mean; Figure 8 B). We collected blood samples from 71 patients in our internal cohort, of whom 58 had non-small cell lung cancer, 9 had small cell lung cancer, 1 had pulmonary lymphoma, 1 had pneumonia, 1 had tuberculosis, and 1 was unknown. Two patients with infectious diseases had low levels of RBC DNA ( Figure 8 C). Except for the T1 stage, the RBC DNA concentration of patients with malignant tumors in stages T2 to 4 gradually increased, and that of patients in stage N3 was significantly higher than that of patients in stage N1, and that of patients in stage M1 was higher than that of patients in stage M0 ( Figure 8 D). From the perspective of clinical staging, patients in stages III to IV have higher levels of RBC DNA. Patients with tumors larger than 30 mm in maximum diameter have higher RBC DNA concentrations ( Figure 8 E), which indicates that there is a positive correlation between RBC DNA and tumor burden and that RBC DNA concentration is a potential marker for the diagnosis of malignant tumors.

[0078] 102: Processing data: inputting the data of the red blood cell DNA characteristics into the constructed malignant tumor diagnosis model, wherein the malignant tumor diagnosis model predicts whether the subject to be tested has a malignant tumor based on the data of the red blood cell DNA characteristics;

[0079] In some embodiments of the present invention, the method of constructing a malignant tumor diagnostic model is known to those skilled in the art, and the step of associating data of red blood cell DNA characteristics with a certain possibility or risk can be implemented and realized in different ways.

[0080] In the context of the present invention, the term "machine learning" refers to the use of computers to simulate or implement human learning activities, and technicians usually use different development tools to build machine learning algorithm models. The development tools include but are not limited to TensorFlow, Scikit Learn, PyTorch, OpenNN, RapidMiner, Azure Machine Learning, ApacheMahout, Shogun, KNIME, VertexAI, H2Oai, Anaconda, Keras, Tableau, Fast.ai, Catalyst, Amazon ML, MLJAR, Spell. The algorithm model includes but is not limited to a linear regression model, a logistic regression model, a Lasso regression model, a Ridge regression model, a linear discriminant analysis model, a neighbor model, a decision tree model, a perceptron model, a neural network model, a support vector machine model, a naive Bayes model, an AdaBoost model, a GBDT model, an XGBoost model, a LightGBM model, a CatBoost model, or a random forest model.

[0081] In one embodiment, after constructing a malignant tumor diagnosis model, ROC curve analysis can be used to evaluate the effectiveness of the model.

[0082] ROC curve is the true positive rate (sensitivity) of the test for the false positive rate (100%-specificity) of the test. It is useful for describing the performance of a specific feature when distinguishing between two populations. Usually, feature data is selected across the entire population in ascending order based on the value of a single feature. Then, for each value of the feature, the true positive and false positive rates of the data are calculated. The true positive rate is determined by counting the number of cases with a value higher than the feature and divided by the total number of cases. The false positive rate is determined by counting the number of controls with a value higher than the feature and divided by the total number of controls. Although this definition refers to the case in which the feature is increased compared to the control, the definition is also applied to the case in which the feature is lower compared to the control (in this case, the sample with a value lower than the feature will be counted). ROC curve can be generated with respect to a single feature, and can be generated with respect to other single outputs, for example, a combination of two or more features can be combined with mathematical methods (for example, addition, subtraction, multiplication, etc.) to provide a single sum value, and this single sum value can be plotted in the ROC curve. Additionally, any combination of multiple features where the combination is derived from separate output values ​​can be plotted in a ROC curve.

[0083] 103: Output results.

[0084] Figure 2 This is a schematic diagram of the system structure for diagnosing malignant tumors provided by the present invention.

[0085] The system is programmed or otherwise configured to include a data acquisition unit 201, a data classification unit 202, and a result output unit 203:

[0086] Data acquisition unit: used to acquire data on DNA characteristics of red blood cells in the sample of the subject to be tested, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient;

[0087] A data classification unit is used to classify and predict the malignant tumor diagnosis model obtained by the construction method described in the first aspect of the present invention on the data obtained by the data acquisition unit, and obtain a classification result of whether the subject to be tested suffers from a malignant tumor;

[0088] Result output unit: used to output classification results.

[0089] The system may be the user's electronic device or a computer system located remotely from the electronic device.

[0090] Figure 3 The present invention provides a method for predicting the efficacy of a malignant tumor patient. Specifically, the method comprises:

[0091] 301: Acquire data, acquire data of red blood cell DNA concentration in a sample of a malignant tumor patient to be tested;

[0092] In one embodiment of the present invention, we observed that the concentration of red blood cell DNA in lung cancer patients was associated with efficacy. In the efficacy evaluation before the most recent blood draw, we found that patients with low RBC DNA levels had a higher DCR rate (92.1% vs 76.9%, Fisher's test P = 0.16, cutoff value was the top 25%; Figure 8 I). We analyzed all eight samples (from four patients) that had sequential blood draws and complete tumor marker information. In patients with increases or decreases in RBC DNA, they also had increases or decreases in tumor markers. Patients with smaller changes in RBC DNA also had smaller changes in tumor marker levels ( Figure 8 J). In conclusion, our results provide a comprehensive reference for understanding RBC DNA and exploring its potential as a clinical marker.

[0093] 302: Processing data, inputting the data of red blood cell DNA concentration into a constructed efficacy prediction model, wherein the efficacy prediction model diagnoses the efficacy of malignant tumor patients based on the data of red blood cell DNA concentration;

[0094] 303: Output prediction results.

[0095] Figure 4This is a schematic diagram of the system structure for predicting the therapeutic effect of patients with malignant tumors provided by the present invention.

[0096] The system is programmed or otherwise configured to include a data acquisition module 401, a data processing module 402, and an output result module 403:

[0097] The data acquisition module 401 is used to acquire the data of the red blood cell DNA concentration in the sample of the patient with malignant tumor to be tested;

[0098] Data processing module 402: used to classify and predict the data obtained in the data acquisition module using the efficacy prediction model obtained by the construction method described in the third aspect of the present invention, and obtain the classification result of the efficacy of the malignant tumor patient to be tested;

[0099] Output result module 403: used to output classification results.

[0100] Figure 5 A schematic diagram of the structure of a computer device provided by the present invention.

[0101] The computer device 600 includes a processor 601 and a memory 602 coupled to the processor 601, wherein the memory 602 stores program instructions, and when the program instructions are executed by the processor 601, the processor 601 executes the above-mentioned method for diagnosing malignant tumors or the method for predicting the therapeutic effect of malignant tumor patients. The processor 601 may also be referred to as a CPU (Central Processing Unit). The processor 601 may be an integrated circuit chip having signal processing capabilities. The processor 601 may also be a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0102] The computer device 600 may be a mobile electronic device.

[0103] It should be understood that the systems, devices and methods described in the present invention can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0104] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0105] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above-mentioned integrated module may be implemented in the form of hardware or in the form of software functional modules.

[0106] The above are only implementation methods of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for diagnosing malignant tumors based on mature red blood cell DNA characteristics, characterized in that: The method is performed by a computer and comprises the following steps: Acquiring data: acquiring data on DNA characteristics of red blood cells in the sample of the tested subject, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient; Processing data: inputting the data of the red blood cell DNA characteristics into a constructed malignant tumor diagnosis model, wherein the malignant tumor diagnosis model predicts whether the subject to be tested has a malignant tumor based on the data of the red blood cell DNA characteristics; Output the result.

2. The method according to claim 1, characterized in that: The steps for constructing the malignant tumor diagnostic model are as follows: Acquire data on red blood cell DNA characteristics, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient; the data on the red blood cell DNA characteristics are from normal healthy people and patients with malignant tumors; input the data on the red blood cell DNA characteristics into a machine learning algorithm to construct a malignant tumor diagnosis model; Preferably, the step of constructing the malignant tumor diagnosis model further includes verifying the model effectiveness, and the method of verifying the model effectiveness includes ROC curve; Preferably, the malignant tumor diagnostic model obtains results through the following criteria: when the DNA concentration or DNA length variation coefficient is higher than a threshold value or more than 50% of the DNA fragment lengths do not fall within the reference value range, a classification result is obtained that the subject to be tested suffers from a malignant tumor; if the DNA concentration or DNA length variation coefficient is lower than a threshold value or more than 50% of the DNA fragment lengths are within the reference value range, a classification result is obtained that the subject to be tested does not suffer from a malignant tumor.

3. A system for diagnosing malignant tumors based on mature red blood cell DNA characteristics, characterized in that: The system comprises: Data acquisition unit: used to acquire data on DNA characteristics of red blood cells in the sample of the subject to be tested, wherein the DNA characteristics include one or more of the following: DNA concentration, DNA fragment length, and DNA length variation coefficient; A data classification unit: used to classify and predict the malignant tumor diagnosis model obtained by the construction method of claim 2 on the data obtained by the data acquisition unit, and obtain a classification result of whether the subject to be tested suffers from a malignant tumor; Result output unit: used to output classification results.

4. The method according to any one of claims 1 to 2 or the system according to claim 3, characterized in that: The malignant tumors include gastric cancer, pancreatic cancer, esophageal cancer, rectal cancer, colon cancer, lung cancer, nasopharyngeal cancer, laryngeal cancer, malignant lymphoma, melanoma, endometrial cancer, cervical cancer, ovarian cancer, bladder cancer, and kidney cancer; Preferably, the malignant tumor is lung cancer.

5. A method for predicting the therapeutic effect of malignant tumor patients based on red blood cell DNA concentration, characterized in that: The method is performed by a computer and comprises the following steps: Acquiring data: Acquiring data on the red blood cell DNA concentration in samples of patients with malignant tumors to be tested; Processing data: inputting the data of red blood cell DNA concentration into a constructed efficacy prediction model, wherein the efficacy prediction model diagnoses the efficacy of malignant tumor patients based on the data of red blood cell DNA concentration; Output the prediction results.

6. The method according to claim 5, characterized in that The steps of constructing the efficacy prediction model are as follows: obtaining data on red blood cell DNA concentration, the data on red blood cell DNA concentration coming from a population of patients with malignant tumors having a high disease control rate after treatment and a population of patients with malignant tumors having a poor disease control rate after treatment; inputting the data on red blood cell DNA concentration into a machine learning algorithm to construct a efficacy prediction model; Preferably, the step of constructing the efficacy prediction model also includes verifying the model effectiveness, and the method of verifying the model effectiveness includes ROC curve; Preferably, the efficacy prediction model obtains the prediction result through the following criteria: when the red blood cell DNA concentration is higher than a threshold value, a classification result is obtained that the disease control rate of the malignant tumor patient is poor after treatment; when the red blood cell DNA concentration is lower than a threshold value, a classification result is obtained that the disease control rate of the malignant tumor patient is high after treatment.

7. A system for predicting the efficacy of malignant tumor patients based on red blood cell DNA concentration, characterized in that: The system comprises: Data acquisition module: used to obtain data on red blood cell DNA concentration in samples of patients with malignant tumors to be tested; A data processing module: used to classify and predict the efficacy prediction model obtained by the construction method of claim 6 on the data obtained in the data acquisition module, and obtain the classification results of the efficacy of the malignant tumor patients to be tested; Output result module: used to output classification results.

8. The method according to any one of claims 5-6 or the system according to claim 7, characterized in that: The malignant tumors include gastric cancer, pancreatic cancer, esophageal cancer, rectal cancer, colon cancer, lung cancer, nasopharyngeal cancer, laryngeal cancer, malignant lymphoma, melanoma, endometrial cancer, cervical cancer, ovarian cancer, bladder cancer, and kidney cancer; Preferably, the malignant tumor is lung cancer.

9. A computer device, characterized in that: The computer device comprises: a memory and a processor, the memory being used to store program instructions; the processor being used to call program instructions, and when the program instructions are executed, the method for diagnosing a malignant tumor as described in any one of claims 1-2, or the method for diagnosing whether a malignant tumor patient is anemic as described in any one of claims 4-5, or the method for predicting the efficacy of a malignant tumor patient as described in any one of claims 7-8 is implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for diagnosing a malignant tumor according to any one of claims 1-2 or the method for predicting the efficacy of a malignant tumor patient according to any one of claims 5-6 is implemented.

Images