Construction method for predicting lesion model based on HPV infection rate

Abstract

The invention belongs to the field of biological information, particularly relates to a construction method for predicting a lesion model based on an HPV infection rate, and more particularly relates to a construction method, device and equipment for predicting a lesion model based on a single infection rate and a co-infection rate of an HPV. The invention provides a construction method for predicting a lesion model based on an HPV infection rate, and the method comprises the steps: obtaining training data, which comprises an HPV typing detection result and a detection result; determining a single infection rate and a co-infection rate according to an HPV typing detection result; according to the HPV typing detection result and the detection result, determining the relationship between the single infection rate and the detection result as well as between the co-infection rate and the detection result; and constructing a model for predicting lesion based on the HPV infection rate according to the relationship with the detection result. By means of the model building method, a prediction model can be built, and lesions such as TCT detection, colposcope detection results and cervical biopsy results can be predicted through the single infection rate and the co-infection rate.

Description

Technical Field

[0001] The present invention belongs to the field of bioinformatics, specifically, to a method for constructing a lesion prediction model based on HPV infection rate, more specifically, to a method for constructing a lesion prediction model based on HPV single infection rate and co-infection rate, and its device and equipment. Background Art

[0002] Cervical cancer is the most common human papillomavirus (HPV)-related disease to date, and more than 90% of cervical cancer cases can be attributed to persistent infection with high-risk HPV. HPV infection is common in the female reproductive tract, but 70% to 90% of infections are asymptomatic and disappear within 1 to 2 years. 5% to 10% of infected women will have persistent HPV infection and may further progress to cervical precancerous lesions or even invasive cancer. After HPV infects the cervical epithelium, its risk of disease is significantly different due to its different genotypes, whether it is persistently infected, and the duration of persistent infection. Therefore, it is of great clinical significance to perform HPV typing tests and determine whether there is persistent infection.

[0003] At present, more than 200 HPV variants have been detected. In addition to single genotype infections, a large number of patients show mixed infections of more than one genotype. Such multi-genotype HPV infections are common in clinical practice, and their virulence has been shown to be relatively strong, which has a negative impact on the prognosis of HPV-related diseases. However, the arrangement of different genotypes presents more possibilities and hypotheses, and the study of multi-genotype HPV co-infection is destined to be more complicated than single genotype infection, and valuable research in this area is relatively insufficient.

[0004] Therefore, there is a need in the art for a method for constructing a model for predicting lesions, which can predict lesions, such as TCT test results, colposcopy test results, and cervical biopsy results, through single infection rates and co-infection rates. Summary of the invention

[0005] In view of this, in a first aspect, the present invention provides a method for constructing a lesion prediction model based on HPV infection rate, the method comprising:

[0006] Obtain training data, including HPV typing test results, and test results;

[0007] Based on the HPV typing test results, the single infection rate and co-infection rate were determined;

[0008] Based on the HPV typing test results and test results, determine the relationship between the single infection rate and the co-infection rate and the test results;

[0009] A model for predicting lesions based on HPV infection rate was constructed according to the relationship with the test results.

[0010] Using the model building method of the present invention, a prediction model can be constructed, which can predict lesions, such as TCT test results, colposcopy test results and cervical biopsy results, through single infection rate and co-infection rate.

[0011] Furthermore, the test results include: TCT test results, colposcopy test results and cervical biopsy results.

[0012] Furthermore, the single infection rate is calculated by the total infection rate.

[0013] Furthermore, TCT test results include low-grade squamous intraepithelial lesion (LSIL), high-grade squamous intraepithelial lesion (HSIL), squamous cell carcinoma (SCC), atypical glandular cells (AGC), atypical glandular cells prone to neoplasia (AGC-FN), adenocarcinoma in situ (AIS), and adenocarcinoma (ADCA).

[0014] Furthermore, colposcopy results include low-grade lesions, high-grade lesions, suspected invasive cancer, and cancer.

[0015] Furthermore, cervical biopsy results include low-grade lesions (CIN1) and high-grade lesions (CIN2+).

[0016] Furthermore, the HPV infection rate includes the infection rates of 23 HPV genotypes, namely, 6, 11, 16, 18, 26, 31, 33, 35, 39, 42, 43, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, 81, and 82.

[0017] Furthermore, the relationship between the total infection rate and the co-infection rate is as follows:

[0018] Co-infection rate i =0.3737*total infection rate i +0.0008, R 2 =0.9763.

[0019] Furthermore, the relationship between the single infection rate and the co-infection rate is as follows:

[0020] Overall infection rate i =Co-infection rate i +Single infection rate i .

[0021] Furthermore, the co-infection rate C of each pair of genotypes ij The relationship satisfied with the single infection rate of each of the two genotypes is as follows:

[0022] C ij=0.0295*I i *I j +0.0072,R 2 =0.9659.

[0023] Furthermore, the relationship between low-grade squamous intraepithelial lesion (LSIL) and single infection rate and co-infection rate is as follows:

[0024] Sensitivity LSIL SIi =0.5999*single infection rate i -0.0032,R 2 =0.9414;

[0025] Sensitivity LSIL CIi =0.4127*co-infection rate i +0.0135,R 2 =0.8735;

[0026] Specificity SIi =-0.7442*single infection rate i +1.0048, R 2 =0.9929;

[0027] Specificity CIi =-0.2274*co-infection rate i +0.9954, R 2 =0.9628.

[0028] Furthermore, the relationship between high-grade squamous intraepithelial lesions (HSIL) and above (squamous cell carcinoma (SCC), atypical glandular cells (AGC), atypical glandular cells prone to neoplasia (AGC-FN), adenocarcinoma in situ (AIS), adenocarcinoma (ADCA)) and the single infection rate and co-infection rate is as follows:

[0029] Sensitivity SI i =0.7892*single infection rate i-0.0135, R2=0.7;

[0030] Sensitivity CI i =0.4127*co-infection rate i+0.0135, R2=0.8735.

[0031] Furthermore, the relationship between low-grade lesions and single infection rate and co-infection rate is as follows:

[0032] Sensitivity LG-SIi =0.6136*single infection rate i -0.0078, R2 =0.9804;

[0033] Specificity SIi =-0.6758*single infection rate i +1.0121, R 2 =0.8223;

[0034] Sensitivity LG-CIi =0.4057*co-infection rate i +0.008, R 2 =0.9644;

[0035] Specificity CIi =-0.4111*co-infection rate i +0.9964, R 2 =0.9578.

[0036] Furthermore, the relationship between high-grade lesions and above (suspicious invasive carcinoma, cancer) and the single infection rate and co-infection rate is as follows:

[0037] Sensitivity HG-SIi =0.6252*single infection rate i -0.0065, R 2 =0.986;

[0038] Sensitivity HG-CIi =0.3133*co-infection rate i +0.0089, R 2 =0.9006.

[0039] Furthermore, the relationship between CIN1 and the single infection rate and co-infection rate is as follows:

[0040] Sensitivity CIN1 SI i =0.5399*single infection rate i -0.0045, R 2 =0.9474;

[0041] Sensitivity CIN1 CI i =0.4306*co-infection rate i +0.0101, R 2 =0.8991;

[0042] Specificity SI i =-0.6021*single infection rate i +1.006, R 2 =0.9725;

[0043] Specificity CI i =-0.3272*co-infection rate i +0.9914, R 2 =0.9366.

[0044] Furthermore, the relationship between CIN2+ and the single infection rate and co-infection rate is as follows:

[0045] Sensitivity CIN2+SI i =0.7787*single infection rate i -0.0216, R 2 =0.7773;

[0046] Sensitivity CIN2+CI i =0.4606*co-infection rate i +0.0008, R 2 =0.9321.

[0047] In a second aspect, the present invention provides a device for constructing a lesion prediction model based on HPV infection rate, comprising:

[0048] A data acquisition module is used to obtain training data, including HPV typing test results and test results;

[0049] An infection rate determination module, used to determine the single infection rate and co-infection rate according to the HPV typing test results;

[0050] An analysis module for determining the relationship between the single infection rate and the co-infection rate and the test results based on the HPV typing test results and the test results;

[0051] A building module is used to construct a model for predicting lesions based on HPV infection rate according to the relationship with the test results.

[0052] Furthermore, the test results include: TCT test results, colposcopy test results and cervical biopsy results.

[0053] Furthermore, the single infection rate is calculated by the total infection rate.

[0054] Furthermore, TCT test results include low-grade squamous intraepithelial lesion (LSIL), high-grade squamous intraepithelial lesion (HSIL), squamous cell carcinoma (SCC), atypical glandular cells (AGC), atypical glandular cells prone to neoplasia (AGC-FN), adenocarcinoma in situ (AIS), and adenocarcinoma (ADCA).

[0055] Furthermore, colposcopy results include low-grade lesions, high-grade lesions, suspected invasive cancer, and cancer.

[0056] Furthermore, cervical biopsy results include low-grade lesions (CIN1) and high-grade lesions (CIN2+).

[0057] Furthermore, the HPV infection rate includes the infection rates of 23 HPV genotypes, namely, 6, 11, 16, 18, 26, 31, 33, 35, 39, 42, 43, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, 81, and 82.

[0058] Furthermore, the relationship between the total infection rate and the co-infection rate is as follows:

[0059] Co-infection rate i =0.3737*total infection rate i +0.0008, R 2 =0.9763.

[0060] Furthermore, the relationship between the single infection rate and the co-infection rate is as follows:

[0061] Overall infection rate i =Co-infection rate i +Single infection rate i .

[0062] Furthermore, the co-infection rate C of each pair of genotypes ij The relationship satisfied with the single infection rate of each of the two genotypes is as follows:

[0063] C ij =0.0295*I i *I j +0.0072,R 2 =0.9659.

[0064] Furthermore, the relationship between low-grade squamous intraepithelial lesion (LSIL) and single infection rate and co-infection rate is as follows:

[0065] Sensitivity LSIL SIi =0.5999*single infection rate i -0.0032,R 2 =0.9414;

[0066] Sensitivity LSIL CIi =0.4127*co-infection rate i +0.0135,R 2 =0.8735;

[0067] Specificity SIi =-0.7442*single infection rate i +1.0048, R 2 =0.9929;

[0068] Specificity CIi =-0.2274*co-infection rate i +0.9954, R 2 =0.9628.

[0069] Furthermore, the relationship between high-grade squamous intraepithelial lesions (HSIL) and above (squamous cell carcinoma (SCC), atypical glandular cells (AGC), atypical glandular cells prone to neoplasia (AGC-FN), adenocarcinoma in situ (AIS), adenocarcinoma (ADCA)) and the single infection rate and co-infection rate is as follows:

[0070] Sensitivity SI i =0.7892*single infection rate i-0.0135, R2=0.7;

[0071] Sensitivity CI i =0.4127*co-infection rate i+0.0135, R2=0.8735.

[0072] Furthermore, the relationship between low-grade lesions and single infection rate and co-infection rate is as follows:

[0073] Sensitivity LG-SIi =0.6136*single infection rate i -0.0078, R 2 =0.9804;

[0074] Specificity SIi =-0.6758*single infection rate i +1.0121, R 2 =0.8223;

[0075] Sensitivity LG-CIi =0.4057*co-infection rate i +0.008, R 2 =0.9644;

[0076] Specificity CIi =-0.4111*co-infection rate i +0.9964, R 2 =0.9578.

[0077] Furthermore, the relationship between high-grade lesions and above (suspicious invasive carcinoma, cancer) and the single infection rate and co-infection rate is as follows:

[0078] Sensitivity HG-SIi =0.6252*single infection rate i -0.0065, R 2 =0.986;

[0079] Sensitivity HG-CIi =0.3133*co-infection rate i +0.0089, R 2 =0.9006.

[0080] Furthermore, the relationship between CIN1 and the single infection rate and co-infection rate is as follows:

[0081] Sensitivity CIN1 SI i =0.5399*single infection rate i -0.0045, R 2 =0.9474;

[0082] Sensitivity CIN1 CI i =0.4306*co-infection rate i +0.0101, R 2 =0.8991;

[0083] Specificity SI i =-0.6021*single infection rate i +1.006, R 2 =0.9725;

[0084] Specificity CI i =-0.3272*co-infection rate i +0.9914, R 2 =0.9366.

[0085] Furthermore, the relationship between CIN2+ and the single infection rate and co-infection rate is as follows:

[0086] Sensitivity CIN2+SI i =0.7787*single infection rate i -0.0216, R 2 =0.7773;

[0087] Sensitivity CIN2+CI i =0.4606*co-infection ratei +0.0008, R 2 =0.9321.

[0088] In a third aspect, the present invention provides a method or device for constructing a lesion prediction model based on HPV infection rate as described above, and its use in preparing a kit or device for predicting a lesion prediction model based on HPV infection rate.

[0089] In a fourth aspect, the present invention provides a device, comprising:

[0090] at least one processor; and

[0091] a memory communicatively connected to at least one of the processors; wherein,

[0092] The memory stores instructions that can be executed by the processor, and the instructions are used to be executed by the processor to implement any of the above-mentioned methods for constructing a lesion prediction model based on HPV infection rate.

[0093] In some embodiments, the device further includes at least one input device and at least one output device; in the device, the processor, memory, input device, and output device are connected via a bus.

[0094] In a fifth aspect, a storage medium is provided, wherein the storage medium stores computer instructions, and the computer instructions are used to be executed by the computer to implement any of the above-mentioned methods for constructing a lesion prediction model based on HPV infection rate.

[0095] In some embodiments, the storage medium is a computer-readable storage medium. BRIEF DESCRIPTION OF THE DRAWINGS

[0096] Figure 1 A schematic flow chart of the construction method of the present invention;

[0097] Figure 2 The single infection rate and co-infection rate of 23 HPV genotypes in the sample;

[0098] Figure 3 is the co-infection rate C of HPV genotypes i and j ij The product of the infection rates of the two genotypes I i *I j relationship;

[0099] Figure 4 The relationship between the sensitivity and specificity of TCT for detecting LSIL lesions and the infection rate of each genotype in positive samples with single infection and co-infection of non-16 / 18 high-risk HPV genotypes. DETAILED DESCRIPTION

[0100] The present invention will be described in detail below in conjunction with specific embodiments and examples, and the advantages and various effects of the present invention will be more clearly presented. It should be understood by those skilled in the art that these specific embodiments and examples are used to illustrate the present invention, rather than to limit the present invention.

[0101] TCT detection classification: no intraepithelial lesion or malignancy (NILM), normal or inflammatory cells, atypical squamous cells, undetermined significance (ASC-US), atypical squamous cells, high-grade squamous intraepithelial lesions (ASC-H) cannot be excluded, low-grade squamous intraepithelial lesions (LSIL), high-grade squamous intraepithelial lesions (HSIL), squamous cell carcinoma (SCC), atypical glandular cells (AGC), atypical glandular cells prone to neoplasia (AGC-FN), adenocarcinoma in situ (AIS), adenocarcinoma (ADCA).

[0102] Classification of colposcopy: The criteria for colposcopy are normal, benign, or abnormal. Abnormalities include low-grade lesions, high-grade lesions, suspected invasive cancer, cancer, etc.

[0103] Cervical biopsy classification: Cervical histopathological diagnosis is divided into normal group, low-grade lesions (CIN1), high-grade lesions (CIN2 / CIN3), adenocarcinoma in situ, microinvasive carcinoma and invasive carcinoma. CIN2+ refers to high-grade lesions (CIN2 / CIN3), adenocarcinoma in situ, microinvasive carcinoma and invasive carcinoma.

[0104] Single infection rate P: the number of samples N that are positive for only genotype i Si Divide by the total number of samples in the test N:P S =N Si / N

[0105] Co-infection rate C: the number of samples N that are simultaneously positive for at least two HPV genotypes ci Divide by the total number of samples in the test N: P C =N ci / N. Co-infection rate of a pair of HPV genotypes i and j (C ij ) is calculated as the number of samples that test positive for both type i and type j (N i∩j ) divided by the total number of samples (N): C ij =N i∩j / N

[0106] Sensitivity of single infection = number of samples of a certain type of lesions detected in positive cases of single infection of a certain HPV genotype / number of all samples of the same type of lesions detected.

[0107] Sensitivity of co-infection = number of samples of a certain type of lesions detected in positive cases of co-infection of a certain HPV genotype with other high-risk genotypes / number of all samples detected with the same type of lesions.

[0108] The specificity of single infection = the number of samples without detection of a certain type of lesions in negative cases of non-single infection of a certain HPV genotype / the number of all samples without detection of a certain type of lesions.

[0109] Co-infection specificity = number of samples without detection of a certain type of lesions in negative cases of co-infection of a certain HPV genotype with other high-risk genotypes / number of all samples without detection of a certain type of lesions.

[0110] An exemplary process of the present invention is as follows Figure 1 shown.

[0111] Example 1. Research objects and methods used in the present invention

[0112] 1. Research subjects

[0113] All clinical observation cases were derived from free examinations for women of appropriate age in the Changsha Health and People's Livelihood Project in 2023. This study was approved by the Medical Ethics Review Committee of Changsha Maternal and Child Health Hospital (EC-20240308-12). All included subjects were informed of the research content and signed informed consent before the examination. Inclusion criteria: The study subjects included women aged 35-64 years living in Changsha, with a history of sexual life, and voluntarily underwent gynecological examinations. Those with any of the following conditions were excluded: (1) menstrual period; (2) acute genital tract inflammation, sexual intercourse or vaginal douching and vaginal medication within 48 hours before sampling; (3) history of cervical precancerous lesions and cervical cancer; (4) other genital malignancies.

[0114] 2. Research Methods

[0115] Use a disposable sterile cervical sampler (including cervical brush and cell preservation solution) to collect samples in the patient's cervical transformation zone. Transport samples in a sealed box with ice or foam for no more than 5 days. Use a magnetic bead method nucleic acid extraction or purification kit (Shengxiang Biotechnology) to extract nucleic acids. Use a human papillomavirus nucleic acid typing detection kit (PCR fluorescent probe method, Shengxiang Biotechnology) to qualitatively detect 23 HPV genotypes: 6, 11, 16, 18, 26, 31, 33, 35, 39, 42, 43, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, 81, 82. Use a PCR amplification instrument for fluorescent quantitative PCR amplification. Negative results should have no amplification curve (no Ct) in the target area or a Ct value > 39. Positive results should have a Ct value ≥ 39 for the corresponding target gene.

[0116] According to the Chinese cervical cancer screening guidelines

[18] , the screening process is as follows:

[0117] 1) If the result is positive for HPV16 / 18, colposcopy should be performed directly;

[0118] 2) If the result is positive for other high-risk HPV types, TCT examination shall be performed again. If the TCT result of the research subject is ASC-US, ASC-H, LSIL, HSIL, SCC, AGC, atypical endocervical glandular cells prone to neoplasia, endocervical carcinoma in situ, or adenocarcinoma, colposcopy shall be performed;

[0119] 3) If any suspicious or abnormal conditions are found during the colposcopy, free histopathological examination will be provided, and timely treatment and follow-up will be provided to subjects with abnormalities in the histopathological examination.

[0120] 3. Data Analysis

[0121] 1) TCT test results They are summarized as NILM, ASC-US, LSIL and HSIL+, where HSIL+ is HSIL and SCC, AGC, AGC-FN, AIS and ADCA. For cases with both HPV and TCT results, the number of TCT results of positive and negative samples of single infection of each HPV genotype and co-infection with other high-risk genotypes was counted, and the sensitivity and specificity of detecting LSIL and HSIL+ were calculated to compare the difference between single infection and co-infection.

[0122] 2) Colposcopy results It is divided into three categories: no lesion found, low-grade lesion, and high-grade lesion. For cases with both HPV and colposcopy results, the number of colposcopy results of positive and negative samples of single infection of each HPV genotype and co-infection with other high-risk genotypes was counted, and the sensitivity and specificity of detecting low-grade and high-grade lesions were calculated to compare the difference between single infection and co-infection.

[0123] 3) Cervical biopsy results It is divided into normal, CIN1, CIN2+ and other benign abnormalities. For cases with both HPV and cervical biopsy, the number of samples of cervical biopsy results of positive and negative samples of single infection of each HPV genotype and co-infection with other high-risk genotypes was counted, and the sensitivity and specificity of detecting CIN1 and CIN2+ were calculated to compare the difference between single infection and co-infection.

[0124] Example 2. Relationship between HPV single infection rate, co-infection rate and total infection rate

[0125] Among the 158,565 HPV typing results, 20,244 cases were single-infection with only one genotype positive, with a single infection rate of 12.77%, and another 6,935 cases were positive for two or more genotypes, with a co-infection rate of 4.37%. Figure 2The total infection rate of single infection and co-infection of 23 HPV genotypes is shown. The average co-infection rate of each genotype is 49.94%, and the 95% CI is 47.72%-52.15%. Among them, the co-infection rate of HPV52 with the highest infection rate is the lowest at 35.8%, while the co-infection rate of HPV26 with the lowest infection rate is the highest at 61.54%.

[0126] Figure 3 The co-infection rate of each pair of genotypes between 23 HPVs is shown in Figure 2. ij The product of the infection rates of the two genotypes I i *I j The proportional linear relationship, C ij =0.0295*I i *I j +0.0072,R 2 =0.9659. This indicates that the co-infection rate of two genotypes infected at the same time is basically consistent with the calculation of probability.

[0127] Example 3: TCT cytology test results and their relationship with HPV typing infection rate

[0128] TCT cytology test results

[0129] A total of 19,707 non-HPV16 / 18 high-risk positive cases underwent TCT cytology testing. Among all samples, 13,384 were NILM (67.9%). All other HPV high-risk positive cases except NILM were recommended for colposcopy, that is, a total of 6,323 non-HPV16 / 18 high-risk positive cases (32.1%) were recommended for colposcopy. Table 1 lists the TCT results of 19,707 non-HPV16 / 18 high-risk genotype single infection and co-infection positive samples.

[0130] Table 1

[0131]

[0132]

[0133] Relationship between the single infection rate / co-infection rate of each genotype and the sensitivity / specificity of TCT in detecting LSIL

[0134] Figure 4 The results showed that the sensitivity and specificity of TCT for detecting LSIL lesions in positive samples with single infection or co-infection of non-16 / 18 high-risk HPV genotypes were linearly correlated with the infection rate of each genotype.

[0135] The sensitivity of detecting LSIL for single infection or co-infection is very well proportional to the infection rate of HPV genotype i:

[0136] Sensitivity LSIL SIi =0.5999*single infection rate i -0.0032,R 2 =0.9414;

[0137] Sensitivity LSIL CIi =0.4127*co-infection rate i +0.0135,R 2 =0.8735.

[0138] The specificity of single infection or co-infection shows a very good negative linear relationship with the infection rate of HPV genotype i:

[0139] Specificity SIi =-0.7442*single infection rate i +1.0048, R 2 =0.9929;

[0140] Specificity CIi =-0.2274*co-infection rate i +0.9954, R 2 =0.9628.

[0141] It can be seen that the risk of TCT detecting LSIL lesions based on single infection or co-infection of each HPV genotype is mainly related to the infection rate of the genotype, and the slope of the linear relationship between the sensitivity and specificity of single infection or co-infection and the infection rate is slightly different.

[0142] Relationship between the single infection rate / co-infection rate of each genotype and the sensitivity / specificity of TCT in detecting HSIL+

[0143] HSIL+ is HSIL and other results (SCC, AGC, AGC-FN, AIS and ADCA). The sensitivity of TCT for detecting HSIL+ lesions in single-infection or co-infection positive samples is also well proportional to the infection rate of HPV genotype i:

[0144] Sensitivity SI i=0.7892*single infection rate i-0.0135, R2=0.7;

[0145] Sensitivity CI i=0.4127*co-infection rate i+0.0135, R2=0.8735.

[0146] It can be seen that the risk of TCT detecting HSIL+ lesions based on single infection and co-infection of each HPV genotype is mainly related to the infection rate of this genotype.

[0147] Example 4. Colposcopy test results and their relationship with HPV typing infection rate

[0148] Colposcopy results

[0149] A total of 7539 cases underwent colposcopy, of which 4249 (56.36%) had low-grade lesions, 2479 (32.88%) had high-grade lesions, and 811 (10.76%) had no lesions. Table 2 lists the colposcopy results of 7539 positive samples for single infection and co-infection of 18 high-risk HPV genotypes.

[0150] Table 2

[0151]

[0152] Relationship between the single infection rate / co-infection rate of each genotype and the sensitivity / specificity of colposcopy for detecting low-grade lesions and high-grade lesions and above

[0153] The sensitivity and specificity of colposcopy for detecting low-grade lesions in samples with single infection and co-infection of 18 high-risk HPV genotypes showed a good linear relationship with the infection rate of each genotype.

[0154] The sensitivity of detecting low-grade lesions by single infection is proportional to the infection rate of HPV genotype i: Sensitivity LG-SIi =0.6136*single infection rate i -0.0078, R 2 =0.9804;

[0155] The sensitivity of detecting high-grade lesions by single infection is proportional to the infection rate of HPV genotype i: Sensitivity HG-SIi =0.6252*single infection rate i -0.0065, R 2 =0.986;

[0156] The specificity of a single infection also shows a very good negative linear relationship with the infection rate of the genotype. SIi =-0.6758*single infection rate i +1.0121, R 2 =0.8223.

[0157] The sensitivity of co-infection for detecting low-grade lesions is proportional to the infection rate of HPV genotype i: sensitivity LG-CIi =0.4057*co-infection rate i +0.008, R 2 =0.9644;

[0158] The sensitivity of co-infection in detecting high-grade lesions is proportional to the infection rate of HPV genotype i: Sensitivity HG-CIi =0.3133*co-infection rate i +0.0089, R 2 =0.9006;

[0159] The specificity of co-infection also showed a very good negative linear relationship with the infection rate of the genotype. CIi =-0.4111*co-infection rate i +0.9964, R 2 =0.9578.

[0160] Example 5: Cervical biopsy results and their relationship with HPV typing infection rate.

[0161] Cervical biopsy results

[0162] A total of 4762 samples were diagnosed by cervical histopathology, of which 2194 (46.07%) were normal, 1569 (32.95%) were CIN1, 809 (16.99%) were CIN2+, including 744 CIN2 / CIN3 and 65 Cancer, and 190 (3.99%) were other benign abnormalities such as inflammation. Table 3 lists the cervical biopsy results of 4762 samples with single infection and co-infection positive samples of 18 high-risk HPV genotypes.

[0163] Table 3

[0164]

[0165]

[0166] Relationship between the single infection rate / co-infection rate of each genotype and the sensitivity / specificity of cervical biopsy for detecting CIN1 and CIN2+

[0167] The sensitivity and specificity of cervical biopsy for detecting CIN1 and CIN2+ in samples with single infection and co-infection of 18 high-risk HPV genotypes showed a good linear relationship with the infection rate of each genotype.

[0168] The sensitivity of detecting CIN1 and CIN2+ in a single infection is proportional to the infection rate of HPV genotype i:

[0169] Sensitivity CIN1 SI i =0.5399*single infection rate i -0.0045, R 2 =0.9474;

[0170] SensitivityCIN2+SI i =0.7787*single infection rate i -0.0216, R 2 =0.7773;

[0171] The sensitivity of co-infection detection of CIN1 and CIN2+ is proportional to the infection rate of HPV genotype i:

[0172] Sensitivity CIN1 CI i =0.4306*co-infection rate i +0.0101, R 2 =0.8991;

[0173] Sensitivity CIN2+CI i =0.4606*co-infection rate i +0.0008, R 2 =0.9321;

[0174] The specificity of single infection and co-infection also shows a very good negative linear relationship with the infection rate of the genotype:

[0175] Specificity SI i =-0.6021*single infection rate i +1.006, R 2 =0.9725;

[0176] Specificity CI i =-0.3272*co-infection rate i +0.9914, R 2 =0.9366.

Claims

1. A method for constructing a lesion prediction model based on HPV infection rate, the method comprising: Obtain training data, including HPV typing test results, and test results; Based on the HPV typing test results, the single infection rate and co-infection rate were determined; Based on the HPV typing test results and test results, determine the relationship between the single infection rate and the co-infection rate and the test results; A model for predicting lesions based on HPV infection rate was constructed according to the relationship with the test results.

2. The construction method according to claim 1, characterized in that: The test results include: TCT test results, colposcopy test results and cervical biopsy results.

3. The construction method according to claim 2, characterized in that: The results of TCT examination include low-grade squamous intraepithelial lesions, high-grade squamous intraepithelial lesions, squamous cell carcinoma, atypical glandular cells, atypical glandular cells, tendency to neoplasia, adenocarcinoma in situ, and adenocarcinoma; the results of colposcopy examination include low-grade lesions, high-grade lesions, suspected invasive carcinoma, and cancer; the results of cervical biopsy include low-grade lesions and high-grade lesions.

4. The construction method according to claim 1, characterized in that: The HPV infection rate includes the infection rates of HPV genotypes 6, 11, 16, 18, 26, 31, 33, 35, 39, 42, 43, 45, 51, 52, 53, 56, 58, 59, 66, 68, 73, 81, and 82.

5. The construction method according to claim 1, characterized in that: The co-infection rate C of each pair of genotypes ij The relationship satisfied with the single infection rate of each of the two genotypes is as follows: C ij =0.0295*I i *I j +0.0072,R 2 =0.9659。 6. The construction method according to claim 1, characterized in that: The relationship between single infection rate and co-infection rate and test results is as follows: Low-grade squamous intraepithelial lesion: Sensitivity LSILSIi =0.5999*single infection rate i -0.0032,R 2 =0.9414; Sensitivity LSILCIi =0.4127*co-infection rate i +0.0135,R 2 =0.8735; Specificity SIi =-0.7442*single infection rate i +1.0048, R 2 =0.9929; Specificity CIi =-0.2274*co-infection rate i +0.9954, R 2 =0.9628; High-grade squamous intraepithelial lesion and above: Sensitivity SIi =0.7892*single infection rate i-0.0135, R2=0.7; Sensitivity CIi =0.4127*co-infection rate i+0.0135, R2=0.8735; Low-grade lesions: Sensitivity LG-SIi =0.6136*single infection rate i -0.0078, R 2 =0.9804; Specificity SIi =-0.6758*single infection rate i +1.0121, R 2 =0.8223; Sensitivity LG-CIi =0.4057*co-infection rate i +0.008, R 2 =0.9644; Specificity CIi =-0.4111*co-infection rate i +0.9964, R 2 =0.9578; High-grade lesions and above: Sensitivity HG-SIi =0.6252*single infection rate i -0.0065, R 2 =0.986; Sensitivity HG-CIi =0.3133*co-infection rate i +0.0089, R 2 =0.9006; CIN1: Sensitivity CIN1SIi =0.5399*single infection rate i -0.0045, R 2 =0.9474; Sensitivity CIN1CIi =0.4306*co-infection rate i +0.0101, R 2 =0.8991; Specificity SIi =-0.6021*single infection rate i +1.006, R 2 =0.9725; Specificity CIi =-0.3272*co-infection rate i +0.9914, R 2 =0.9366; CIN2+: Sensitivity CIN2+SIi =0.7787*single infection rate i -0.0216, R 2 =0.7773; Sensitivity CIN2+CIi =0.4606*co-infection rate i +0.0008, R 2 =0.9321.

7. Use of the method for constructing a lesion prediction model based on HPV infection rate according to any one of claims 1 to 6 in preparing a kit for predicting a lesion prediction model based on HPV infection rate.

8. A device for constructing a lesion prediction model based on HPV infection rate, comprising: A data acquisition module is used to obtain training data, including HPV typing test results and test results; An infection rate determination module, used to determine the single infection rate and co-infection rate according to the HPV typing test results; An analysis module for determining the relationship between the single infection rate and the co-infection rate and the test results based on the HPV typing test results and the test results; A building module is used to construct a model for predicting lesions based on HPV infection rate according to the relationship with the test results.

9. A device comprising: at least one processor; as well as a memory communicatively connected to at least one of the processors; wherein, The memory stores instructions executable by the processor, and the instructions are used to be executed by the processor to implement the method for constructing a lesion prediction model based on HPV infection rate as described in any one of claims 1 to 6.

10. A storage medium storing computer instructions, wherein the computer instructions are used to be executed by the computer to implement the method for constructing a lesion prediction model based on HPV infection rate as described in any one of claims 1 to 6.

Images