Search

CN-121975957-A - Flora prediction model for detecting premature risk, kit and application thereof

CN121975957ACN 121975957 ACN121975957 ACN 121975957ACN-121975957-A

Abstract

The invention relates to a flora prediction model for detecting premature birth risk, a kit and application thereof. The related strain of premature birth is obtained through collection and analysis of the data of the early mNGS, a set of optimal primer probe groups are finally determined through screening and optimization of a large number of experiments, namely lactobacillus crispatus, lactobacillus grignard, lactobacillus inertia, lactobacillus jensenii, escherichia coli, enterococcus faecalis, enterobacter aerogenes, group B hemolytic streptococcus, ureaplasma minum, chlamydia trachomatis, diplococcus gonorrhoeae and gardnerella vaginalis, and a fluorescence quantitative PCR method and a premature birth risk calculation model aiming at the 12 flora detection are established. Compared with the existing detection method, the method is simple, convenient and rapid, high in sensitivity and strong in specificity, reduces the risk of false positive and false negative, and achieves the purpose of batch detection. The change of the microbial flora in the reproductive system of the pregnant woman can be detected rapidly in clinic, corresponding treatment measures are taken, the risk of PTB occurrence is reduced, and the method has good application prospect.

Inventors

  • YING HAO
  • JIANG XIANG
  • Mi Yabing
  • XIE HAN
  • JIA YUANHUI
  • MAO XIAOYUAN
  • LI YAN

Assignees

  • 上海市第一妇婴保健院

Dates

Publication Date
20260505
Application Date
20231212

Claims (10)

  1. 1. The method for constructing the flora prediction model for detecting the risk of premature birth is characterized by comprising the following steps: S1, collecting samples to obtain strain category and abundance information of different crowds; s2, identifying specific strain information and abundance of the premature pregnant women; S3, screening strain types with obvious differences through machine learning; S4, designing a primer probe group and determining a threshold value; S5, sequencing strains and constructing a qPCR premature risk prediction model by combining a threshold value; and S6, evaluating the model to verify accuracy.
  2. 2. The method according to claim 1, wherein the sample in step S1 is a genital tract swab sample from a healthy pregnant woman and a premature pregnant woman.
  3. 3. The method according to claim 1, wherein the step S2 of identifying means that the sample data are divided into two parts, one part is 70% as a training set for screening different strains and abundance, and the other part is 30% as a verification set for verifying whether the strains obtained by screening are significantly different in distribution among different people.
  4. 4. The construction method according to claim 1, wherein the machine learning algorithm in the step S3 includes random forest, naive bayes, support vector machine, decision tree, etc., the screening is to select the best marker of the strain marker with higher occurrence frequency according to the feature importance, and based on the screened best marker, four machine learning methods are adopted to conduct risk prediction, and the optimal parameters are selected through a cross-validation mode to conduct model construction, and finally the formed model is validated in a validation set.
  5. 5. The method according to claim 4, wherein the strain markers are Lactobacillus crispatus, lactobacillus gasseri, lactobacillus inertia, lactobacillus jensenii, escherichia coli, enterococcus faecalis, enterobacter aerogenes, group B hemolytic Streptococcus, microureaplasma, chlamydia trachomatis, diplococcus gonorrhoeae, and Gardnerella vaginalis, respectively.
  6. 6. The construction method according to claim 1, wherein the threshold in step S4 is 0.7/0.3, the risk value is greater than or equal to 0.7 and is high, the risk value is between 0.3 and 0.7 and is low, and the risk value is less than or equal to 0.3.
  7. 7. A model for flora prediction for detecting risk of premature birth, characterized in that the model is constructed by the construction method according to any one of claims 1-6.
  8. 8. A kit for detecting a pre-term risk flora, comprising a detection reagent for detecting a marker of a species as defined in claim 5, or applying the model as defined in claim 7.
  9. 9. The kit of claim 8, wherein the detection kit further comprises primers and probes for detecting the strain markers of claim 5.
  10. 10. Use of a model according to any one of claims 8 and 9 for the preparation of a kit for detecting a pre-term risk flora prediction.

Description

Flora prediction model for detecting premature risk, kit and application thereof Technical Field The invention relates to the field of biotechnology, in particular to a flora prediction model for detecting premature birth risk, a kit and application thereof. Background Premature birth refers to a live infant born before 37 weeks of gestation is completed, as defined by the world health organization. Premature birth can be classified into ultra premature birth (less than 28 weeks), extremely premature birth (28 to 32 weeks), and mid-late premature birth (32 to 37 weeks) according to gestational age. The detection of the risk of premature delivery mainly comprises the following methods of ① symptom and cervical length joint evaluation, wherein regular uterine contraction occurs at least once for 10 minutes from 28 weeks to less than 37 weeks of gestation, cervical canal shortening is accompanied, and the premonitory premature delivery can be diagnosed. Regular uterine contractions (20 min no less than 4 times for no less than 30 seconds) occur between 28 weeks and less than 37 weeks of gestation with cervical shortening no less than 75%, cervical dilatation over 2cm, diagnosed as premature labor, and ② fetal fibronectin (fFN) detection, which is a fetal fibronectin, is a matrix component outside chorionic cells of uterus, is glycoprotein, exists between chorion and decidua, is mainly produced by trophoblast cells, and exists in amniotic fluid in a high concentration. Levels of fFN in cervical-vaginal secretions between 22 and 35 weeks of gestation have a great correlation with the occurrence of premature labor. ③ Phosphorylated insulin-like growth factor binding protein (phIGFBP-1) phosphorylated insulin-like growth factor binding protein (phIGFBP-1) is produced by placental decidua cells, which are released into the cervical vaginal fluid upon injury to decidua tissue cells. IGFBP-1 diagnostic premature rupture has higher sensitivity and specificity than conventional methods. With the development of the technology level, more new technologies are added to the army of premature birth risk monitoring. For example, "use of gene markers in predicting preeclampsia risk in pregnant women" in the university institute is to evaluate preeclampsia risk using a noninvasive method, and preeclampsia gene markers including a set of mRNA markers and lncRNA markers can be used as detection targets for the prediction of preeclampsia diseases. In addition, entry of microorganisms into the uterus is also believed to be an important cause of premature birth, including disruption of maternal immune balance, spontaneous premature birth, and/or premature rupture of the membranes by release of microbial products (e.g., collagenases, proteases, or toxins) that disrupt membrane integrity. Various studies have shown that an imbalance in genital microbiota leads to a serious risk of premature delivery, and that a high diversity of vaginitis-associated vaginal microbiota is also associated with HPV infection, cervical dysplasia, fertilization failure, abortion and premature delivery. The female genital tract consists of vagina, cervix, uterus, fallopian tubes and ovaries, with the cervix connecting the upper genital tract and vagina. As one of the most prominent bacterial stores in the human body, the effects of vaginal flora on female physiological and reproductive health are apparent. The vaginal microbiota of healthy non-pregnant women consists of a variety of anaerobic and aerobic bacteria, the most important of which is lactic acid bacteria. A sequencing study based on the 16S rRNA gene revealed 5 microbiota composition types (Community STATE TYPE, CST) from healthy female vaginas and were designated CST I-V, respectively, the dominant species were Lactobacillus crispatus, lactobacillus gasseri, lactobacillus inertia, lactobacillus jensenii, and a variety of strict anaerobes, respectively. Lactobacillus (Lactobacillus) is usually the most abundant taxonomic group in the Vaginal Microbiome (VMB), producing lactic acid and bacteriocins to inhibit microorganisms associated with dysbiosis, maintain homeostasis, and reduce disease risk. Lactobacillus-based VMB is a hallmark of female reproductive health. VMB in late gestation tends to be more stable and lactobacillus dominated than in early gestation, which may be an evolutionary selection mechanism that ensures success of pregnancy. VMBs, which are based on other taxa, such as VMBs associated with Bacterial Vaginosis (BV) are generally considered undesirable and associated with higher risks of adverse health, including increased risk of bacterial, viral and parasitic transmission infections (STIs), and adverse pregnancy outcomes, including but not limited to premature birth (Preterm Birth, PTB). Inflammation caused by microorganisms invading the upper genital tract may lead to poor pregnancy outcomes. Disruption of the dynamic balance of the microflora structure of the female reproduct