Rapid intelligent prediction method for FUT3 gene mutation site and morbidity

Abstract

The invention provides a rapid intelligent prediction method for FUT3 gene mutation sites and morbidity, and the method comprises the steps: carrying out the collection of the occurrence distribution probability data of the mutation sites and the collection of the morbidity data of patients and normal people, and carrying out the recognition of the morbidity through an artificial intelligence recognition model. According to the method, the probability of illness is predicted only by detecting four or less types of optional mutation points of a to-be-detected patient, so that screening and risk prevention and control are carried out in time, a good reference is provided for adopting a pre-illness intervention mechanism, and the cost of completely detecting five gene mutation points in the prior art is saved.

Description

[0001] This case is a series of applications of the prior application with application number "2022113687863" and invention name "FUT3 gene mutation site detection primers and probes, detection kits and PCR detection methods", the full text of which is introduced into this article by reference. Technical Field

[0002] The present invention relates to the detection of FUT3 gene mutation sites, and in particular to a rapid intelligent prediction method for detecting the genotype and incidence rate of five mutation sites of the FUT3 gene associated with Lewis antigen negativity, and belongs to the field of gene mutation detection. Background Art

[0003] The gene mutation detection of Lewis antigen negativity caused by FUT3 gene mutation mainly includes fluorescence measurement of the distribution of T59G, T202C, C314T, G508A and T1067A mutations with five corresponding primers and probes, thereby analyzing the amplification results. However, five expensive primers and probe reagents are required, and it is impossible to confirm the prediction of the incidence rate at different age levels. Therefore, it can only provide a reference for the overall population risk of pancreatic cancer.

[0004] A better reference method is to study the mutation distribution of these five sites and the different incidence rates in different age groups, especially the incidence of other digestive system organ cancers other than pancreatic cancer in Lewis antigen-negative patients. This will enable the risk of more diseases to be predicted while reducing the cost of using detection reagents. Summary of the invention

[0005] In view of the above-mentioned defects of the prior art, the main purpose of the present invention is to provide a rapid intelligent prediction method for determining whether a sample is Lewis antigen negative and the incidence rate for non-diagnostic or therapeutic purposes, comprising:

[0006] 1. Data Collection and Modeling Steps

[0007] A. Recruit healthy people and patients to form volunteer groups, divide the groups by age, and conduct the following tests:

[0008] (1) Extracting whole blood genomic DNA from samples;

[0009] (2) using whole blood genomic DNA of the sample as a template, using up to 4 real-time fluorescent PCR primers and probes in the following combination: a real-time fluorescent PCR primer and probe of T59G and a mixture of up to 3 real-time fluorescent PCR primers and probes of C314T, T202C, G508A, and T1067A to establish a fluorescent PCR amplification system for TaqMan PCR amplification;

[0010] (3) Collecting fluorescence signals and exporting data results for amplification result analysis;

[0011] B. Based on the analysis of the amplification results, the base sequence distribution data of the mutations corresponding to up to 4 real-time fluorescent PCR primers and probes in each age group, as well as the statistics of the incidence rate, are obtained. Artificial intelligence models are constructed respectively, and the incidence rate is trained and learned to obtain the probability prediction of the disease in each age group under each distribution condition.

[0012] 2. Prediction Steps

[0013] C. Perform steps (1) to (3) on the subject to be tested, and substitute the obtained distribution data into the artificial intelligence model of the age group corresponding to the subject to be tested to predict the incidence rate, wherein:

[0014] The amplification result analysis includes that if the amplification result shows any one of the following (a)-(b), the sample is 100% Lewis antigen negative, and if it is (c), the sample has a 1 / 14 probability of being Lewis antigen negative:

[0015] (a) Homozygous mutation: any one of up to three homozygous mutations except T59G;

[0016] (b) T59G heterozygous mutation and up to 3 heterozygous mutations other than T59G;

[0017] (c) Compound heterozygous mutations in which the T59G homozygous mutation is combined with heterozygous mutations in up to three other mutations except T59G.

[0018] It should be noted that in case (a), no matter what mixture is selected, when only the HEX signal is detected, it must satisfy any one of the three conditions except T59G, so case (a) can be judged as negative. Similarly, case (c) is also judged as negative.

[0019] However, this is not the case in (b). In addition to T59G, at most three other mutations exist. It is not necessarily that C314T and T202C are heterozygous mutations at the same time, but it is possible that one of them is heterozygous. In this case, it is still impossible to determine that the test result is negative.

[0020] It should be emphasized that: although T202C was not selected, T59G and C314T were detected as heterozygous mutations at the same time, but it is possible that T202C is also mutated, or T202C is not mutated. Although it is still impossible to determine whether it is negative or not, that is, this situation does not belong to the combination of primers and probes that can determine the mutant genotype, it does not hinder the prediction of the incidence rate that is most concerned (see below), because the purpose of detecting mutations is to predict the incidence rate.

[0021] Preferably, the real-time fluorescent PCR primers and probes are selected from T59G, and any one of C314T, T202C, G508A, and T1067A corresponding PCR primers and probes.

[0022] More preferably, the real-time fluorescent PCR primers and probes are selected from T59G, C314T, and any one or any two corresponding PCR primers and probes among T202C, G508A, and T1067A.

[0023] Optionally, the analysis of the amplification results includes performing Gaussian fitting of HEX signals and FAM signals on the fluorescence peaks emitted by the probes in the mixture to obtain the presence or absence of the spectrum peaks of the corresponding probes, and judging the presence of mutation sites accordingly.

[0024] Optionally, the method for determining the situation (a) is that when only the HEX signal amplification is successful when detecting at most three midpoint sites other than T59G, it can be determined as a homozygous mutation;

[0025] The method for determining the situation (b) is to detect the T59G heterozygous mutation and when at most three sites other than T59G, both FAM signal and HEX signal are successfully amplified, and determine that at most three sites other than T59G have heterozygous mutations;

[0026] The determination method for the above situation (c) is that when detecting the T59G site, only when the FAM signal is successfully amplified, the T59G site can be determined as a homozygous mutation, and when detecting up to three sites other than T59G, when both the FAM signal and the HEX signal are successfully amplified, it is a compound heterozygous mutation;

[0027] The acquisition of the distribution data comprises the following steps:

[0028] S1 assigns different preset grayscale or RGB color value combinations to the bases A, T, C, and G, respectively, and considers the alleles, and presets one base sequence in each group as the starting sequence and the other base sequence as the relative sequence. The two groups of FUT3 gene sequences of the alleles are pixelated in sequence in the order of starting sequence-relative sequence according to the preset different grayscale or RGB color value combinations, and are converted into a square matrix image, and multiple square matrix images corresponding to multiple volunteers in different age groups are obtained in this way, wherein the base pairs of up to 4 mutation points are additionally represented by two groups, and each group has two identical preset grayscale or preset RGB color value combinations of pixels to represent the overall presence of heterozygous mutations and homozygous mutations respectively;

[0029] S2 divides the square array images into training set and validation set.

[0030] The construction of the artificial intelligence model includes:

[0031] S3 builds a convolutional neural network with a residual mechanism (RESNET), inputs the training set into the network, passes it through the fully connected layer at the output, and uses the softmax function for classification;

[0032] S4 uses the validation set to verify, calculates the loss function, performs backpropagation, optimizes the network parameters, and continues training until the loss function stabilizes and stops training.

[0033] Optionally, the incidence rate is an incidence rate including multiple diseases.

[0034] One of the secondary objects of the present invention is to provide a mixture of real-time fluorescent PCR primers and probes for detecting the genotype and incidence prediction of up to 4 mutation sites of the FUT3 gene associated with Lewis antigen negativity based on the above-mentioned rapid intelligent prediction method, wherein the up to 4 mutation sites are selected from the real-time fluorescent PCR primers and probes of T59G and a mixture of up to 3 real-time fluorescent PCR primers and probes of C314T, T202C, G508A, and T1067A, wherein:

[0035] (1) The real-time fluorescent PCR primers for detecting the T59G mutation site are:

[0036] 5'CGCAGGTAGGAGAAGAAACACA3'

[0037] 5'CATGGCGCCGCTGTCT3'

[0038] The real-time fluorescent PCR probe for detecting the T59G mutation site is:

[0039] 5'FAM-CTGAAATAGCCGTGCGG-MGB 3'

[0040] 5'HEX-CTGAAATAGCAGTGCGG-MGB 3'

[0041] (2) The real-time fluorescence PCR primers for detecting the T202C mutation site are:

[0042] 5'CATCTCTGAACAGCGGGACAG 3'

[0043] 5'GCTGTGTGTTTCTTCTCCTACCT 3'

[0044] The probe for real-time fluorescent PCR to detect the T202C mutation site is:

[0045] 5'FAM-CATGTCCATAGCAGGAT-MGB 3'

[0046] 5'HEX-CATGTCCGTAGCAGGA-MGB 3'

[0047] (3) The real-time fluorescent PCR primers for detecting the C314T mutation site are:

[0048] 5'AGGCGTGACTTAGGGTTGGA 3'

[0049] 5'GCCGACCGCAAGGTGTAC 3'

[0050] The probe for real-time fluorescent PCR to detect the C314T mutation site is:

[0051] 5'FAM-ACGATGACCGTGTCT-MGB 3'

[0052] 5'HEX-ATGACCATGTCTGCC-MGB 3'

[0053] (4) The real-time fluorescent PCR primers for detecting the G508A mutation site are:

[0054] 5'AGCGCTGGATCTGGTTCAACT 3'

[0055] 5'GTCTTGGCCGAGAGGTTGAG 3'

[0056] The probe for real-time fluorescent PCR to detect the G508A mutation site is:

[0057] 5'FAM-CGCCCTACGGCTG-MGB 3'

[0058] 5'HEX-TTCACGCCCTACAGC-MGB 3'

[0059] (5) The real-time fluorescent PCR primers for detecting the T1067A mutation site are:

[0060] 5'CCATGCCGGCCTCTCA 3'

[0061] 5'ACTGCAGCAGGAATCCAGGTA 3'

[0062] The probe for real-time fluorescent PCR to detect the T1067A mutation site is:

[0063] 5'FAM-CCAAGCCGCTTTG-MGB 3'

[0064] 5'HEX-AACCAAGCCCGCTATG-MGB 3'.

[0065] The second secondary purpose of the present invention is to provide a real-time fluorescent PCR kit for determining whether a sample is Lewis antigen negative and a rapid intelligent prediction method for the incidence rate, comprising: DNA polymerase, Mg 2+ Buffer, dNTPs, quality control products, water, and real-time fluorescent PCR primers and probes for detecting up to 4 mutation sites, including the real-time fluorescent PCR primers and probes for T59G as described above and a mixture of up to 3 real-time fluorescent PCR primers and probes among C314T, T202C, G508A, and T1067A, wherein the quality control products are quality control products in which the genotypes of the up to 4 mutation sites of the FUT3 gene are heterozygous.

[0066] The TaqMan PCR amplification reaction procedure in step (2) may be: after pre-denaturation at 95°C for 10 min, run 45 cycles under the following conditions: denaturation at 95°C for 15 s, and annealing / extension at 60°C for 1 min.

[0067] The present invention designs real-time fluorescent PCR primers and probes for at most three or at most four mutation sites of the FUT3 gene T59G and C314T, T202C, G508A, and T1067A associated with Lewis antigen negativity, selects a small number of real-time fluorescent PCR primers and probe mixtures, utilizes Gaussian peak separation technology, obtains four types of Lewis antigen negativity-related mutation genotype judgment results, and simultaneously performs square matrix image conversion through base pair distribution data, and inputs the trained artificial intelligence model to determine the prediction of specific incidence rate, so as to provide a more accurate reference for the type of onset.

[0068] This small but special combination of four types of primers and probes can effectively avoid various interactions between primers and probes. On this basis, a TaqMan-MGB probe PCR detection method is established. By determining the genotype of only four mutation sites of the FUT3 gene associated with Lewis antigen negativity, it is determined whether the sample to be tested is Lewis antigen negative (situations (a) and (c)), or the incidence rate (especially for (b)). The present invention effectively improves the specificity of detection, greatly reduces false positives, thereby reducing detection errors, and can predict the incidence of various diseases, and also reduces detection costs. The detection of the genotype of the five mutation sites of the FUT3 gene associated with Lewis antigen negativity has high specificity and detection sensitivity.

[0069] From the patient's psychological perspective, when the test result (b) is negative, it gives the patient some confidence. At the same time, combined with the reference of the incidence rate, the patient can have a rough prediction of the onset of his or her disease, so as to prepare psychologically. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] Figure 1 This is a scatter plot of the T59G site amplification of the FUT3 gene in 100 cases of genomic DNA in the prior art.

[0071] Figure 2 This is a scatter plot of the T202C site amplification of the FUT3 gene in 100 genomic DNA samples in the prior art.

[0072] Figure 3 This is a scatter plot of the amplification of the C314T site of the FUT3 gene in 100 cases of genomic DNA in the prior art.

[0073] Figure 4 This is a graph showing the amplification of heterozygous genomic DNA at the T59G site of the FUT3 gene in the prior art.

[0074] Figure 5 This is a graph showing the amplification curve of the genomic DNA of the T59G site mutant of the FUT3 gene in the prior art.

[0075] Figure 6 This is a general flow chart of a method for rapidly and intelligently predicting whether a sample is Lewis antigen negative and the incidence rate for non-diagnostic or therapeutic purposes according to the present invention. DETAILED DESCRIPTION

[0076] 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 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.

[0077] Embodiment 1

[0078] 1. Preparation of mixture and kit

[0079] A mixture of real-time fluorescent PCR primers and probes for detecting the genotype of three mutation sites of the FUT3 gene associated with Lewis antigen negativity, wherein the three mutation sites are selected from the real-time fluorescent PCR primers and probes of T59G and the mixture of real-time fluorescent PCR primers and probes of C314T and T202C, wherein:

[0080] (1) The real-time fluorescent PCR primers for detecting the T59G mutation site are:

[0081] 5'CGCAGGTAGGAGAAGAAACACA3'

[0082] 5'CATGGCGCCGCTGTCT3'

[0083] The real-time fluorescent PCR probe for detecting the T59G mutation site is:

[0084] 5'FAM-CTGAAATAGCCGTGCGG-MGB 3'

[0085] 5'HEX-CTGAAATAGCAGTGCGG-MGB 3'

[0086] (2) The real-time fluorescence PCR primers for detecting the T202C mutation site are:

[0087] 5'CATCTCTGAACAGCGGGACAG 3'

[0088] 5'GCTGTGTGTTTCTTCTCCTACCT 3'

[0089] The probe for real-time fluorescent PCR to detect the T202C mutation site is:

[0090] 5'FAM-CATGTCCATAGCAGGAT-MGB 3'

[0091] 5'HEX-CATGTCCGTAGCAGGA-MGB 3'

[0092] (3) The real-time fluorescent PCR primers for detecting the C314T mutation site are:

[0093] 5'AGGCGTGACTTAGGGTTGGA 3'

[0094] 5'GCCGACCGCAAGGTGTAC 3'

[0095] The probe for real-time fluorescent PCR to detect the C314T mutation site is:

[0096] 5'FAM-ACGATGACCGTGTCT-MGB 3'

[0097] 5'HEX-ATGACCATGTCTGCC-MGB 3'

[0098] A real-time fluorescent PCR kit for determining whether a sample is Lewis antigen negative and a rapid intelligent prediction method for the incidence rate, comprising: DNA polymerase, Mg 2+ Buffer, dNTPs, quality control product, water, and a mixture of the above-mentioned real-time fluorescent PCR primer and probe including T59G and the real-time fluorescent PCR primer and probe including C314T and T202C, wherein the quality control product is a quality control product in which the genotype of the at most 4 mutation sites of the FUT3 gene is heterozygous.

[0099] 2. Detection Method

[0100] The main steps include:

[0101] (I) Obtaining genomic DNA of the sample to be tested

[0102] 1. Collection of volunteer specimens: Whole blood specimens are routinely collected by taking 2 ml of venous blood in EDTA anticoagulant tubes. After blood collection, the sample can be prepared for testing immediately or stored at -20°C for subsequent use.

[0103] 2. Extraction of whole blood genomic DNA: Use a blood genomic DNA extraction kit to extract genomic DNA and dilute the sample to 50 ng / μL for the next PCR reaction.

[0104] (ii) using the primers and probes in the above kit and genomic DNA as a template, performing fluorescent PCR amplification and collecting fluorescent signals.

[0105] 1. The TaqMan amplification reaction system includes: ultrapure water, DNA amplification enzyme premix, sample genomic DNA, and specific primers and probes for the above three sites of the FUT3 gene.

[0106] 2. TaqMan amplification reaction program: After pre-denaturation at 95°C for 10 min, run 45 cycles under the following conditions: denaturation at 95°C for 15 s, annealing / extension at 60°C for 1 min.

[0107] (III) After the fluorescence signal is collected, the data results are exported and Gaussian fitting of the spectrum peak is performed. The following test results can be determined as Lewis antigen negative or suspected negative genotypes.

[0108] (a) Homozygous mutation: a pure homozygous mutation of the mixture; that is, when only the HEX signal is successfully amplified when detecting the C314T and T202C mixed sites, it can be determined that there is a pure homozygous mutation. This situation is that either C314T or T202C is homozygous. At this time, it is a 100% Lewis antigen negative judgment.

[0109] (b) T59G heterozygous mutation, while C314T and T202C mixed sites also have heterozygous mutations; when detecting T59G, C314T and T202C mixed sites, when both FAM signal and HEX signal amplification are successful, it can be determined that 100% Lewis antigen is negative, that is, suspected negative.

[0110] (c) T59G homozygous mutation, C314T, T202C mixed sites have heterozygous mutations; this situation is that at least one of C314T and T202C is a heterozygous mutation. In this case, it is considered 100% Lewis antigen negative.

[0111] When detecting the T59G site, only when the FAM signal is successfully amplified can the T59G site be determined to be a homozygous mutation. When detecting the C314T and T202C mixed sites, when both the FAM signal component and the HEX signal component are amplified successfully in the Gaussian fitting results, it can be determined that the above four sites have heterozygous mutations.

[0112] Peripheral blood was collected from 100 patients with digestive tract tumors, genomic DNA was extracted, and the three sites of FUT3 gene T59G, T202C, and C314T were detected according to the above detection steps and methods.

[0113] Test results: 100 samples, with 10 years old as one age group, the detection rate of 40-50 years old for situation (a) is still 100% compared with the existing technology. The detection scatter diagram states of the corresponding sites of the existing technology are Figure 1 to Figure 3 Taking the T59G site detection results as reference, the amplification curves of heterozygous and mutant samples are shown in Figure 4-5 In this embodiment, the corresponding mutation type is determined based on this shape. Figure 6 The flow chart of the detection method is given.

[0114] Embodiment 2

[0115] The acquisition of the distribution data comprises the following steps:

[0116] S1 assigns different preset gray levels to bases A, T, C, and G. Figure 6 Under the premise of a pair of chromosome alleles α and β shown in the figure, the two groups of FUT3 gene sequences of the alleles are preset as the starting sequence and the other base sequence as the relative sequence in the zigzag order of the starting sequence-relative sequence, and each base in the base sequence of the two groups of alleles is arranged in pixel form in sequence according to the preset different grayscale or RGB color value combinations, and converted into a square matrix image.

[0117] Figure 6 Take a base A in the starting sequence as an example, and the base in the relative sequence is T, then the base sequence is given in a zigzag order...-AT-... The base sequence is pixelated according to different preset grayscale or RGB color value combinations, and the two groups of allele sequences are arranged in sequence from right to left and from top to bottom according to the arrow direction in the figure. The T59G, C314T and T202C sites are pixelated with another preset grayscale. For simplicity, Figure 6 A partial pixel grid in a square array image is given as an example.

[0118] Multiple square matrix images corresponding to multiple volunteers in different age groups are obtained in this way, wherein the base pairs of up to 4 mutation points are additionally represented by two groups of two identical preset grayscale or preset RGB color value combinations in each group to respectively represent the overall presence of heterozygous mutations and all homozygous mutations;

[0119] S2 divides the square array images into training set and validation set.

[0120] The construction of the artificial intelligence model includes:

[0121] S3 established a convolutional neural network with a residual mechanism (RESNET), input the training set into the network and then pass it through the fully connected layer at the output end, and used the softmax function to perform four classifications of pancreatic cancer, gastric cancer, intestinal cancer, and liver cancer;

[0122] S4 uses the validation set to verify, calculates the loss function, performs backpropagation, optimizes the network parameters, and continues training until the loss function stabilizes and stops training.

[0123] Table 1 shows the accuracy distribution of the predicted incidence rate of the trained model verified by the validation set compared with the actual calculated incidence rate statistics.

[0124] Table 1 Age distribution of the statistical accuracy of incidence prediction

[0125] 20-30 years old 30-40 years old 40-50 years old 50-60 years old 60-70 years old 70-80 years old 82％ 90％ 96％ 98％ 84％ 78％

[0126] It can be seen that the accuracy rate for the middle-aged group 40-60 years old is higher than that for the young group and the elderly group.

Claims

1. A rapid intelligent prediction method for determining whether a sample is Lewis antigen negative and the incidence rate for non-diagnostic or therapeutic purposes, characterized in that: include:

1. Data Collection and Modeling Steps A. Recruit healthy people and patients to form volunteer groups, divide the groups by age, and conduct the following tests: (1) Extracting whole blood genomic DNA from samples; (2) using whole blood genomic DNA of the sample as a template, using up to 4 real-time fluorescent PCR primers and probes in the following combination: a real-time fluorescent PCR primer and probe of T59G and a mixture of up to 3 real-time fluorescent PCR primers and probes of C314T, T202C, G508A, and T1067A to establish a fluorescent PCR amplification system for TaqMan PCR amplification; (3) Collecting fluorescence signals and exporting data results for amplification result analysis; B. According to the analysis of the amplification results, the base sequence distribution data of the mutations corresponding to the at most 4 real-time fluorescent PCR primers and probes in each age group, as well as the statistics of the incidence rate, are obtained, and artificial intelligence models are respectively constructed to train and learn the incidence rate to obtain the probability prediction of the disease in each age group under each distribution condition; 2. Prediction Steps C. Perform steps (1) to (3) on the subject to be tested, and substitute the obtained distribution data into the artificial intelligence model of the age group corresponding to the subject to be tested to predict the incidence rate, wherein: The amplification result analysis includes that if the amplification result shows any one of the following (a)-(b), the sample is 100% Lewis antigen negative, and if it is (c), the sample has a 1 / 14 probability of being Lewis antigen negative: (a) Homozygous mutation: any one of up to three homozygous mutations except T59G; (b) T59G heterozygous mutation and up to 3 heterozygous mutations other than T59G; (c) Compound heterozygous mutations in which the T59G homozygous mutation is combined with heterozygous mutations in up to three other mutations except T59G.

2. The fast intelligent prediction method according to claim 1, characterized in that: The real-time fluorescent PCR primers and probes are T59G, and any one of C314T, T202C, G508A, and T1067A corresponding PCR primers and probes.

3. The fast intelligent prediction method according to claim 2, characterized in that: The real-time fluorescent PCR primers and probes are selected from T59G, C314T, and any one or any two corresponding PCR primers and probes among T202C, G508A, and T1067A.

4. The fast intelligent prediction method according to claim 2, characterized in that: The analysis of the amplification results includes Gaussian fitting of the HEX signal and the FAM signal of the fluorescence peak emitted by the probe in the mixture to obtain the presence or absence of the spectrum peak of the corresponding probe, and judge the existence of the mutation site based on this.

5. The fast intelligent prediction method according to any one of claims 1 to 4, characterized in that: The TaqMan PCR amplification reaction procedure described in step (2) is: after preliminary denaturation at 95°C for 10 min, run 45 cycles of the following conditions: denaturation at 95°C for 15 s, annealing / extension at 60°C for 1 min; The determination method of the above-mentioned situation (a) is that when only the HEX signal amplification is successful when detecting at most three midpoint sites other than T59G, it can be determined as a homozygous mutation; The method for determining the situation (b) is to detect the T59G heterozygous mutation and when at most three sites other than T59G, both FAM signal and HEX signal are successfully amplified, and determine that at most three sites other than T59G have heterozygous mutations; The method for determining the situation (c) is that when detecting the T59G site, only when the FAM signal is successfully amplified, the T59G site can be determined to be a homozygous mutation, and when detecting up to three sites other than T59G, when both the FAM signal and the HEX signal are successfully amplified, it is a compound heterozygous mutation.

6. The fast intelligent prediction method according to claim 5, characterized in that: The acquisition of the distribution data comprises the following steps: S1 assigns different preset grayscale or RGB color value combinations to the bases A, T, C, and G, respectively, and considers the alleles, and presets one base sequence in each group as the starting sequence and the other base sequence as the relative sequence. The two groups of FUT3 gene sequences of the alleles are pixelated in sequence in the order of starting sequence-relative sequence according to the preset different grayscale or RGB color value combinations, and are converted into a square matrix image, and multiple square matrix images corresponding to multiple volunteers in different age groups are obtained in this way, wherein the base pairs of up to 4 mutation points are additionally represented by two groups, and each group has two identical preset grayscale or preset RGB color value combinations of pixels to represent the overall presence of heterozygous mutations and homozygous mutations respectively; S2 divides the square matrix images into training set and validation set; The construction of the artificial intelligence model includes: S3 builds a convolutional neural network with a residual mechanism (RESNET), inputs the training set into the network, passes it through the fully connected layer at the output, and uses the softmax function for classification; S4 uses the validation set to verify, calculates the loss function, performs backpropagation, optimizes the network parameters, and continues training until the loss function stabilizes and stops training.

7. The fast intelligent prediction method according to claim 6, characterized in that: The morbidity rate includes the morbidity rate of multiple diseases.

8. The fast intelligent prediction method according to claim 7, characterized in that: Various diseases include pancreatic cancer, stomach cancer, intestinal cancer, and liver cancer.

9. A mixture of real-time fluorescent PCR primers and probes for detecting the genotype and incidence prediction of up to 4 mutation sites of the FUT3 gene associated with Lewis antigen negativity, for use in the rapid intelligent prediction method according to any one of claims 1 to 8, characterized in that: The at most 4 mutation sites are selected from the real-time fluorescent PCR primers and probes of T59G and a mixture of at most 3 real-time fluorescent PCR primers and probes of C314T, T202C, G508A, and T1067A, wherein: (1) The real-time fluorescent PCR primers for detecting the T59G mutation site are: 5'CGCAGGTAGGAGAAGAAACACA3' 5'CATGGCGCCGCTGTCT3' The real-time fluorescent PCR probe for detecting the T59G mutation site is: 5'FAM-CTGAAATAGCCGTGCGG-MGB 3' 5'HEX-CTGAAATAGCAGTGCGG-MGB 3' (2) The real-time fluorescence PCR primers for detecting the T202C mutation site are: 5'CATCTCTGAACAGCGGGACAG 3' 5'GCTGTGTGTTTCTTCTCCTACCT 3' The probe for real-time fluorescent PCR to detect the T202C mutation site is: 5'FAM-CATGTCCATAGCAGGAT-MGB 3' 5'HEX-CATGTCCGTAGCAGGA-MGB 3' (3) The real-time fluorescent PCR primers for detecting the C314T mutation site are: 5'AGGCGTGACTTAGGGTTGGA 3' 5'GCCGACCGCAAGGTGTAC 3' The probe for real-time fluorescent PCR to detect the C314T mutation site is: 5'FAM-ACGATGACCGTGTCT-MGB 3' 5'HEX-ATGACCATGTCTGCC-MGB 3' (4) The real-time fluorescent PCR primers for detecting the G508A mutation site are: 5'AGCGCTGGATCTGGTTCAACT 3' 5'GTCTTGGCCGAGAGGTTGAG 3' The probe for real-time fluorescent PCR to detect the G508A mutation site is: 5'FAM-CGCCCTACGGCTG-MGB 3' 5'HEX-TTCACGCCCTACAGC-MGB 3' (5) The real-time fluorescent PCR primers for detecting the T1067A mutation site are: 5'CCATGCCGGCCTCTCA 3' 5'ACTGCAGCAGGAATCCAGGTA 3' The probe for real-time fluorescent PCR to detect the T1067A mutation site is: 5'FAM-CCAAGCCGCTTTG-MGB 3' 5'HEX-AACCAAGCCCGCTATG-MGB 3'.

10. A real-time fluorescent PCR kit for determining whether a sample is Lewis antigen negative and a rapid intelligent prediction method for morbidity, comprising: DNA polymerase, Mg 2+ Buffer, dNTPs, quality control products, water, and real-time fluorescent PCR primers and probes for detecting up to 4 mutation sites, including real-time fluorescent PCR primers and probes for T59G and a mixture of up to 3 real-time fluorescent PCR primers and probes among C314T, T202C, G508A, and T1067A, wherein the real-time fluorescent PCR primers and probes are the real-time fluorescent PCR primers and probes as claimed in claim 9, and the quality control products are quality control products whose genotypes of up to 4 mutation sites of the FUT3 gene are heterozygous.

Images