Pair of biomarkers for predicting preterm birth

Abstract

To provide biomarker pairs for predicting preterm birth.SOLUTION: Provided is a pair of isolated biomarkers selected from a group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, where the pair of biomarkers exhibits a change in reversal value between pregnant females at risk of preterm birth and term controls. Also provided is a method of determining probability of preterm birth in a pregnant female, the method comprising measuring a reversal value for at least one pair of biomarkers selected from the aforementioned group in a biological sample obtained from the pregnant female.SELECTED DRAWING: None

Description

[Technical Field]

[0001] This application claims the benefit of priority to U.S. Provisional Patent Application No. 62 / 290,796, filed February 3, 2016, U.S. Provisional Patent Application No. 62 / 387,420, filed December 24, 2015, and U.S. Provisional Patent Application No. 62 / 182,349, filed June 19, 2015, the entire contents of each of which are incorporated herein by reference.

[0002] The present invention relates generally to the field of precision medicine, and more specifically to compositions and methods for determining the probability for preterm birth in pregnant women. [Background technology]

[0003] background According to the World Health Organization, an estimated 15 million babies are born prematurely (before 37 weeks of gestation) each year. Premature birth rates are increasing in almost every country with reliable data. World Health Organization; March of Dimes; The Partnership for Maternal, Newborn & Child Health; Save the See Children, Born Too Soon: The Global Action Report on Preterm Birth, ISBN 9789241503433 (2012). An estimated 1 million infants die each year from complications of preterm birth. Globally, premature birth is the number one cause of death in newborns (infants in the first four weeks of life) and the second leading cause of death in children under five, after pneumonia. Many survivors face a lifetime of disabilities, including learning disabilities and vision and hearing problems.

[0004] Across 184 countries with reliable data, preterm birth rates range from 5% to 18% of infants born. Blencowe et al., "National, regional, and worldwide estimates of preterm birth," The Lancet, 9;379(9832):2162-72 (2012). Preterm birth is nonetheless a global problem, with over 60% of preterm births occurring in Africa and South Asia. Countries with the highest numbers include Brazil, India, Nigeria, and the United States. Of the 11 countries with preterm birth rates above 15%, all but two are in sub-Saharan Africa. In the poorest countries, on average, 12% of infants are born too early, compared with 9% in high-income countries. Within countries, poor families are at higher risk. Over three-quarters of premature babies could be saved with affordable, cost-effective care, such as prenatal steroid injections given to pregnant women at risk of preterm labor to strengthen the baby's lungs.

[0005] Infants born preterm are at greater risk than infants born at term for death and a variety of health and developmental problems. Complications include acute respiratory, gastrointestinal, immune, central nervous system, auditory, and visual problems, as well as long-term motor, cognitive, visual, auditory, behavioral, social-emotional, health, and growth problems. The birth of a preterm infant also carries significant emotional and economic costs for families and can have implications for public services, such as health insurance, education, and other social support systems. The greatest risk of death and morbidity is for infants born at the earliest gestational age. However, infants born closer to term represent the largest number of infants born preterm and experience more complications than infants born at term.

[0006] To prevent preterm birth in women less than 24 weeks pregnant who have an open cervix on ultrasound, a surgical procedure known as cervical cerclage can be used to close the cervix with strong sutures.Women less than 34 weeks pregnant and in active preterm labor may require hospitalization and the administration of medications to temporarily stop preterm labor and / or promote fetal lung development.If a pregnant woman is determined to be at risk for preterm labor, health care providers can implement various clinical strategies, which may include preventive oral medications, such as hydroxyprogesterone caproate (Makena) injections and / or vaginal progesterone gel, cervical pessaries, restrictions on sexual activity and / or other physical activities, and changes in treatment for chronic conditions that increase the risk of preterm labor, such as diabetes and hypertension.

[0007] There is a significant need to identify women at risk for preterm birth and provide them with appropriate prenatal care. Women identified as high-risk can be planned for more intensive prenatal surveillance and preventive interventions. Current strategies for risk assessment are based on obstetric and medical histories and clinical examinations, but these strategies only identify a small percentage of women at risk for preterm birth. Currently, a prior history of spontaneous PTB (sPTB) is the single strongest predictor of subsequent PTB. Having experienced one previous sPTB increases the likelihood of a second PTB to 30–50%. Other maternal risk factors include black race, low maternal body mass index, and short cervical length. Studies of amniotic fluid, cervicovaginal fluid, and serum biomarkers to predict sPTB suggest that abnormalities in multiple molecular pathways are present in women who ultimately deliver preterm. Reliable early identification of risk for preterm birth would allow for appropriate monitoring and clinical management planning to prevent preterm birth. Such monitoring and management may include more frequent prenatal visits, serial cervical length measurements, increased education about the signs and symptoms of early preterm labor, lifestyle interventions for modifiable risk behaviors, smoking cessation, cervical pessaries, and progesterone treatment. Finally, reliable prenatal identification of risk for preterm birth is also crucial for the cost-effective allocation of monitoring resources. Despite the intense research being conducted to identify at-risk women, PTB prediction algorithms that are based solely on clinical and demographic factors or that use serum or vaginal biomarkers have not yet been developed into clinically useful tests.To enable clinical intervention, more accurate methods are needed to identify at-risk women during their first pregnancy and early pregnancy.The present invention addresses this need by providing compositions and methods for determining whether a pregnant woman is at risk for preterm birth.Related advantages are also provided. [Prior art documents] [Non-patent literature]

[0008] [Non-Patent Document 1] World Health Organization; March of Dimes; The Partnership for Maternal, Newborn & Child Health; Save the Children, Born Too Soon: The Global Action Report on Preterm Birth, ISBN 9789241503433 (2012) [Non-patent document 2] Blencowe et al., "National, regional and worldwide estimates of preterm birth." The Lancet, 9;379(9832):2162-72 (2012) Summary of the Invention [Problem to be solved by the invention]

[0009] overview The present invention provides compositions and methods for predicting the likelihood of preterm birth in pregnant women.

[0010] The present invention provides isolated biomarkers selected from the group set forth in Table 26. The biomarkers of the present invention can predict the risk of preterm birth in a pregnant woman. In some embodiments, the isolated biomarker is selected from the group consisting of IBP4, SHBG, PSG3, LYAM1, IGF2, CLUS, IBP3, INHBC, PSG2, PEDF, CD14, and APOC3.

[0011] The present invention provides surrogate peptides of isolated biomarkers selected from the group shown in Table 26. In some embodiments, the surrogate peptides of isolated biomarkers are selected from the group of surrogate peptides shown in Table 26. The biomarkers of the present invention and their surrogate peptides can be used in methods for predicting the risk of preterm birth in pregnant women. In some embodiments, the surrogate peptides correspond to isolated biomarkers selected from the group consisting of IBP4, SHBG, PSG3, LYAM1, IGF2, CLUS, IBP3, INHBC, PSG2, PEDF, CD14, and APOC3.

[0012] The present invention provides stable isotope-labeled standard peptides (SIS peptides) corresponding to surrogate peptides selected from the group set forth in Table 26. The biomarkers of the present invention, their surrogate peptides, and SIS peptides can be used in methods for predicting the risk of preterm birth in pregnant women. In some embodiments, the SIS peptides correspond to surrogate peptides of isolated biomarkers selected from the group consisting of IBP4, SHBG, PSG3, LYAM1, IGF2, CLUS, IBP3, INHBC, PSG2, PEDF, CD14, and APOC3.

[0013] The present invention provides pairs of isolated biomarkers selected from the group consisting of the isolated biomarkers listed in Table 26, wherein the pairs of biomarkers exhibit altered ratio values ​​between pregnant women at risk of preterm birth and full-term controls.

[0014] The present invention provides isolated biomarker pairs selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, which biomarker pairs exhibit altered ratio values ​​between pregnant women at risk of preterm birth and full-term controls.

[0015] The present invention provides isolated biomarker pairs selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein the biomarker pairs exhibit altered ratio values ​​between pregnant women at risk of preterm birth and full-term controls.

[0016] In one embodiment, the present invention provides an isolated biomarker pair selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein the biomarker pair exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls.

[0017] In one embodiment, the present invention provides an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein each of said pair of biomarkers exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls.

[0018] In one embodiment, the present invention provides a composition comprising a pair of surrogate peptides corresponding to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein the pair of biomarkers exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the composition comprises stable isotope-labeled standard peptides (SIS peptides) for each of the surrogate peptides.

[0019] In one embodiment, the present invention provides a composition comprising a pair of surrogate peptides corresponding to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein the pair of biomarkers exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the composition comprises stable isotope-labeled standard peptides (SIS peptides) for each of the surrogate peptides.

[0020] In certain embodiments, the present invention provides an isolated biomarker pair IBP4 / SHBG that exhibits an inverse value change among pregnant women at risk of preterm birth compared to full-term controls. In further embodiments, the present invention provides an isolated biomarker pair IBP4 / SHBG that exhibits a higher ratio in pregnant women at risk of preterm birth compared to full-term controls.

[0021] In one embodiment, the present invention provides a composition comprising a pair of surrogate peptides corresponding to the biomarker pair IBP4 / SHBG, wherein the pair of biomarkers exhibits a higher ratio in pregnant women at risk of preterm birth compared to full-term controls. In one embodiment, the composition comprises stable isotope-labeled standard peptides (SIS peptides) for each of the surrogate peptides.

[0022] In a further embodiment, the invention provides a panel of at least two pairs of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, each of which pairs exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the panel comprises stable isotope-labeled standard peptides (SIS peptides) of surrogate peptides derived from each of the biomarkers.

[0023] In a further embodiment, the invention provides a panel of at least two pairs of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, each pair exhibiting inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the panel comprises stable isotope-labeled standard peptides (SIS peptides) of surrogate peptides derived from each of the biomarkers.

[0024] In an additional embodiment, the invention provides a panel of at least two pairs of surrogate peptides, each of which corresponds to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein each of the pairs exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the panel comprises a stable isotope-labeled standard peptide (SIS peptide) for each of the surrogate peptides.

[0025] In additional embodiments, the invention provides a panel of at least two pairs of surrogate peptides, each of which corresponds to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein each of the pairs exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the panel comprises a stable isotope-labeled standard peptide (SIS peptide) for each of the surrogate peptides.

[0026] In a further embodiment, the present invention provides a panel of at least two pairs of surrogate peptides, each of which corresponds to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein at least one of the pairs exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the composition comprises a stable isotope-labeled standard peptide (SIS peptide) for each of the surrogate peptides.

[0027] In a further embodiment, the present invention provides a panel of at least two pairs of surrogate peptides, each corresponding to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein at least one of the pairs exhibits inverse changes in values ​​between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the composition comprises a stable isotope-labeled standard peptide (SIS peptide) for each of the surrogate peptides.

[0028] In an additional embodiment, the present invention provides a panel of at least two pairs of surrogate peptides, each of which corresponds to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein a calculated score derived from the panel of at least two pairs of biomarkers indicates a change in value between pregnant women and full-term controls. In one embodiment, the composition comprises a stable isotope-labeled standard peptide (SIS peptide) for each of the surrogate peptides.

[0029] In additional embodiments, the present invention provides a panel of at least two pairs of surrogate peptides, each of which corresponds to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein a calculated score derived from the panel of at least two pairs of biomarkers indicates a change in value between pregnant women and full-term controls. In one embodiment, the composition comprises a stable isotope-labeled standard peptide (SIS peptide) for each of the surrogate peptides.

[0030] In one embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring the ratio of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman. 2 and 37 kg / m 2 have a body mass index (BMI) equal to or less than

[0031] In one embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring the ratio of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman. 2 and 37 kg / m 2 In some embodiments, the method includes an initial step of obtaining a biological sample. In some embodiments, the method includes detecting, measuring, or quantifying SIS surrogate peptides for each of the biomarkers.

[0032] In some embodiments, determining a pregnant woman's probability of preterm birth involves an initial step comprising forming a probability / risk index by measuring the ratio of isolated biomarkers selected from a group of cohorts of preterm and term pregnancies where gestational age at birth is known. In further embodiments, a preterm birth risk index is formed by measuring the ratio of IBP4 / SHBG in a cohort of preterm and term pregnancies where gestational age at birth has been recorded. In some embodiments, determining a pregnant woman's probability of preterm birth involves measuring the ratio of IBP4 / SHBG and comparing that value to the index to derive the risk of preterm birth, using the same isolation and measurement techniques as used to derive IBP4 / SHBG in the index group.

[0033] In one embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring in a biological sample obtained from the pregnant woman at least one pair of reversible values ​​of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS to determine the pregnant woman's probability of preterm birth. In some embodiments, the pregnant woman has a blood pressure of 22 kg / m 2 and 37 kg / m 2 In some embodiments, the method includes an initial step of obtaining a biological sample. In some embodiments, the method includes detecting, measuring, or quantifying SIS surrogate peptides for each of the biomarkers.

[0034] In one embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring in a biological sample obtained from the pregnant woman at least one pair of reversible values ​​of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 to determine the probability of preterm birth in the pregnant woman. 2 and 37 kg / m 2 In some embodiments, the method includes an initial step of obtaining a biological sample. In some embodiments, the method includes detecting, measuring, or quantifying SIS surrogate peptides for each of the biomarkers.

[0035] In another embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring changes in reversal values ​​of a panel of at least two pairs of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman. In another embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring changes in inversion values ​​of at least two pairs of a panel of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman. In some embodiments, the inversion values ​​reveal a change in the relative intensity of individual biomarkers between the pregnant woman and a full-term control, indicating the probability of preterm birth in the pregnant woman. In additional embodiments, the measuring step comprises measuring surrogate peptides of the biomarkers in a biological sample obtained from the pregnant woman. In some embodiments, a pregnant woman may have a blood sugar level of 22 kg / m 2 and 37 kg / m 2 In some embodiments, the method includes an initial step of obtaining a biological sample. In some embodiments, the method includes detecting, measuring, or quantifying SIS surrogate peptides for each of the biomarkers.

[0036] In one embodiment, the present invention provides a method for determining the likelihood of preterm birth in a pregnant woman, comprising measuring inverted values ​​of a pair of biomarkers consisting of IBP4 and SHBG in a biological sample obtained from the pregnant woman to determine the likelihood of preterm birth in the pregnant woman. 2 and 37 kg / m 2 In some embodiments, the method includes an initial step of obtaining a biological sample. In some embodiments, the method includes detecting, measuring, or quantifying SIS surrogate peptides for each of the biomarkers.

[0037] In one embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring the inverted value of the biomarker pair consisting of the ratio of IBP4 to SHBG (IBP4 / SHBG) in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman, wherein a higher ratio in the pregnant woman compared to full-term controls indicates an increased risk of preterm birth. In a further embodiment, the pregnant woman has a blood glucose level of 22 kg / m 2 and 37 kg / m 2 In some embodiments, the method includes an initial step of obtaining a biological sample. In some embodiments, the method includes detecting, measuring, or quantifying SIS surrogate peptides for each of the biomarkers.

[0038] In one embodiment, the present invention provides a method for determining the likelihood of preterm birth in a pregnant woman, comprising measuring the inverse values ​​of the biomarker pair IBP4 and SHBG in a biological sample obtained from the pregnant woman to determine the likelihood of preterm birth in the pregnant woman. 2 and 37 kg / m 2In some embodiments, the method includes an initial step of obtaining a biological sample. In some embodiments, the method includes detecting, measuring, or quantifying SIS surrogate peptides for each of the biomarkers.

[0039] The present invention also provides a method for detecting an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a pregnant woman, comprising: a. obtaining a biological sample from said pregnant female; b. detecting whether the isolated pair of biomarkers is present in the biological sample by contacting the biological sample with a first capture agent that specifically binds the first member of the pair and a second capture agent that specifically binds the second member of the pair; and detecting binding between a first biomarker of the pair and the first capture agent, and between the second member of the pair and the second capture agent. The present invention provides a method comprising:

[0040] In one embodiment, the present invention provides a method for detecting IBP4 and SHBG in a pregnant woman, comprising: a. obtaining a biological sample from the pregnant woman; b. contacting the biological sample with a capture agent that specifically binds IBP4 and a capture agent that specifically binds SHBG to detect whether IBP4 and SHBG are present in the biological sample; and c. detecting binding between IBP4 and the capture agent and between SHBG and the capture agent. In one embodiment, the method comprises measuring reversal values ​​of the pair of biomarkers. In a further embodiment, the presence of a change in reversal value between the pregnant woman and a full-term control indicates the likelihood of preterm birth in the pregnant woman. In one embodiment, the sample is obtained at 19-21 weeks of gestation. In a further embodiment, the capture agent is selected from the group consisting of an antibody, an antibody fragment, a nucleic acid-based protein-binding reagent, a small molecule, or a variant thereof. In a further embodiment, the method is performed by an assay selected from the group consisting of an enzyme immunoassay (EIA), an enzyme-linked immunosorbent assay (ELISA), and a radioimmunoassay (RIA).

[0041] The present invention also provides a method for detecting an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a pregnant woman, comprising: a. obtaining a biological sample from said pregnant female; and b. detecting whether said isolated biomarker pair is present in said biological sample, comprising subjecting said sample to a proteomic workflow consisting of mass spectrometry quantification. The present invention provides a method comprising:

[0042] In one embodiment, the present invention provides a method for detecting IBP4 and SHBG in a pregnant woman, comprising: a. obtaining a biological sample from said pregnant female; and b. detecting whether said isolated biomarker pair is present in said biological sample, comprising subjecting said sample to a proteomic workflow consisting of mass spectrometry quantification. The present invention provides a method comprising:

[0043] In some embodiments, the inverted values ​​reveal a change in the relative intensity of the individual biomarkers between the pregnant woman and full-term controls, indicating the pregnant woman's likelihood of preterm birth. In additional embodiments, the measuring step comprises measuring surrogate peptides of the biomarkers in a biological sample obtained from the pregnant woman. In one embodiment, a preterm birth risk index is formed by measuring the ratio of IBP4 / SHBG in a cohort of preterm and full-term pregnancies for which gestational age at delivery has been recorded. Clinical practice then compares the measured ratio of IBP4 / SHBG in each individual pregnancy to the index to derive the risk of preterm birth, using the same isolation and measurement techniques as used to derive IBP4 / SHBG in the index group.

[0044] Other features and advantages of the invention will become apparent from the detailed description, and from the claims. In certain embodiments, for example, the following are provided: (Item 1) A composition comprising an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein said pair of biomarkers exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. (Item 2) A composition comprising a pair of surrogate peptides of a pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein said pair of biomarkers exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. (Item 3) 3. The composition of claim 2, further comprising stable isotope-labeled standard peptides (SIS peptides) corresponding to each of the surrogate peptides. (Item 4) A panel of at least two pairs of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein each of the pairs of biomarkers shows an inverse value change between pregnant women at risk of preterm birth and full-term controls. (Item 5) A panel of at least two pairs of surrogate peptides, each of the pairs of surrogate peptides corresponding to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1, wherein each of the pairs of biomarkers shows an inverse change in values ​​between pregnant women at risk of preterm birth and full-term controls. (Item 6) Item 6. The panel of item 5, further comprising a SIS peptide corresponding to each of the surrogate peptides. (Item 7) 1. A method for determining the probability of preterm birth in a pregnant woman, comprising measuring inverted values ​​of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman. (Item 8) 1. A method for determining the probability of preterm birth in a pregnant woman, comprising measuring changes in reversal values ​​of at least two pairs of a panel of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a biological sample obtained from the pregnant woman, thereby determining the probability of preterm birth in the pregnant woman. (Item 9) 9. The method according to item 7 or 8, wherein the presence of a change in inversion value between the pregnant woman and a full-term control is indicative of the probability of preterm birth in the pregnant woman. (Item 10) 9. The method of claim 7 or 8, wherein said measuring comprises measuring a surrogate peptide of said biomarker in said biological sample obtained from said pregnant woman. (Item 11) Item 11. The method of item 10, wherein the measuring further comprises measuring stable isotope-labeled standard peptides (SIS peptides) of each of the surrogate peptides. (Item 12) 9. The method of claim 7 or 8, wherein the probability is expressed as a risk score. (Item 13) 9. The method according to item 7 or 8, wherein the biological sample is selected from the group consisting of whole blood, plasma, and serum. (Item 14) 9. The method according to item 7 or 8, wherein the biological sample is serum. (Item 15) Item 15. The method according to item 14, wherein the sample is obtained at 19 to 21 weeks of gestation. (Item 16) 9. The method according to item 7 or 8, wherein the measuring comprises mass spectrometry (MS). (Item 17) 9. The method according to item 7 or 8, wherein the measuring comprises an assay using a capture agent. (Item 18) 18. The method of claim 17, wherein the capture agent is selected from the group consisting of an antibody, an antibody fragment, a nucleic acid-based protein binding reagent, a small molecule, or a variant thereof. (Item 19) 18. The method of claim 17, wherein the assay is selected from the group consisting of an enzyme immunoassay (EIA), an enzyme-linked immunosorbent assay (ELISA), and a radioimmunoassay (RIA). (Item 20) 9. The method of claim 7 or 8, further comprising an initial step of detecting one or more measurable characteristics of a risk indicator. (Item 21) 18. The method of claim 17, wherein the risk indicator is selected from the group consisting of body mass index (BMI), parity, and fetal sex. (Item 22) 22. The method of item 21, wherein the risk indicator is BMI. (Item 23) 9. The method of claim 7 or 8, further comprising predicting gestational age at birth (GAB) prior to said determining the probability of preterm birth. (Item 24) The pregnant woman is 22 kg / m 2 and 37 kg / m 2 23. The method of claim 22, wherein the subject has a body mass index (BMI) less than or equal to: (Item 25) 10. A method for determining a pregnant woman's likelihood of preterm birth, comprising measuring inverted values ​​of a pair of biomarkers selected from the biomarkers listed in Table 26 in a biological sample obtained from said pregnant woman to determine said pregnant woman's likelihood of preterm birth. (Item 26) A method for determining the probability of preterm birth in a pregnant woman, comprising measuring the inverted values ​​of the biomarker pair IBP4 and SHBG in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman. (Item 27) 27. The method of item 25 or 26, wherein the presence of a change in inversion value between the pregnant woman and a full-term control is indicative of the probability of preterm birth in the pregnant woman. (Item 28) 27. The method of claim 25 or 26, wherein said measuring comprises measuring a surrogate peptide of said biomarker in said biological sample obtained from said pregnant woman. (Item 29) 27. The method of item 25 or 26, wherein the probability is expressed as a risk score. (Item 30) 27. The method of claim 25 or 26, wherein the biological sample is selected from the group consisting of whole blood, plasma, and serum. (Item 31) 27. The method of item 25 or 26, wherein the biological sample is serum. (Item 32) 31. The method according to item 30, wherein the sample is obtained at 19 to 22 weeks of gestation. (Item 33) 27. The method of claim 25 or 26, wherein the measuring comprises mass spectrometry (MS). (Item 34) 27. The method of claim 25 or 26, wherein the measuring comprises an assay using a capture agent. (Item 35) 35. The method of claim 34, wherein the capture agent is selected from the group consisting of an antibody, an antibody fragment, a nucleic acid-based protein binding reagent, a small molecule, or a variant thereof. (Item 36) 36. The method of claim 35, wherein the assay is selected from the group consisting of an enzyme immunoassay (EIA), an enzyme-linked immunosorbent assay (ELISA), and a radioimmunoassay (RIA). (Item 37) 27. The method of claim 25 or 26, further comprising detecting one or more measurable characteristics of a risk indicator. (Item 38) 38. The method of claim 37, wherein the risk indicia are selected from the group consisting of body mass index (BMI), parity, and fetal sex. (Item 39) 39. The method of claim 38, wherein the risk indicator is BMI. (Item 40) 27. The method of claim 25 or 26, further comprising predicting gestational age at birth (GAB) prior to said determining the probability of preterm birth. (Item 41) The pregnant woman is 22 kg / m 2 and 37 kg / m 2 40. The method of claim 39, wherein the subject has a body mass index (BMI) less than or equal to: (Item 42) 27. The method of claim 26, wherein the inverted value is based on the ratio of IBP4 to SHBG (IBP4 / SHBG) to determine the probability of preterm birth in the pregnant woman, wherein a higher ratio in the pregnant woman compared to a full-term control indicates an increased risk of preterm birth. (Item 43) 1. A method for detecting an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a pregnant woman, comprising: a. obtaining a biological sample from said pregnant female; b. detecting whether the isolated pair of biomarkers is present in the biological sample by contacting the biological sample with a first capture agent that specifically binds the first member of the pair and a second capture agent that specifically binds the second member of the pair; and c. detecting binding between a first biomarker of the pair and the first capture agent, and between the second member of the pair and the second capture agent. A method comprising: (Item 44) 44. The method of claim 43, further comprising measuring the inverted values ​​of said pair of biomarkers. (Item 45) 45. The method of claim 44, wherein the presence of a change in inverse value between the pregnant woman and a full-term control is indicative of the probability of preterm birth for the pregnant woman. (Item 46) Item 46. The method of item 45, wherein the probability is expressed as a risk score. (Item 47) 44. The method of claim 43, wherein the biological sample is selected from the group consisting of whole blood, plasma, and serum. (Item 48) 48. The method of item 47, wherein the biological sample is serum. (Item 49) Item 48. The method according to item 47, wherein the sample is obtained at 19 to 21 weeks of gestation. (Item 50) 44. The method of claim 43, wherein the capture agent is selected from the group consisting of an antibody, an antibody fragment, a nucleic acid-based protein binding reagent, a small molecule, or a variant thereof. (Item 51) 44. The method of claim 43, wherein the method is performed by an assay selected from the group consisting of an enzyme immunoassay (EIA), an enzyme-linked immunosorbent assay (ELISA), and a radioimmunoassay (RIA). (Item 52) 45. The method of claim 44, further comprising detecting one or more measurable characteristics of a risk indicator. (Item 53) 53. The method of claim 52, wherein the risk indicia are selected from the group consisting of body mass index (BMI), parity, and fetal sex. (Item 54) 54. The method of item 53, wherein the risk indicator is BMI. (Item 55) The pregnant woman is 22 kg / m 2 and 37 kg / m 2 55. The method of claim 54, wherein the subject has a body mass index (BMI) less than or equal to: (Item 56) 44. The method of claim 43, wherein the isolated biomarker pair is IBP4 / SHBG. (Item 57) 1. A method for detecting IBP4 and SHBG in a pregnant woman, comprising: a. obtaining a biological sample from said pregnant female; b. detecting whether IBP4 and SHBG are present in the biological sample by contacting the biological sample with a capture agent that specifically binds IBP4 and a capture agent that specifically binds SHBG; and c. Detecting binding between IBP4 and the capture agent and between SHBG and the capture agent. A method comprising: (Item 58) 1. A method for detecting an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, IBP4 / PSG3, IBP4 / LYAM1, IBP4 / IGF2, CLUS / IBP3, CLUS / IGF2, CLUS / LYAM1, INHBC / PSG3, INHBC / IGF2, PSG2 / LYAM1, PSG2 / IGF2, PSG2 / LYAM1, PEDF / PSG3, PEDF / SHBG, PEDF / LYAM1, CD14 / LYAM1, and APOC3 / LYAM1 in a pregnant woman, comprising: a. obtaining a biological sample from said pregnant female; and b. A method comprising detecting whether said isolated pair of biomarkers is present in said biological sample, comprising subjecting said sample to a proteomics workflow consisting of mass spectrometry quantification. (Item 59) 1. A method for detecting IBP4 and SHBG in a pregnant woman, comprising: a. obtaining a biological sample from said pregnant female; and b. detecting whether said IBP4 and SHBG are present in said biological sample, comprising subjecting said sample to a proteomic workflow consisting of mass spectrometry quantification. A method comprising: (Item 60) 60. The method of claim 58 or 59, further comprising measuring the inverted values ​​of said pair of biomarkers. (Item 61) 61. The method of claim 60, wherein the presence of a change in inverse value between the pregnant woman and a full-term control is indicative of the probability of preterm birth for the pregnant woman. (Item 62) Item 62. The method of item 61, wherein the probability is expressed as a risk score. (Item 63) 60. The method of claim 58 or 59, wherein the biological sample is selected from the group consisting of whole blood, plasma, and serum. (Item 64) 64. The method of item 63, wherein the biological sample is serum. (Item 65) Item 64. The method according to item 63, wherein the sample is obtained at 19 to 21 weeks of gestation. (Item 66) 60. The method of claim 58 or 59, wherein the proteomics workflow comprises quantifying SIS peptides. (Item 67) 60. The method of claim 58 or 59, further comprising detecting one or more measurable characteristics of a risk indicator. (Item 68) 68. The method of claim 67, wherein the risk indicia are selected from the group consisting of body mass index (BMI), parity, and fetal sex. (Item 69) 69. The method of item 68, wherein the risk indicator is body mass index (BMI). (Item 70) The pregnant woman is 22 kg / m 2 and 37 kg / m 2 70. The method of claim 69, wherein the subject has a body mass index (BMI) less than or equal to: [Brief explanation of the drawings]

[0045] [Figure 1] Blood collection window. Individual inversion performance is shown across the blood collection window. Inversions shown: IBP4 / SHBG; VTNC / VTDB; IBP4 / SHBG; VTNC / SHBG; IBP4 / SHBG; CATD / SHBG; PSG2 / ITIH4; CHL1 / ITIH4; PSG2 / C1QB; PSG2 / FBLN3; HEMO / IBP6; HEMO / PTGDS.

[0046] [Figure 2] Discovery, supporting, and verifying cases of GABD.

[0047] [Figure 3]Protein Expression During Pregnancy. Various proteins can be analyzed based on known protein behavior and knowledge of proteins / pathways that are not affected by preterm birth. Figure 3 shows the expression of pregnancy-associated proteins during pregnancy. These proteins and their networks are not affected by preterm birth pathology at the gestational ages shown.

[0048] [Figure 4] Protein Expression During Pregnancy Figure 4 shows an expanded version of the graph shown in Figure 3 for placenta-specific growth hormone.

[0049] [Figure 5] Protein pathology during pregnancy. Insulin-like growth factor binding protein 4 (IBP4) was overexpressed by at least 10% during the blood collection window of 19-21 weeks. Sex hormone-binding globulin (SHBG) was underexpressed by at least 10% during the blood collection window of 19-21 weeks.

[0050] [Figure 6] Supporting Selection Criteria. Figure 6 describes the criteria for implementing a high performance preterm birth test that is clinically and analytically robust.

[0051] [Figure 7] Monte Carlo Cross Validation (MCVV). MCCV is a conservative method for estimating how well a classifier will perform on an independent set of samples drawn from the same population (e.g., PAPR).

[0052] [Figure 8] Analysis of [IBP4] / [SHBG CHL1 CLUS]. CHL1 and CLUS increased performance by 0.03 relative to IBP4 / SHBG alone.

[0053] [Figure 9]Power and sample size analysis. Power and sample size analysis predicts the likelihood that a study will have sufficient power to reject the null hypothesis (AUC=0.5) at a given sample size and performance estimate threshold.

[0054] [Figure 10] Gestational age and time to delivery. Gestational age can be biochemically determined using multiple analytes that increase during pregnancy but do not differ between PTB cases and controls. Biochemical dating can be useful to confirm dating by the date of the last menstrual period or ultrasound dating, or prior to later determining sPTB risk, TTB, or GAB prediction.

[0055] [Figure 11] Classifier Development. Figure 11 shows the criteria for developing a classifier.

[0056] [Figure 12] Pathway coverage in discovery assays. Figure 12 shows the distribution of proteins by pathway.

[0057] [Figure 13] PCA of the discovery data detected changes across the blood collection window, thus demonstrating that this highly multiplexed assay is sensitive to gestational age.

[0058] [Figure 14] Hierarchical clustering of proteins measured in discovery samples.

[0059] [Figure 15] Placenta-specific protein branches within larger clusters. The right panel lists the module of genes expressed during pregnancy identified by Thompson, and the left panel shows that a discovery serum proteomics assay recapitulates the associated expression of this module (Thompson et al., Genome Res. 12(10):1517-1522 (2002)).

[0060] [Figure 16] Dysregulated Protein PreTRM™ Samples.

[0061] [Figure 17] Sex hormone-binding globulin (SHBG) biology highlighted. SHBG is expressed in placental cells (right). SHBG may be involved in regulating the levels of free testosterone and estrogen levels in the placental fetal compartment (left).

[0062] [Figure 18] Interactions between IBP4, IGF2, PAPP-A, and PRG2. IBP4 is a negative regulator of IGF2. IBP4 is released from IGF2 by PAPPA-mediated proteolysis. Low levels of PAPPA have been implicated in IUGR and PE. High levels of IBP4 suppress IGF2 activity. PTB cases suppress PAPPA and PRG2 levels and increase IBP4 levels.

[0063] [Figure 19] Insulin-like growth factor binding protein 4 (IBP4). IBP4 is upregulated in cases of PTB. IGF2 stimulates EVT proliferation, differentiation, and invasion during early pregnancy. IGF activity is essential for normal placentation and fetal growth. IBP4 mediates autocrine and paracrine regulation of IGF2 activity at the maternal-fetal interface. The activity of IGF2 expressed by cytotrophoblasts is balanced by IBP produced by decidual cells. Elevated IBP4 and decreased IGF2 during the first trimester correlated with placental insufficiency (e.g., IUGR / SGA).

[0064] [Figure 20] MS vs. ELISA correlation for IBP4, SHBG, and CHL1. Mass spectrometry and ELISA for key analytes showed good agreement. Agreement between two orthogonal platforms ensures reliability of analyte measurements.

[0065] [Figure 21] PTB classification by IBP4 / SHBG in discovery samples from 19-21 weeks' gestational age births. Discovery samples from 19-21 weeks' gestation were classified by high and low BMI. IBP4 / SHBG reversal values ​​were higher in the high BMI category because SHBG values ​​were lower. Separation of cases and controls was greater at lower BMI.

[0066] [Figure 22] Suppressed SHBG levels in PTB cases with low BMI. Linear fitting of SHBG serum levels in PAPR subjects across GABD. SHBG levels are suppressed by high BMI. SHBG levels increase throughout pregnancy. PTB cases with low BMI have reduced SHBG levels, which increase at an accelerated rate throughout pregnancy.

[0067] [Figure 23] Figure 23 summarizes the distribution of PAPR studied.

[0068] [Figure 24] FIG. 24 shows the ROC curves and corresponding AUC values ​​using the IBP4 / SHBG predictors to classify the BMI-stratified validation sample set.

[0069] [Figure 25] Figure 25 shows the prevalence-adjusted positive predictive value (PPV), a measure of clinical risk, as a function of predictor score. The calculated association between predictor score and PPV allows the probability of sPTB risk to be determined for any unknown subject. The top (purple) line under the risk curve graph corresponds to GAB <35 0 / 7 weeks, the second line (red) from the top corresponds to GAB 35 0 / 7 to 37 0 / 7 / week, the third line (green) corresponds to GAB 37 0 / 7 to 39 0 / 7 / week, and the fourth line (blue) from the top corresponds to GAB ≤ 39 0 / 7 weeks.

[0070] [Figure 26] Figure 26 displays the proportion of births in the high-risk and low-risk groups as events in the Kaplan-Meier analysis. High and low risk were defined as a relative risk greater than or less than two times the average population risk of sPTB (=14.6%) according to the data in Figure 25.

[0071] [Figure 27] Figure 27 shows the ROC curve corresponding to predictor performance using a combination of subjects from the blinded corroboration and validation analyses within the optimal BMI and GA intervals. The ROC curve for the combined sample corresponds to an AUROC of 0.72 (p=0.013).

[0072] [Figure 28] FIG. 28 shows that 44 proteins were either up- or down-regulated in the overlapping 3-week GA intervals and passed the analytical filters.

[0073] [Figure 29] FIG. 29 shows that the overall top performing inverse IBP4 / SHBG had an AUROC=0.74 in the interval from 19 0 / 7 to 21 6 / 7.

[0074] [Figure 30] Figure 30 shows that the average AUROC obtained from 2,000 bootstrap replicates was 0.76. The blinded IBP4 / SHBG AUROC performance for the validation samples was 0.77 and 0.79 for all subjects and BMI-stratified subjects, respectively, which showed good agreement with the performance obtained in discovery. After blind validation, the discovery and validation samples were combined for bootstrap performance determination.

[0075] [Figure 31] FIG. 31 shows the separation of sPTB cases versus controls derived by MS versus ELISA score values.

[0076] [Figure 32]FIG. 32 shows immunoassay vs. MS ROC analysis without BMI restriction.

[0077] [Figure 33] FIG. 33 shows the immunoassay vs. MS ROC analysis when BMI is greater than 22 and less than or equal to 37.

[0078] [Figure 34] FIG. 34 shows the correlation between MS- and ELISA-derived IBP4 / SHBG score values ​​within GABD133-146 for BMI-stratified subjects (left panel) and all subjects (right panel).

[0079] [Figure 35] Figure 35 shows ELISA and MS separation of controls and cases (BMI stratified).

[0080] [Figure 36] Figure 36 shows the Elisa and MS separation of controls and cases (all BMI).

[0081] [Figure 37] FIG. 37 shows a comparison of SHBG measurements using an Abbott Architect CMIA, a semi-automated immunoassay machine, and by Sera Prognostics' proteomic analysis method, which involves immunodepletion of samples, enzymatic digestion, and analysis on an Agilent 6490 mass spectrometer.

[0082] [Figure 38] Figure 38 shows a comparison of SHBG measurements using a Roche cobas e602 analyzer, a semi-automated immunoassay machine, and a proteomic analysis method from Sera Prognostics that involves immunodepletion of samples, enzymatic digestion, and analysis on an Agilent 6490 mass spectrometer.

[0083] [Figure 39]Figure 39 shows a comparison of SHBG measurements from the Abbott Architect CMIA and Roche cobas e602 analyzers, both semi-automated immunoassay machines.

[0084] [Figure 40] Figure 40 shows the domain and structural features of the longest isoform of the IBP4 protein (Uniprot: P22692). The IBP4 QCHPALDGQR (aa, 214-223) peptide (SEQ ID NO: 2) is located within the thyroglobulin type 1 domain. IBP4 has a single N-linked glycosylation site at residue 125.

[0085] [Figure 41] FIG. 41 highlights the location of the QCHPALDGQR peptide (SEQ ID NO: 2) in two IBP4 isoforms (SEQ ID NOs: 158 and 159, respectively, in order of appearance).

[0086] [Figure 42] Figure 42 shows the domain and structural features of the longest isoform of the SHBG protein (Uniprot: P04278). The SHBG IALGGLLFPASNLR (aa, 170-183) peptide (SEQ ID NO: 18) is located in the first lamin G-like domain. SHBG has three glycosylation sites: two N-linked sites at residues 380 and 396; and one O-linked site at residue 36.

[0087] [Figure 43] Figure 43 highlights the location of the IALGGLLFPASNLR peptide (SEQ ID NO: 18) in exon 4 within the seven isoforms of SHBG (SEQ ID NOs: 160, 160, 160, 160, 160, 160, and 161, respectively, in order of appearance).

[0088] [Figure 44]Figure 44 shows the mean response ratios of IBP4 levels for sPTB cases and full-term controls separately across gestational age at blood collection (GABD). Cross-sectional discovery data were analyzed by smoothing over a sliding 10-day window. The case-to-control signal corresponds to approximately a maximum 10% difference.

[0089] [Figure 45] Figure 45 shows the mean response ratios of SHBG levels for sPTB cases and full-term controls separately across gestational age at blood collection (GABD). Cross-sectional discovery data were analyzed by smoothing over a sliding 10-day window. The case-to-control signal corresponds to approximately a maximum 10% difference.

[0090] [Figure 46] Figure 46 shows the IBP4 / SHBG predictor scores for sPTB cases and full-term controls separately across gestational age at blood collection (GABD). Cross-sectional discovery data were analyzed by smoothing over a sliding 10-day window. The maximum difference between the two curves corresponds to nearly a 20% difference compared to the approximately 10% difference in the individual analyte signals (Figures 45 and 46). These data demonstrate that the diagnostic signal obtained is amplified by using the IBP4 / SHBG reversal strategy.

[0091] [Figure 47]Figure 47 shows the amplification of diagnostic signals as a result of the formation of multiple different inversions. To investigate whether the amplification of diagnostic signals due to the formation of inversions is common, we examined the diagnostic performance of inversions formed with multiple different proteins by ROC analysis. The range of AUC values ​​(sPTB cases vs. term controls) using datasets derived from samples collected at 19 / 0 to 21 / 6 weeks of gestation is shown in the top panel. The adjacent boxplots show the range of ROC performance for the individual up- and down-regulated proteins used to form the associated inversions. Similarly, the bottom panel shows that p-values ​​derived from Wilcoxon tests (sPTB cases vs. term controls) for inversions are more significant than those for the corresponding individual proteins.

[0092] [Figure 48] Figure 48 shows the analytical coefficient of variation (CV) of individual IBP4 and SHBG response ratio measurements and the corresponding calculated reversal scores. Pooled control serum samples (pHGS) from pregnant donors with no biological variability were analyzed in multiple batches over several days. The reversal variability is less than the variability for individual proteins. These data demonstrate the formation of a reversal control for analytical variability that occurs during laboratory processing of samples. Analytical variability is not a biological phenomenon.

[0093] [Figure 49]Figure 49 shows the analytical CVs of multiple inversions and their individual up- and down-regulated proteins. To investigate whether the amplification of diagnostic signals due to the formation of inversions is common, we examined the ROC performance (AUC) of high-performing inversions (AUC > 0.6) formed by the ratio of multiple proteins. The range of AUC values ​​(sPTB cases vs. term controls) using datasets derived from samples collected at 19 / 0 to 21 / 6 weeks of gestation is shown in the top panel. The adjacent box plots show the range of ROC performance of the individual up- and down-regulated proteins used to form the associated inversions. Similarly, p-values ​​derived from Wilcoxon tests of inversions (sPTB cases vs. term controls) are more significant than those of the corresponding individual proteins.

[0094] [Figure 50] FIG. 50 shows PreTRM™ score comparisons for subjects annotated as medically indicated to have pre-eclampsia compared to other indications.

[0095] [Figure 51] Figure 51 shows a table of IBP4 / SHBG predictor performance metrics for the validation sample set (BMI > 22 <= 37). Predictor sensitivity, specificity, area under the ROC curve (AUC), and odds ratio were determined for defining cases (below the cutoff) from controls (above the cutoff) using different boundaries.

[0096] [Figure 52] Figure 52 shows an inverted intensity heatmap of diabetes annotations. Red arrows indicate diabetes cases. Samples are listed at the bottom, with PTB cases on the right side of the screen and full-term births on the left. Diabetic patients are clustered on the right, indicating that it is possible to build a diagnostic test from biomarkers to predict gestational diabetes.

[0097] [Figure 53] Figure 53 shows hierarchical clustering of analyte response ratios.

[0098] [Figure 54] Figure 54 shows differentially expressed proteins that function in extracellular matrix interactions.

[0099] [Figure 55] Figure 55 shows a kinetic plot of differentially expressed proteins that function in the IGF-2 pathway, showing maximum separation at 18 weeks.

[0100] [Figure 56] Figure 56A shows a schematic of the interactions between IGF-2, IBP4, PAPP1, and PRG2 proteins in sPTB that affect the bioavailability of these proteins. Figure 56B shows a schematic of the intracellular signals preferentially activated by insulin binding to IR-B and by insulin and IGF binding to either IR-A or IGF1R.

[0101] [Figure 57] FIG. 57 shows kinetic plots of differentially expressed proteins with functions related to metabolic hormone balance.

[0102] [Figure 58] FIG. 58 shows the kinetic plot of differentially expressed proteins with functions in angiogenesis.

[0103] [Figure 59] FIG. 59 shows kinetic plots of differentially expressed proteins with functions related to innate immunity.

[0104] [Figure 60] Figure 60 shows kinetic plots of differentially expressed proteins with functions related to coagulation.

[0105] [Figure 61] Figure 61 shows the kinetic plots of differentially expressed serum / secreted proteins.

[0106] [Figure 62] Figure 62 shows the kinetic plots of differentially expressed PSG / IBP.

[0107] [Figure 63] Figure 63 shows kinetic plots of differentially expressed ECM / cell surface proteins.

[0108] [Figure 64] Figure 64 shows kinetic plots of differentially expressed complement / acute phase protein-1.

[0109] [Figure 65] Figure 65 shows kinetic plots of differentially expressed complement / acute phase protein-2.

[0110] [Figure 66] Figure 66 shows kinetic plots of differentially expressed complement / acute phase protein-3.

[0111] [Figure 67] Figure 67 shows kinetic plots of differentially expressed complement / acute phase protein-4.

[0112] [Figure 68] Figure 68 shows kinetic plots of the analytes designated in panels AI using data from gestational age at blood draw (GABD) from 17 weeks 0 days to 28 weeks 6 days.

[0113] [Figure 69] Figure 69 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0114] [Figure 70] Figure 70 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0115] [Figure 71] Figure 71 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0116] [Figure 72] Figure 72 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0117] [Figure 73] Figure 73 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0118] [Figure 74] Figure 74 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0119] [Figure 75] Figure 75 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0120] [Figure 76] Figure 76 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0121] [Figure 77] Figure 77 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0122] [Figure 78] Figure 78 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0123] [Figure 79] Figure 79 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0124] [Figure 80] Figure 80 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0125] [Figure 81] Figure 81 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0126] [Figure 82] Figure 82 shows kinetic plots of the analytes designated in panels AI using data from 17 weeks 0 days to 28 weeks 6 days.

[0127] [Figure 83] Figure 83 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0128] [Figure 84] Figure 84 shows kinetic plots of the analytes designated in panels AI using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0129] [Figure 85] Figure 85 shows kinetic plots of the peptide transitions designated in panels A-G using GABD data from 17 weeks 0 days to 28 weeks 6 days.

[0130] [Figure 86] Figure 86 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0131] [Figure 87] Figure 87 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0132] [Figure 88] Figure 88 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0133] [Figure 89] Figure 89 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0134] [Figure 90] Figure 90 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0135] [Figure 91] Figure 91 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0136] [Figure 92] Figure 92 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0137] [Figure 93] Figure 93 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0138] [Figure 94] Figure 94 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0139] [Figure 95] Figure 95 shows kinetic plots of the peptide transitions designated in panels AC using a gestational age at term cutoff of <37 0 / 7 weeks vs. >=37 0 / 7 weeks.

[0140] [Figure 96] Figure 96 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0141] [Figure 97] Figure 97 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0142] [Figure 98] Figure 98 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0143] [Figure 99] Figure 99 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0144] [Figure 100] Figure 100 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0145] [Figure 101] Figure 101 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0146] [Figure 102] Figure 102 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0147] [Figure 103] Figure 103 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0148] [Figure 104] Figure 104 shows kinetic plots of the peptide transitions designated in panels AI using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0149] [Figure 105] Figure 105 shows kinetic plots of the peptide transitions designated in panels AC using a gestational age at term cutoff of <35 0 / 7 weeks vs. >=35 0 / 7 weeks.

[0150] [Figure 106] FIG. 106 shows IBP4 and SHBG levels and IBP4 / SHBG reversal values ​​for sPTB cases and controls separately.

[0151] [Figure 107] Figure 107 shows the correlation of MSD results from a commercial ELISA kit and MS-MRM.

[0152] [Figure 108]Figure 108 provides a boxplot showing an example of an inversion with good performance in 19-20 week preterm labor (PTL) without PPROM.

[0153] [Fig. 109A-B] Figure 109 provides a boxplot showing an example of a reversal with good performance in preterm premature rupture of membranes (PPROM) at 19-20 weeks. [Fig. 109C-D] Figure 109 provides a boxplot showing an example of a reversal with good performance in preterm premature rupture of membranes (PPROM) at 19-20 weeks.

[0154] [Figure 110] Figure 110 shows risk curves illustrating the relationship between predictor score (ln IBP4 / SHBG) and prevalence-adjusted relative risk (positive predictive value) of sPTB using a cutoff of <37 0 / 7 gestational weeks vs. >=37 0 / 7 gestational weeks. The top line (purple) under the risk curve graph corresponds to sPTB (GAB<35 weeks), the second line (red) from the top corresponds to sPTB (35≦GAB<37 weeks), the third line (green) corresponds to TERM (37≦GAB<39 weeks), and the fourth line (blue) from the top corresponds to TERM (39 weeks≦GAB).

[0155] [Figure 111] Figure 111 shows risk curves illustrating the relationship between predictor score (ln IBP4 / SHBG) and prevalence-adjusted relative risk (positive predictive value) of sPTB using a cutoff of <35 0 / 7 gestational weeks vs. >=35 0 / 7 gestational weeks. The top line (purple) under the risk curve graph corresponds to sPTB (GAB<35 weeks), the second line (red) from the top corresponds to sPTB (35≦GAB<37 weeks), the third line (green) corresponds to TERM (37≦GAB<39 weeks), and the fourth line (blue) from the top corresponds to TERM (39 weeks≦GAB). DETAILED DESCRIPTION OF THE INVENTION

[0156] (Detailed explanation) The present disclosure is generally based on the discovery that certain proteins and peptides in biological samples obtained from pregnant women are differentially expressed in pregnant women who exhibit an increased risk of preterm birth compared to controls. More particularly, the present disclosure is based in part on the unexpected discovery that the inverted values ​​of the biomarker pairs disclosed herein can be used in a method for determining the probability of preterm birth in pregnant women with high sensitivity and specificity. The proteins and peptides disclosed herein as component ratios and / or inverted pairs serve as biomarkers for classifying test samples, predicting the probability of preterm birth, predicting the probability of full-term birth, predicting gestational age at birth (GAB), predicting time to birth (TTB), and / or monitoring the progress of preventive therapy for pregnant women at risk of PTB, either individually, by ratio, by inverted pair, or by panel of biomarkers / inverted pairs. The inverted value is the ratio of the relative peak area of ​​the up-regulated biomarker to the relative peak area of ​​the down-regulated biomarker, and serves both as a variability normalization and diagnostic signal amplification. The present invention resides, in part, in the selection of specific biomarkers that, when paired together, can predict the probability of preterm birth based on the inverted value. Thus, underlying the present invention is human ingenuity in selecting specific biomarkers that, when paired in novel inversions, provide useful information.

[0157] The term "inversion value" refers to the ratio of the relative peak area of ​​an upregulated analyte to the relative peak area of ​​a downregulated analyte, and serves both to normalize variability and amplify diagnostic signals. Among all possible inversions within a narrow window, a subset can be selected based on their individual univariate performance. As disclosed herein, the ratio of the relative peak area of ​​an upregulated biomarker to the relative peak area of ​​a downregulated biomarker, referred to herein as the inversion value, can be used to identify robust and accurate classifiers, predict the probability of preterm birth, predict the probability of full-term birth, predict gestational age at birth (GAB), predict time to delivery, and / or monitor the progress of preventive therapy in pregnant women. Thus, the present invention is based, in part, on identifying biomarker pairs whose relative expression is inverted between PTB and non-PTB, indicating a change in inversion value. The use of biomarker ratios in the methods disclosed herein corrects for variability resulting from artificial manipulations after removal of a biological sample from a pregnant woman. Such variability may be introduced, for example, during sample collection, processing, depletion, digestion, or any other step of the method used to measure biomarkers present in a sample, and is independent of how the biomarker behaves in nature. Thus, the present invention generally encompasses the use of inversion pairs in diagnostic or prognostic methods to reduce variability and / or to amplify, standardize, or clarify diagnostic signals.

[0158] The term "inverted value" refers to the ratio of the relative peak area of ​​an upregulated analyte to the relative peak area of ​​a downregulated analyte, and serves both to normalize for variability and to amplify the diagnostic signal, although it is also contemplated that biomarker pairs of the present invention can be measured by any other means, such as by subtraction, addition, or multiplication of relative peak areas. The methods disclosed herein encompass measuring biomarker pairs by such other means.

[0159] This method is advantageous because it does not rely on data standardization, helps avoid overfitting, and provides the simplest possible classifier, resulting in a very simple experimental test that is easily performed in the clinic. The use of marker pairs based on changes in inverse values ​​that do not rely on data standardization has enabled the development of clinically relevant biomarkers as disclosed herein. Because quantification of any single protein is subject to uncertainty caused by measurement variability, normal variability, and individual-related variations in baseline expression, identifying pairs of markers that may be under coordinate and systematic regulation enables robust methods of personalized diagnosis and prognosis.

[0160] The present disclosure provides biomarker reversal pairs and related panels of reversal pairs, methods, and kits for determining the probability of preterm birth in pregnant women.One major advantage of the present disclosure is that the risk of developing preterm birth can be assessed early in pregnancy, so that appropriate monitoring and clinical management to prevent preterm birth can be initiated accordingly.The present invention is particularly beneficial for women who lack any risk factors for preterm birth and would otherwise not be identified and treated.

[0161] By way of example, the present disclosure includes a method for generating a result useful for determining the probability of preterm birth in a pregnant woman by obtaining a dataset associated with a sample, the dataset including at least quantitative data regarding the relative expression of a pair of biomarkers identified to exhibit a reversal change in value predictive of preterm birth, and inputting the dataset into an analytical process that uses the dataset to generate a result useful for determining the probability of preterm birth in the pregnant woman. As described further below, the quantitative data can include amino acids, peptides, polypeptides, proteins, nucleotides, nucleic acids, nucleosides, sugars, fatty acids, steroids, metabolites, carbohydrates, lipids, hormones, antibodies, regions of interest that serve as surrogates for biological macromolecules, and combinations thereof.

[0162] In addition to the specific biomarkers identified in this disclosure, for example, by accession numbers in public databases, sequences, or references, the present invention also contemplates the use of biomarker variants, now known or later discovered, that are at least 90%, or at least 95%, or at least 97% identical to the exemplary sequences that have utility for the methods of the present invention. These variants may represent polymorphisms, splice variants, mutations, and the like. In this regard, the present specification discloses several art-known proteins in the context of the present invention and provides exemplary accession numbers associated with one or more public databases and exemplary references to published journal articles related to these art-known proteins. However, those of skill in the art will readily identify additional accession numbers and journal articles that can provide additional characteristics of the disclosed biomarkers, and will understand that the exemplary references are in no way limiting with respect to the disclosed biomarkers. As described herein, a variety of techniques and reagents find use in the methods of the present invention. Suitable samples in the context of the present invention include, for example, blood, plasma, serum, amniotic fluid, vaginal secretions, saliva, and urine. In some embodiments, the biological sample is selected from the group consisting of whole blood, plasma, and serum.In certain embodiments, the biological sample is serum.As described herein, biomarkers can be detected through various assays and techniques known in the art.As further described herein, such assays include, but are not limited to, mass spectrometry (MS)-based assays, antibody-based assays, and assays that combine aspects of both.

[0163] Protein biomarkers that are components of inverted pairs described herein include, for example, insulin-like growth factor binding protein 4 (IBP4), sex hormone binding globulin (SHBG), vitronectin (VTNC), group-specific component (vitamin D binding protein) (VTDB), cathepsin D (lysosomal aspartyl protease) (CATD), pregnancy-specific beta-1-glycoprotein 2 (PSG2), inter-alpha-trypsin inhibitor heavy chain family member 4 (ITIH4), cell adhesion molecule L1-like (CHL1), complement component 1, Q subcomponent B chain (C1QB), fibulin 3 (FBLN3), hemopexin (HEMO or HPX), insulin-like growth factor binding protein 6 (IBP6), and prostaglandin D2 synthase 21 kDa (PTGDS).

[0164] In some embodiments, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring in a biological sample obtained from the pregnant woman the reversal values ​​of at least one pair of biomarkers selected from the group comprising the pairs listed in any of the accompanying figures, including FIG. 1 , and tables.

[0165] In some embodiments, the present invention provides methods for determining the probability of preterm birth in a pregnant woman, comprising measuring reversal values ​​of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman.

[0166] The present invention provides isolated biomarkers selected from the group shown in Table 26. The biomarkers of the present invention can predict the risk of preterm birth in pregnant women. In some embodiments, the isolated biomarkers are selected from the group consisting of IBP4, SHBG, VTNC, VTDB, CATD, PSG2, ITIH4, CHL1, C1QB, FBLN3, HPX, and PTGDS. In some embodiments, the isolated biomarkers are selected from the group consisting of IBP4, SHBG, PSG3, LYAM1, IGF2, CLUS, IBP3, INHBC, PSG2, PEDF, CD14, and APOC3.

[0167] The present invention provides surrogate peptides of isolated biomarkers selected from the group shown in Table 26. In some embodiments, the surrogate peptides of isolated biomarkers are selected from the group of surrogate peptides shown in Table 26. The biomarkers of the present invention and their surrogate peptides can be used in methods for predicting the risk of preterm birth in pregnant women. In some embodiments, the surrogate peptides correspond to isolated biomarkers selected from the group consisting of IBP4, SHBG, VTNC, VTDB, CATD, PSG2, ITIH4, CHL1, C1QB, FBLN3, HPX, and PTGDS. In some embodiments, the surrogate peptides correspond to isolated biomarkers selected from the group consisting of IBP4, SHBG, PSG3, LYAM1, IGF2, CLUS, IBP3, INHBC, PSG2, PEDF, CD14, and APOC3.

[0168] The present invention provides stable isotope-labeled standard peptides (SIS peptides) corresponding to surrogate peptides selected from the group shown in Table 26. The biomarkers of the present invention, their surrogate peptides, and SIS peptides can be used in methods for predicting the risk of preterm birth in pregnant women. In some embodiments, the SIS peptides correspond to surrogate peptides of isolated biomarkers selected from the group consisting of IBP4, SHBG, VTNC, VTDB, CATD, PSG2, ITIH4, CHL1, C1QB, FBLN3, HPX, and PTGDS. In some embodiments, the SIS peptides correspond to surrogate peptides of isolated biomarkers selected from the group consisting of IBP4, SHBG, PSG3, LYAM1, IGF2, CLUS, IBP3, INHBC, PSG2, PEDF, CD14, and APOC3.

[0169] In some embodiments, the invention provides an isolated biomarker pair IBP4 / SHBG that exhibits an inverse value change among pregnant women at risk of preterm birth compared to full-term controls. In further embodiments, the invention provides an isolated biomarker pair IBP4 / SHBG that exhibits a higher ratio in pregnant women at risk of preterm birth compared to full-term controls.

[0170] In some embodiments, the present invention provides methods for determining the likelihood of preterm birth in a pregnant woman, comprising measuring reversal values ​​of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, and CATD / SHBG in a biological sample obtained from the pregnant woman to determine the likelihood of preterm birth in the pregnant woman. In additional embodiments, the sample is obtained at 19-21 weeks GABD. In further embodiments, the sample is obtained at 19-22 weeks GABD.

[0171] In some embodiments, the present invention provides methods for determining the likelihood of preterm birth in a pregnant woman, comprising measuring the inverse level of IBP4 / SHBG in a biological sample obtained from the pregnant woman to determine the likelihood of preterm birth in the pregnant woman. In additional embodiments, the sample is obtained at 19-21 weeks GABD. In further embodiments, the sample is obtained at 19-22 weeks GABD.

[0172] In some embodiments, the present invention provides a method for determining the likelihood of preterm birth in a pregnant woman, comprising measuring reversal values ​​of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the likelihood of preterm birth in the pregnant woman, wherein the presence of a change in reversal values ​​between the pregnant woman and a full-term control determines the likelihood of preterm birth in the pregnant woman. In additional embodiments, the sample is obtained at 19-21 weeks GABD. In further embodiments, the sample is obtained at 19-22 weeks GABD.

[0173]

[0010] Included within embodiments of the present invention is an iterative method for determining the probability of preterm birth in a pregnant woman, comprising measuring inverted values ​​of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, HPX / PTGDS, and any other pair of biomarkers selected from the proteins described and / or exemplified herein, in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman, wherein the presence of a change in inverted values ​​between the pregnant woman and a full-term control determines the probability of preterm birth in the pregnant woman. Iterative performance of the methods described herein includes subsequent measurements obtained from a single sample, as well as obtaining subsequent samples for measurement. For example, if a pregnant woman's probability of preterm birth, which can be expressed as a risk score, is determined to exceed a specified value, the above method can be repeated using different inversion pairs of the same sample, or the same or different inversion pairs of subsequent samples, to further stratify the risk of sPTB.

[0174] In addition to specific biomarkers, the present disclosure further includes biomarker variants that are about 90%, about 95%, or about 97% identical to the exemplary sequences. Variants, as used herein, include polymorphisms, splice variants, mutations, etc. Although inversion value changes are described with reference to protein biomarkers, they may also be identified at the protein expression level or at the gene expression level of biomarker pairs.

[0175] Additional markers can be selected from one or more risk indicators, including, but not limited to, maternal characteristics, medical history, past pregnancy history, and obstetric history. Such additional markers can include, for example, previous low birth weight or preterm birth, multiple second-trimester spontaneous abortions, previous first-trimester induced abortions, familial and intergenerational factors, infertility history, nulliparity, placental abnormalities, cervical and uterine abnormalities, short cervical length measurement, pregnancy bleeding, intrauterine growth restriction, intrauterine diethylstilbestrol exposure, multiple pregnancy, infant sex, short stature, low prepregnancy weight, low or high body mass index, diabetes, hypertension, genitourinary infections (i.e., urinary tract infections), asthma, anxiety and depression, asthma, hypertension, hypothyroidism, and other demographic risk indicators for preterm birth can include, for example, maternal age, race / ethnicity, single marital status, low socioeconomic status, maternal age, employment-related physical activity, occupational and environmental exposures, and stress. Additional risk indicators may include inadequate prenatal care, smoking, marijuana and other illicit drug use, cocaine use, alcohol consumption, caffeine intake, maternal weight gain, dietary intake, sexual activity during the third trimester, and leisure-time physical activity (Preterm (Birth: Causes, Consequences, and Prevention, Institute of Medicine (US) Committee on Understanding Premature Birth and Assurance Healthy Outcomes; Behrman RE, Butler AS, eds. Washington (DC): National Academies Press (US); 2007). Additional risk signatures useful as markers can be identified using learning algorithms known in the art, such as linear discriminant analysis, support vector machine classification, recursive feature elimination, predictive analysis of microarrays, logistic regression, CART, FlexTree, LART, random forest, MART, and / or survival analysis regression, which are known to those skilled in the art and further described herein.

[0176] It must be noted that as used in this specification and the appended claims, the singular forms "a," "an," and "the" include plural referents unless the content clearly dictates otherwise. Thus, for example, reference to a "biomarker" includes a mixture of two or more biomarkers, and the like.

[0177] The term "about," particularly with reference to a given amount, is meant to encompass deviations of plus or minus 5 percent.

[0178] As used in this application, including the appended claims, the singular forms "a," "an," and "the" include plural referents unless the content clearly dictates otherwise, and are used interchangeably with "at least one" and "one or more."

[0179] As used herein, the terms "comprises," "comprising," "includes," "including," "contains," "containing," and any variations thereof, are intended to cover a non-exclusive inclusion, and a process, method, product-by-process, or composition that comprises, includes, or contains an element or list of elements does not include only those elements, but may include other elements not expressly listed in or inherent to such process, method, product-by-process, or composition.

[0180] As used herein, the term "panel" refers to a composition, such as an array or collection, that includes one or more biomarkers. The term can also refer to a profile or index of the expression pattern of one or more biomarkers described herein. The number of biomarkers useful for a biomarker panel is based on the sensitivity and specificity values ​​for a particular combination of biomarker values.

[0181] As used herein, unless otherwise indicated, the terms "isolated" and "purified" generally describe a composition that has been removed from its native environment (e.g., the natural environment if it occurs in nature) and therefore has been artificially altered from its native state so as to have significantly different characteristics with respect to at least one of structure, function, and properties. An isolated protein or nucleic acid is different from the manner in which it occurs in nature and includes synthetic peptides and proteins.

[0182] The term "biomarker" refers to a biological molecule, or a fragment of a biological molecule, the alteration and / or detection of which can be correlated with a particular physical condition or state. The terms "marker" and "biomarker" are used interchangeably throughout this disclosure. For example, the biomarkers of the present invention correlate with an increased likelihood of preterm birth. Such biomarkers include any suitable analyte, including, but not limited to, biological molecules containing nucleotides, nucleic acids, nucleosides, amino acids, sugars, fatty acids, steroids, metabolites, peptides, polypeptides, proteins, carbohydrates, lipids, hormones, antibodies, regions of interest that serve as surrogates for biological macromolecules, and combinations thereof (e.g., glycoproteins, ribonucleoproteins, lipoproteins). The term also encompasses portions or fragments of biological molecules, for example, peptide fragments of proteins or polypeptides comprising at least 5 consecutive amino acid residues, at least 6 consecutive amino acid residues, at least 7 consecutive amino acid residues, at least 8 consecutive amino acid residues, at least 9 consecutive amino acid residues, at least 10 consecutive amino acid residues, at least 11 consecutive amino acid residues, at least 12 consecutive amino acid residues, at least 13 consecutive amino acid residues, at least 14 consecutive amino acid residues, at least 15 consecutive amino acid residues, at least 5 consecutive amino acid residues, at least 16 consecutive amino acid residues, at least 17 consecutive amino acid residues, at least 18 consecutive amino acid residues, at least 19 consecutive amino acid residues, at least 20 consecutive amino acid residues, at least 21 consecutive amino acid residues, at least 22 consecutive amino acid residues, at least 23 consecutive amino acid residues, at least 24 consecutive amino acid residues, at least 25 consecutive amino acid residues, or more consecutive amino acid residues.

[0183] As used herein, the term "surrogate peptide" refers to a peptide selected to serve as a surrogate for quantifying a biomarker of interest in an MRM assay configuration. Quantification of the surrogate peptide is best achieved using a stable isotope-labeled standard surrogate peptide ("SIS surrogate peptide" or "SIS peptide") in conjunction with MRM detection techniques. The surrogate peptide may be synthetic. For example, an SIS surrogate peptide can be synthesized in which arginine or lysine, or any other amino acid at the C-terminus of the peptide that serves as an internal standard for the MRM assay, is heavily labeled. SIS surrogate peptides are not naturally occurring peptides and have significantly different structures and properties compared to their naturally occurring counterparts.

[0184] In some embodiments, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring the ratio of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman, wherein a change in the ratio between the pregnant woman and a full-term control indicates the probability of preterm birth in the pregnant woman. In some embodiments, the ratio may include an upregulated protein in the numerator and a downregulated protein in the denominator or both. For example, as exemplified herein, IBP4 / SHBG is the ratio of the upregulated protein in the numerator and the downregulated protein in the denominator, defined herein as "inversion." When a ratio includes an upregulated protein in the numerator or a downregulated protein in the denominator, the unregulated protein will play a normalization role (e.g., reduce pre-analysis or analytical variability). In the specific case where the ratio is "inverted," both amplification and normalization are possible. It is understood that the method of the present invention is not limited to a subset of inversions, but also encompasses biomarker ratios.

[0185] As used herein, the term "inversion" refers to the ratio of an upregulated analyte measurement to a downregulated analyte measurement. In some embodiments, the analyte value itself is the ratio of the endogenous analyte peak area to the ratio of the peak area of ​​the corresponding stable isotope standard analyte, referred to herein as the response ratio or relative ratio.

[0186] As used herein, the term "inversion pair" refers to a pair of biomarkers that show a change in value between the classes being compared. Detecting an inversion in protein concentration or gene expression level eliminates the need for data standardization or the establishment of a threshold across the entire population. In some embodiments, the inversion pair is the isolated biomarker pair IBP4 / SHBG, which shows an inversion value change between pregnant women at risk of preterm birth compared to full-term controls. In a further embodiment, the inversion pair IBP4 / SHBG shows a higher ratio in pregnant women at risk of preterm birth compared to full-term controls. The definition of any inversion pair encompasses corresponding inversion pairs in which the individual biomarkers are swapped between the numerator and the denominator. Those skilled in the art will recognize that such corresponding inversion pairs provide equally useful information regarding their predictive power.

[0187] The term "inversion value" refers to the ratio of the relative peak area of ​​up-regulated analytes to the relative peak area of ​​down-regulated analytes, and serves both to normalize variability and amplify diagnostic signals.From all possible inversions within a narrow window, a subset can be selected based on individual univariate performance.As disclosed herein, the ratio of the relative peak area of ​​up-regulated biomarkers to the relative peak area of ​​down-regulated biomarkers, referred to herein as inversion value, can be used to identify robust and accurate classifiers, predict the probability of preterm birth, predict the probability of full-term birth, predict gestational age at birth (GAB), predict time to delivery, and / or monitor the progress of preventive therapy in pregnant women.

[0188] This inversion method is advantageous because it does not rely on data standardization, helps avoid overfitting, and provides the simplest possible classifier, resulting in a very simple experimental test that can be easily performed in the clinic. The use of inversion-based biomarker pairs that do not rely on data standardization, as described herein, has very high potential as a method for identifying clinically relevant PTB biomarkers. Because the quantification of any single protein is subject to uncertainty caused by measurement variability, normal variability, and individual-related variations in baseline expression, identifying marker pairs that may be under coordinate and systematic regulation should prove to be a more robust method for personalized diagnosis and prognosis.

[0189] The present invention provides a composition comprising an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein the pair of biomarkers exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls. In one embodiment, the composition comprises stable isotope-labeled standard peptides (SIS peptides) of surrogate peptides derived from each of the biomarkers.

[0190] In certain embodiments, the present invention provides an isolated biomarker pair consisting of IBP4 and SHBG that exhibit inverse value changes between pregnant women at risk of preterm birth and full-term controls.

[0191] IBP4 is a member of the insulin-like growth factor binding protein (IBP) family that negatively regulates the insulin-like growth factors IGF1 and IGF2 (Forbes et al., Insulin-like growth factor I and II regulate the life cycle of trophoblasts). in the developing human placenta. Am J Physiol, Cell Physiol. 2008;294(6):C1313-22). IBP4 is expressed by the syncytiotrophoblast (Crosley et al., IGFBP-4 and -5 are expressed in first-trimester villus and differentially regulate the migration of HTR-8 / SVneo cells. Reprod Biol Endocrinol. 2014;12(1):123) and is the predominant IBP expressed by the extravillous trophoblast (Qiu et al., Significance of IGFBP-4 in the development of fetal growth restriction. J Clin Endocrinol Metab. 2012;97(8):E1429-39). Maternal IBP4 levels during early pregnancy are higher in pregnancies complicated by fetal growth restriction and preeclampsia compared with full-term pregnancies. (Qiu et al. above, 2012)

[0192] SHBG regulates the availability of biologically active unbound steroid hormones. Hammond GL., Diverse roles for sex hormone-binding globulin in reproduction. Biol Reprod., 2011;85(3):431-41. Plasma SHBG levels are , which increases 5-10 fold during pregnancy (Anderson DC. Sex-hormone-binding globulin. Clin Endocrinol (Oxf)., 1974;3(1):69-96), there is evidence of extrahepatic expression, including in placental trophoblast cells (Larrea et al., Evidence that human placenta is a (Site of sex hormone-binding globulin gene expression. J Steroid Biochem Mol Biol., 1993;46(4):497-505). Physiologically, SHBG levels are negatively correlated with triglyceride levels, insulin levels, and BMI (Simo et al., Novel insights in SHBG regulation and clinical implications. Trends Endocrinol Metab., 2015;26(7):376-83). SHB The effect of BMI on G levels may, in part, explain the improved predictive performance with BMI stratification.

[0193] Intra-amniotic infection and inflammation have been associated with PTB, as they exhibit increased levels of pro-inflammatory cytokines, including TNF-α and IL1-β. (Mendelson CR. Minireview: fetal-maternal hormonal signaling in pregnancy and labor. Mol Endocrinol. 2009;23(7):947-54; Gomez-Lopez et al., Immune cells in term and preterm labor., Cell Mol Immunol., 2014;11(6) ): pp. 571-81). SHBG transcription in the liver is mediated by IL1-β and NF-kB-mediated TNF-α signaling (Simo et al., Novel insights in SHBG regulation and clinical implications. Trends Endocrinol Metab. 2015;26(7):37 6-83), a pathway involved in the initiation of normal and abnormal labor (Lindstrom TM, Bennett PR. The role of nuclear factor kappa B in human labor. Reproduction. 2005;130(5):569-81). Low SHBG levels in women predisposed to sPTB may be the result of infection and / or inflammation. Thus, SHBG may be important in regulating androgen and estrogen action in the placenta-fetal unit in response to upstream inflammatory signals.

[0194] In one embodiment, the present invention provides a composition comprising a pair of surrogate peptides corresponding to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein the pair of biomarkers exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls.

[0195] In a further embodiment, the present invention provides a panel of at least two pairs of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein each of the pairs exhibits an inverse value change between pregnant women at risk of preterm birth and full-term controls.

[0196] In additional embodiments, the present invention provides a panel of at least two pairs of surrogate peptides, each of which corresponds to a pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS, wherein each of the pairs exhibits inverse value changes between pregnant women at risk of preterm birth and full-term controls.

[0197] In one embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring the reversal values ​​of at least one pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman.

[0198] In another embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring a change in inversion values ​​of at least two pairs of a panel of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman. In some embodiments, the inversion values ​​reveal the presence of a change in inversion values ​​between the pregnant woman and a full-term control, indicating the probability of preterm birth in the pregnant woman. In some embodiments, the measuring step comprises measuring surrogate peptides of the biomarkers in a biological sample obtained from the pregnant woman.

[0199] In one embodiment, the present invention provides a method for determining a pregnant woman's likelihood of preterm birth, comprising measuring, in a biological sample obtained from the pregnant woman, the reversal values ​​of at least one pair of biomarkers selected from the group consisting of the biomarkers listed in any of Tables 1-77 and Figures 1-111 of the pregnant woman to determine the pregnant woman's likelihood of preterm birth.

[0200] In an additional embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring in a biological sample obtained from the pregnant woman the reversal values ​​of at least one pair of biomarkers selected from the group consisting of the biomarker pairs designated in Tables 27-59, 61-72, 76, and 77 to determine the pregnant woman's probability of preterm birth.

[0201] In a further embodiment, the present invention provides a method for determining the probability of preterm birth in a pregnant woman, comprising measuring, in a biological sample obtained from the pregnant woman, the reversal values ​​of at least one pair of biomarkers selected from the group consisting of the biomarkers listed in Table 26 of the pregnant woman to determine the pregnant woman's probability of preterm birth.

[0202] In another embodiment, the present invention provides a method for determining a pregnant woman's likelihood of preterm birth, comprising measuring, in a biological sample obtained from the pregnant woman, a change in inversion values ​​of a panel of at least two pairs of biomarkers selected from the group consisting of the biomarker pairs specified in any of Tables 1-77 and Figures 1-111 to determine the pregnant woman's likelihood of preterm birth. In some embodiments, the inversion values ​​reveal the presence of a change in inversion values ​​between the pregnant woman and a full-term control, indicating the pregnant woman's likelihood of preterm birth. In some embodiments, the measuring step comprises measuring surrogate peptides of the biomarkers in the biological sample obtained from the pregnant woman.

[0203] In another embodiment, the invention provides a method for determining a pregnant woman's likelihood of preterm birth, comprising measuring a change in inversion values ​​of a panel of at least two pairs of biomarkers selected from the group consisting of the biomarker pairs specified in Tables 27-59, 61-72, 76, and 77 in a biological sample obtained from the pregnant woman to determine the pregnant woman's likelihood of preterm birth. In some embodiments, the inversion values ​​reveal the presence of a change in inversion values ​​between the pregnant woman and a full-term control, indicating the pregnant woman's likelihood of preterm birth. In some embodiments, the measuring step comprises measuring surrogate peptides of the biomarkers in the biological sample obtained from the pregnant woman.

[0204] In another embodiment, the invention provides a method for determining a pregnant woman's likelihood of preterm birth, comprising measuring a change in inversion values ​​of at least two pairs of a panel of biomarkers selected from the group consisting of the biomarkers specified in Table 26 in a biological sample obtained from the pregnant woman to determine the pregnant woman's likelihood of preterm birth. In some embodiments, the inversion values ​​reveal the presence of a change in inversion values ​​between the pregnant woman and a full-term control, indicating the pregnant woman's likelihood of preterm birth. In some embodiments, the measuring step comprises measuring surrogate peptides of the biomarkers in a biological sample obtained from the pregnant woman.

[0205] For methods aimed at predicating the time to birth, see "Birth, "Birth" is understood to mean childbirth following the spontaneous onset of labor, with or without rupture of membranes.

[0206] Although described and exemplified with reference to methods for determining the probability of preterm birth in pregnant women, the present disclosure is equally applicable to methods for predicting gestational age at birth (GAB), predicting full-term birth, determining the probability of full-term birth in pregnant women, and predicting time to birth (TTB) in pregnant women. It will be apparent to those skilled in the art that each of the above methods has particular substantial utility and benefits related to maternal-fetal health considerations.

[0207] Additionally, while the present disclosure is described and exemplified with reference to methods for determining the probability of preterm birth in a pregnant woman, it is equally applicable to methods for determining the probability of preterm birth in a pregnant woman, including abnormal glucose tests, gestational diabetes, hypertension, preeclampsia, intrauterine growth retardation, stillbirth, fetal growth retardation, HELLP syndrome, oligohydramnios, chorioamnionitis, placenta previa, placenta accreta, abruption, and premature birth. Applicable methods for predicting placenta abruption, placental hemorrhage, preterm rupture of membranes, preterm labor, adverse cervix, post-term pregnancy, cholelithiasis, uterine over-distention, and stress As described in more detail below, the classifiers described herein are sensitive to components of PTB that are medically indicated based on conditions such as, for example, pre-eclampsia or gestational diabetes.

[0208] In some embodiments, the present disclosure provides biomarkers, biomarker pairs, and / or inverses that are strong predictors of time to birth (TTB), exemplified herein by the use of ITIH4 / CSH (Figure 10). TTB is defined as the difference between GABD and gestational age at birth (GAB). This discovery allows for the prediction of TTB or GAB, either individually or in mathematical combinations of such analytes. Analytes that lack case-control differences but show changes in analyte intensity throughout pregnancy are useful for pregnancy clocks according to the methods of the present invention. It may be possible to determine gestational age using calibration of multiple analytes that cannot be used to diagnose preterm birth for other disorders. Such pregnancy clocks are useful for confirming age determinations by other measurements (e.g., last menstrual period and / or ultrasound age determinations) or for post hoc and more accurate predictions of, for example, sPTB, GAB, or TTB alone. These analytes, also referred to herein as "clock proteins," can be used without or in conjunction with other age determination methods to determine gestational age. Table 60 provides a list of clock proteins useful in the pregnancy clock of the present invention for predicting TTB and GAB.

[0209] In additional embodiments, the method for determining the probability of preterm birth in a pregnant woman further comprises detecting a measurable characteristic for one or more risk indicators associated with preterm birth, in which the risk indicators are selected from the group consisting of previous low birth weight or preterm birth, multiple second-trimester spontaneous abortions, previous first-trimester induced abortions, familial and intergenerational factors, a history of infertility, nulliparity, gravitational, primigravida, multiparous women, placental abnormalities, cervical and uterine abnormalities, pregnancy bleeding, intrauterine growth restriction, intrauterine diethylstilbestrol exposure, multiple pregnancy, infant sex, short stature, low prepregnancy weight, low or high body mass index, diabetes, hypertension, and genitourinary infection.

[0210] A "measurable characteristic" is any characteristic, feature, or aspect that can be determined and correlated with the probability of preterm birth in a subject. This term also encompasses any characteristic, feature, or aspect that can be determined and correlated in relation to predicting GAB in pregnant women, predicting full-term birth, or predicting time to birth. For biomarkers, such measurable characteristics include, for example, the presence, absence, or concentration of the biomarker or a fragment thereof in a biological sample; an altered structure, such as the presence or amount of post-translational modifications, such as oxidation at one or more positions on the amino acid sequence of the biomarker, or the presence of an altered conformation, for example, compared to the conformation of the biomarker in a term control subject; and and / or the presence, amount, or altered structure of a biomarker as part of a profile of more than one biomarker.

[0211] In addition to biomarkers, measurable characteristics can further include risk indicators, including, for example, maternal characteristics, age, race, ethnicity, medical history, previous reproductive history, and obstetric history. For risk indicators, measurable characteristics can include, for example, previous low birth weight or preterm birth, multiple second-trimester spontaneous abortions, previous first-trimester induced abortions, familial and intergenerational factors, infertility history, nulliparity, placental abnormalities, cervical and uterine abnormalities, short cervical length measurement, pregnancy bleeding, intrauterine growth restriction, intrauterine diethylstilbestrol exposure, multiple pregnancy, infant sex, short stature, low pre-pregnancy weight / low body mass index, diabetes, hypertension, genitourinary infections, hypothyroidism, asthma, poor academic performance, smoking, drug use, and alcohol consumption.

[0212] In some embodiments, the methods of the present invention comprise calculating the body mass index (BMI).

[0213] In some embodiments, the disclosed methods for determining the probability of preterm birth involve detecting and / or quantifying one or more biomarkers using mass spectrometry, a capture agent, or a combination thereof.

[0214] In an additional embodiment, the disclosed method of determining the probability for preterm birth in a pregnant woman includes the initial step of providing a biological sample from the pregnant woman.

[0215] In some embodiments, the disclosed methods for determining the probability of preterm birth in a pregnant woman include communicating the probability to a healthcare provider. The disclosed methods for predicting GAB, predicting full-term birth, determining the probability of full-term birth in a pregnant woman, and determining the time to delivery in a pregnant woman similarly include communicating the probability to a healthcare provider. While described and exemplified above with reference to determining the probability of preterm birth in a pregnant woman, all embodiments described throughout this disclosure are equally applicable to methods for predicting GAB, predicting full-term birth, determining the probability of full-term birth in a pregnant woman, and determining the time to delivery in a pregnant woman. In particular, the biomarkers and panels listed throughout this application that explicitly reference methods for preterm birth can also be used in methods for predicting GAB, predicting full-term birth, determining the probability of full-term birth in a pregnant woman, and determining the time to delivery in a pregnant woman. It will be apparent to those skilled in the art that each of the above methods has specific and substantial utility and benefits related to maternal-fetal health considerations.

[0216] In additional embodiments, the communication informs subsequent treatment decisions for the pregnant woman. In some embodiments, the method of determining the probability of preterm birth in a pregnant woman includes an additional feature that expresses the probability as a risk score.

[0217] In the methods disclosed herein, determining the probability of preterm birth in a pregnant woman involves an initial step of forming a probability / risk index by measuring the ratio of isolated biomarkers selected from a group of cohorts of preterm and term pregnancies with known gestational age at birth. For an individual pregnancy, determining the probability of preterm birth in a pregnant woman involves measuring the ratio of the isolated biomarkers using the same measurement method as used in the initial step of generating the probability / risk index, and comparing the measured ratios of the risk index to derive an individualized risk for the individual pregnancy. In one embodiment, a preterm birth risk index is formed by measuring the ratio of IBP4 / SHBG in a cohort of preterm and term pregnancies with recorded gestational age at birth. Then, in clinical practice, the measured ratio of IBP4 / SHBG in the individual pregnancy is compared to the index using the same isolation and measurement techniques used to derive IBP4 / SHBG in the index group to derive the risk of preterm birth.

[0218] As used herein, the term "risk score" refers to a score that can be assigned based on comparing the amount or inverted value of one or more biomarkers in a biological sample obtained from a pregnant woman with a standard score or reference score that represents the average amount of one or more biomarkers calculated from biological samples obtained from a random pool of pregnant women. In some embodiments, the risk score is expressed as the inverted value, i.e., the logarithm of the ratio of the relative intensities of the individual biomarkers. Those skilled in the art will understand that risk scores can be expressed based on various data transformations as well as as the ratio itself. Furthermore, particularly with respect to inverted pairs, those skilled in the art will understand that any ratio will provide similarly useful information if the numerator and denominator biomarkers are swapped, or if a related data transformation (e.g., subtraction) is applied. Because biomarker levels may not be static throughout pregnancy, the standard score or reference score must be obtained for a gestational time point that corresponds to that of the pregnant woman when the sample is collected. The standard score or reference score can be determined in advance and incorporated into the predictive model, allowing the comparison to be indirect rather than being performed each time a probability is determined for a subject. The risk score can be a standard (e.g., a number) or a threshold (e.g., a line on a graph). The value of the risk score correlates with an upward or downward deviation from the average amount of one or more biomarkers calculated from biological samples obtained from a random pool of pregnant women. In certain embodiments, if the risk score is greater than the standard risk score or reference risk score, the pregnant woman may have an increased likelihood of preterm birth. In some embodiments, the magnitude of a pregnant woman's risk score, or an amount above the reference risk score, may indicate or correlate with the level of risk for that pregnant woman.

[0219] As exemplified herein, the PreTRM™ classifier identifies the IBP4 peptide transition (QCHPALDGQR_394.5_475.2 (SEQ ID NO: 2) ) and SHBG peptide transition (IALGGLLFPASNLR_481.3_657.4 (SEQ ID NO: 18) ) is defined as the natural logarithm of the SIS standardized intensity. Score = ln(P 1 n / P 2 n ) where P 1 n and P 2 n are the SIS-normalized peak area values ​​of the IBP4 and SHBG transitions, respectively. SIS normalization is defined as the relative ratio of the endogenous peak area divided by the corresponding SIS peak area. For example, P 1 n =P 1 e / P 1 SIS where P 1 e = IBP4 endogenous transition peak area, and P 1 SIS =IBP4 SIS transition peak area. By identifying the correlation between the distribution of PreTRM™ scores and the corresponding prevalence-adjusted positive predictive values, a probability of sPTB can be assigned to unknown subjects based on their score determinations. This relationship or association is shown in Figure 25 and links laboratory measurements to clinical predictions.

[0220] The PreTRM™ classifier identified IBP4 peptide transitions (QCHPALDGQR_394.5_475.2 (SEQ ID NO: 2) ) and SHBG peptide transition (IALGGLLFPASNLR_481.3_657.4 (SEQ ID NO: 18)While the present invention encompasses classifiers that include multiple inversions, improved performance can be achieved by constructing predictors formed from more than one inversion. Thus, in additional embodiments, the methods of the present invention include multiple inversions that demonstrate strong predictive performance, for example, for different GABD windows, preterm rupture of membranes (PPROM) versus preterm labor without PPROM (PTL), fetal gender, and primiparas versus multiparas. This embodiment is illustrated in Example 10 and Table 61 for inversions that produced strong predictive performance either early (e.g., 17-19 weeks) or late (e.g., 19-21 weeks) in the gestational age range. As illustrated, the performance of predictors formed from the combination of multiple inversions (SumLog) was evaluated across the entire blood draw range, and the predictor score was derived from the sum of the logarithm values ​​of the individual inversions (SumLog). One skilled in the art could select other models (e.g., logistic regression) to construct predictors formed from more than one inversion.

[0221] The methods of the present invention further include a classifier that includes an indicator variable that selects one or a subset of reversals based on known clinical factors, such as blood sampling period, fetal sex, parity, and any discriminatory patient characteristics and / or risk factors described throughout this application. This embodiment is exemplified in Example 10, Tables 61-64, which independently illustrate the reversal performance (17-21 weeks) of two distinct phenotypes of sPTB: sPTB, PPROM, and PTL. Similarly, this embodiment is exemplified in Example 10, Tables 76 and 77, and Figures 108 and 109, which independently illustrate the reversal performance (19-21 weeks) of two distinct phenotypes of sPTB: preterm premature rupture of membranes (PPROM) and preterm labor without PPROM (PTL). Thus, the methods of the present invention include selecting reversals to construct independent predictors of PPROM and PTL, or to maximize overall performance using a combination of more than one reversal of a single predictor, as described above. This embodiment is further illustrated in Example 10, Tables 65-68, which independently illustrate the reversal performance (17-21 weeks) of two different types of sPTB: primipara and multipara. This embodiment is further illustrated in Example 10, Tables 69-72, and Figure 106, which independently illustrate the reversal performance (17-21 weeks) of two different types of sPTB based on fetal gender. While the methods of the present invention are illustrated with respect to PPROM and PTL, parity, and fetal gender, they also include classifiers containing indicator variables that select one or a subset of reversals based on GABD, or any known clinical / risk factors described herein or otherwise known to those of skill in the art. As an alternative to having a classifier containing an indicator variable, the present invention further provides separate classifiers that fit subsets of pregnant women based on GABD, or any known clinical / risk factors described herein or otherwise known to those of skill in the art. For example, this embodiment encompasses separate classifiers for consecutive and / or overlapping time windows of GABD based on the best-performing reversal in each time window.

[0222] As exemplified herein, the predicted performance of the claimed method is 22 kg / m 2 and 37kg / m 2 BMI stratification can be improved by stratifying BMIs equal to or less than 100 mg / kg. Therefore, in some embodiments, the method of the present invention can be performed on samples obtained from pregnant women with a specified BMI. Briefly, BMI is an individual's weight in kilograms divided by the square of their height in meters. Although BMI does not directly measure body fat, some studies have shown that BMI correlates with more direct measurements of body fat obtained from skinfold thickness measurement, bioelectrical impedance analysis, densitometry (underwater weighing), dual-energy X-ray absorptiometry (DXA), and other methods. Furthermore, because BMI is a more direct measure of physical adiposity, it is believed to be strongly correlated with various metabolic and disease outcomes. Generally, individuals with a BMI below 18.5 are considered underweight, individuals with a BMI of 18.5 or greater to 24.9 are considered normal weight, individuals with a BMI of 25.0 or greater to 29.9 are considered overweight, and individuals with a BMI of 30.0 or greater are considered obese. In some embodiments, the predictive performance of the claimed methods can be improved by BMI stratification of 18 or greater, 19 or greater, 20 or greater, 21 or greater, 22 or greater, 23 or greater, 24 or greater, 25 or greater, 26 or greater, 27 or greater, 28 or greater, 29 or greater, or 30 or greater. In other embodiments, the predictive performance of the claimed methods can be improved by BMI stratification of 18 or less, 19 or less, 20 or less, 21 or less, 22 or less, 23 or less, 24 or less, 25 or less, 26 or less, 27 or less, 28 or less, 29 or less, or 30 or less.

[0223] In the context of the present invention, the term "biological sample" encompasses any sample collected from a pregnant woman and contains one or more of the biomarkers disclosed herein. Suitable samples in the context of the present invention include, for example, blood, plasma, serum, amniotic fluid, vaginal secretions, saliva, and urine. In some embodiments, the biological sample is selected from the group consisting of whole blood, plasma, and serum. In certain embodiments, the biological sample is serum. As will be understood by those skilled in the art, the biological sample may include any fraction or component of blood, including, but not limited to, T cells, monocytes, neutrophils, red blood cells, platelets, and microvesicles, such as exosomes and exosome-like vesicles. In certain embodiments, the biological sample is serum.

[0224] As used herein, the term "preterm birth" refers to birth or delivery at less than 37 completed weeks of gestation. Other commonly used subcategories of preterm birth have been established, delineating moderately preterm birth (birth at 33-36 weeks of gestation), very preterm birth (birth at <33 weeks of gestation), and extremely preterm birth (birth at ≤28 weeks of gestation). With respect to the methods disclosed herein, those skilled in the art will understand that the cutoffs delineating preterm birth from full-term birth, as well as the cutoffs delineating subcategories of preterm birth, can be adjusted when practicing the methods disclosed herein to, for example, maximize specific health benefits. In various embodiments of the present invention, the cutoff defining preterm birth includes, for example, birth at ≤ 37 weeks gestation, birth at ≤ 36 weeks gestation, birth at ≤ 35 weeks gestation, birth at ≤ 34 weeks gestation, birth at ≤ 33 weeks gestation, birth at ≤ 32 weeks gestation, birth at ≤ 30 weeks gestation, birth at ≤ 29 weeks gestation, birth at ≤ 28 weeks gestation, birth at ≤ 27 weeks gestation, birth at ≤ 26 weeks gestation, birth at ≤ 25 weeks gestation, birth at ≤ 24 weeks gestation, birth at ≤ 23 weeks gestation, or birth at ≤ 22 weeks gestation. In some embodiments, the cutoff defining preterm birth is ≤ 35 weeks gestation. It is further understood that such adjustments are well within the skill of those skilled in the art and are encompassed within the scope of the invention disclosed herein. Gestational age is a surrogate for the degree of fetal development and fetal readiness for birth. Gestational age is typically defined as the length of time from the date of the last normal menstrual period to the date of delivery. However, obstetric measures and ultrasound estimates can also be helpful in estimating gestational age. Preterm birth is generally classified into two separate subgroups. One is spontaneous preterm birth, which occurs following the spontaneous onset of preterm labor or premature rupture of membranes, regardless of subsequent labor augmentation or cesarean section. The other is medically indicated preterm birth, which occurs following induction or cesarean section for one or more conditions that female caregivers determine to be threatening to the health or life of the mother and / or fetus. In some embodiments, the method disclosed herein is directed to determining the probability of spontaneous preterm birth or medically indicated preterm birth. In some embodiments, the method disclosed herein relates to determining the probability of spontaneous preterm birth or medically indicated preterm birth.In some embodiments, the methods disclosed herein relate to determining the probability of spontaneous preterm birth. In additional embodiments, the methods disclosed herein relate to medically indicated preterm birth. In additional embodiments, the methods disclosed herein are directed to predicting pregnancy delivery.

[0225] As used herein, the term "estimated gestational age" or "estimated GA" refers to the GA determined based on the date of the last normal menstrual period and additional obstetric measures, ultrasound estimates, or other clinical parameters, including but not limited to those described in the previous section. In contrast, the term "predicted gestational age at birth" or "predicted GAB" refers to the GAB determined based on the inventive method disclosed herein. As used herein, "term birth" refers to birth at or beyond 37 full weeks of gestation.

[0226] In some embodiments, the pregnant woman is between 17 and 28 weeks gestation at the time the biological sample is collected (also referred to as GABR (gestational age at blood collection)). In other embodiments, the pregnant woman is between 16 and 29 weeks gestation, between 17 and 28 weeks gestation, between 18 and 27 weeks gestation, between 19 and 26 weeks gestation, between 20 and 25 weeks gestation, between 21 and 24 weeks gestation, or between 22 and 23 weeks gestation at the time the biological sample is collected. In further embodiments, the pregnant woman is between about 17 and 22 weeks gestation, between about 16 and 22 weeks gestation, between about 22 and 25 weeks gestation, between about 13 and 25 weeks gestation, between about 26 and 28 weeks gestation, or between about 26 and 29 weeks gestation at the time the biological sample is collected. Thus, the gestational age of the pregnant woman at the time the biological sample is collected can be 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, or 30 weeks. In certain embodiments, the biological sample is obtained at 19-21 weeks of gestation. In certain embodiments, the biological sample is collected at 19-22 weeks of gestation. In certain embodiments, the biological sample is collected at 19-21 weeks of gestation. In certain embodiments, the biological sample is collected at 19-22 weeks of gestation. In certain embodiments, the biological sample is collected at 18 weeks of gestation. In further embodiments, the best-performing inversions over consecutive or overlapping time windows can be combined into a single classifier to predict the probability of sPTB over a wider window of gestational age at the time of blood collection.

[0227] The term "amount" or "level," as used herein, refers to the amount of a biomarker that is detectable or measurable in a biological sample and / or control. The amount of a biomarker can be, for example, the amount of a polypeptide, the amount of a nucleic acid, or the amount of a fragment or surrogate. The term can alternatively include combinations thereof. The term "amount" or "level" of a biomarker is a measurable characteristic of that biomarker.

[0228] The present invention also provides a method for detecting an isolated pair of biomarkers selected from the group consisting of the biomarker pairs specified in any of Tables 1-77 and Figures 1-111 in a pregnant woman, the method comprising: a. obtaining a biological sample from the pregnant woman; b. detecting whether the isolated pair of biomarkers is present in the biological sample by contacting the biological sample with a first capture agent that specifically binds to the first member of the pair and a second capture agent that specifically binds to the second member of the pair; and detecting binding between the first biomarker of the pair and the second member of the pair and the second capture agent.

[0229] The present invention also provides a method for detecting an isolated pair of biomarkers selected from the group consisting of the biomarker pairs designated in Tables 27-59, 61-72, 76, and 77 in a pregnant woman, the method comprising: a. obtaining a biological sample from the pregnant woman; b. detecting whether the isolated pair of biomarkers is present in the biological sample by contacting the biological sample with a first capture agent that specifically binds to the first member of the pair and a second capture agent that specifically binds to the second member of the pair; and detecting binding between the first biomarker of the pair and the second member of the pair and the second capture agent.

[0230] The present invention also provides a method for detecting an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a pregnant woman, the method comprising: a. obtaining a biological sample from the pregnant woman; b. detecting whether the isolated pair of biomarkers is present in the biological sample by contacting the biological sample with a first capture agent that specifically binds to the first member of the pair and a second capture agent that specifically binds to the second member of the pair; and detecting binding between the first biomarker of the pair and the second member of the pair. In one embodiment, the present invention provides a method for detecting IBP4 and SHBG in a pregnant woman, comprising: a. obtaining a biological sample from the pregnant woman; b. contacting the biological sample with a capture agent that specifically binds IBP4 and a capture agent that specifically binds SHBG to detect whether IBP4 and SHBG are present in the biological sample; and c. detecting binding between IBP4 and the capture agent and between SHBG and the capture agent. In one embodiment, the method comprises measuring reversal values ​​of the pair of biomarkers. In a further embodiment, the presence of a change in reversal value between the pregnant woman and a full-term control indicates the likelihood of preterm birth in the pregnant woman. In one embodiment, the sample is obtained between 19 and 21 weeks of gestation. In a further embodiment, the capture agent is selected from the group consisting of an antibody, an antibody fragment, a nucleic acid-based protein-binding reagent, a small molecule, or a variant thereof. In a further embodiment, the method is performed by an assay selected from the group consisting of an enzyme immunoassay (EIA), an enzyme-linked immunosorbent assay (ELISA), and a radioimmunoassay (RIA).

[0231] The present invention provides a method for detecting an isolated pair of biomarkers selected from the group consisting of IBP4 / SHBG, VTNC / VTDB, VTNC / SHBG, CATD / SHBG, PSG2 / ITIH4, CHL1 / ITIH4, PSG2 / C1QB, PSG2 / FBLN3, HPX / IBP4, and HPX / PTGDS in a pregnant woman, the method comprising the steps of: a. obtaining a biological sample from the pregnant woman; and b. detecting whether the isolated pair of biomarkers is present in the sample, comprising subjecting the biological sample to a proteomic workflow consisting of mass spectrometry quantification.

[0232] In one embodiment, the present invention provides a method for detecting IBP4 and SHBG in a pregnant woman, comprising: a. obtaining a biological sample from the pregnant woman; and b. detecting whether the isolated pair of biomarkers is present in the sample, comprising subjecting the biological sample to a proteomic workflow consisting of mass spectrometry quantification.

[0233] A "proteomics workflow" generally involves one or more of the following steps: thawing a serum sample and depleting the 14 most abundant proteins by immunoaffinity chromatography. The depleted serum is digested with a protease, such as trypsin, to generate peptides. The digest is then enriched with a mixture of SIS peptides, desalted, and subjected to LC-MS / MS with a triple quadrapole instrument operated in MRM mode. A response ratio is generated from the area ratio of the endogenous peptide peak and the corresponding SIS peptide equivalent peak. Those skilled in the art will recognize that other types of MS, such as MALDI-TOF or ESI-TOF, can be used in the methods of the present invention. Additionally, those skilled in the art can modify the proteomics workflow by selecting specific reagents (such as proteases) or by omitting or changing the order of certain steps. For example, immunodepletion may not be necessary, SIS peptides may be added earlier or later, or stable isotope-labeled proteins may be used as standards instead of peptides.

[0234] Any existing, available, or conventional separation, detection, and quantification method can be used herein to measure the presence or absence (e.g., readout is present vs. absent; or detectable amount vs. undetectable amount) and / or amount (e.g., readout is absolute amount or relative amount, such as absolute concentration or relative concentration) of biomarkers, peptides, polypeptides, proteins, and / or fragments thereof, and optionally one or more other biomarkers or fragments thereof in a sample. In some embodiments, the detection and / or quantification of one or more biomarkers comprises an assay utilizing a capture agent. In further embodiments, the capture agent is an antibody, antibody fragment, nucleic acid-based protein-binding reagent, small molecule, or variant thereof. In additional embodiments, the assay is an enzyme immunoassay (EIA), enzyme-linked immunosorbent assay (ELISA), and radioimmunoassay (RIA). In some embodiments, the detection and / or quantification of one or more biomarkers further comprises mass spectrometry (MS). In yet a further embodiment, the mass spectrometry is co-immunoprecipitation-mass spectrometry (co-IP MS), in which co-immunoprecipitation, a technique suitable for the isolation of whole protein complexes, is followed by mass spectrometry.

[0235] As used herein, the term "mass spectrometer" refers to a device capable of volatilizing / ionizing analytes, forming gas-phase ions, and determining their absolute or relative molecular mass. Suitable methods of volatilization / ionization are matrix-assisted laser desorption / ionization (MALDI), electrospray, laser / light, thermal, electrical, atomization / nebulization, etc., or combinations thereof. Suitable forms of mass spectrometry include, but are not limited to, ion trap instruments, quadrupole instruments, electrostatic sector instruments, time-of-flight instruments, time-of-flight tandem mass spectrometers (TOF MS / MS), Fourier transform mass spectrometers, Orbitrap, and hybrid instruments composed of various combinations of these types of mass analyzers. These instruments can then be adapted with various other instruments to fractionate the sample (e.g., liquid chromatography or solid-phase adsorption techniques based on chemical or biological features) and ionize the sample for introduction into the mass spectrometer (including matrix-assisted laser desorption (MALDI), electrospray, or nanospray ionization (ESI), or combinations thereof).

[0236] Generally, any mass spectrometry (MS) technique (e.g., tandem mass spectrometry, MS / MS; or post-source decay, TOF MS) that can provide accurate information about peptide mass, preferably also about the fragmentation and / or (partial) amino acid sequence of selected peptides, can be used in the methods disclosed herein. Suitable peptide MS and MS / MS techniques and systems are well known per se (see, for example, Methods in Molecular Biology, Vol. 146: "Mass Spectrometry of Proteins and Peptides", Chapman, ed., Humana Press, 2000; Biemann, 1990, Methods Enzymol, Vol. 193: pp. 455-79; or Methods in Enzymology, Vol. 402: "Biological Mass Spectrometry", Burlingame, ed., Academic Press, 2005) and can be used in carrying out the methods disclosed herein. Thus, in some embodiments, the disclosed methods comprise performing quantitative MS to measure one or more biomarkers. Such quantitative methods can be performed in an automated (Villanueva et al., Nature Protocols (2006) 1(2):880-891) or semi-automated format. In certain embodiments, the MS can be operably linked to a liquid chromatography device (LC-MS / MS or LC-MS) or a gas chromatography device (GC-MS or GC-MS / MS). Other methods useful in this context include isotope-coded affinity tagging (ICAT), tandem mass tagging (TMT), or stable isotope labeling with amino acids in cell culture (SILAC), followed by chromatography and MS / MS.

[0237] As used herein, the terms "multiple reaction monitoring (MRM)" or "selected reaction monitoring (SRM)" refer to MS-based quantification methods that are particularly useful for quantifying low-abundance analytes. In an SRM experiment, a predefined precursor ion and one or more of its fragments are selected by two mass filters in a triple quadrupole instrument and monitored over time for accurate quantification. Multiple SRM precursor and fragment ion pairs can be measured within the same experiment on a chromatographic timescale by rapidly toggling between different precursor / fragment pairs to perform an MRM experiment. A series of transitions (precursor / fragment ion pairs) combined with the retention time of a target analyte (e.g., a peptide or small molecule, such as a chemical, steroid, hormone, etc.) can constitute the final assay. Multiple analytes can be quantified during a single LC-MS experiment. The terms "schedule" or "dynamic" in reference to MRM or SRM refer to a variation of the assay in which transitions for a particular analyte are acquired only in a time window surrounding the expected retention time, significantly increasing the number of analytes that can be detected and quantified in a single LC-MS experiment and contributing to the selectivity of the test. This is because retention time is a characteristic that depends on the physical properties of the analyte. A single analyte can also be monitored using more than one transition. Finally, a standard corresponding to the analyte of interest (e.g., the same amino acid sequence) but differing due to the inclusion of a stable isotope can be included in the assay. A stable isotope standard (SIS) can be incorporated into the assay at precise levels and used to quantify the corresponding unknown analyte. An additional level of specificity is contributed by the coelution of an unknown analyte and its corresponding SIS and the characteristics of their transitions (e.g., similarity in the ratio of the levels of two transitions in the unknown and the ratio of two transitions in the corresponding SIS).

[0238] Suitable mass spectrometry assays, instruments, and systems for biomarker peptide analysis include, but are not limited to, matrix-assisted laser desorption / ionization time-of-flight (MALDI-TOF) MS; MALDI-TOF post-source decay (PSD); MALDI-TOF / TOF; surface-enhanced laser desorption / ionization time-of-flight mass spectrometry (SELDI-TOF) MS; electrospray ionization mass spectrometry (ESI-MS); ESI-MS / MS; ESI-MS / (MS) n (n is an integer greater than zero); ESI 3D or linear (2D) ion trap MS; ESI triple quadrupole MS; ESI quadrupole orthogonal time-of-flight (Q-TOF); ESI Fourier transform MS system; desorption / ionization on silicon (DIOS); secondary ion mass spectrometry (SIMS); atmospheric pressure chemical ionization mass spectrometry (APCI-MS); APCI-MS / MS; APCI-(MS) n ;Ion mobility spectroscopy (IMS);Inductively coupled plasma mass spectrometry (ICP-MS);Atmospheric pressure photoionization mass spectrometry (APPI-MS);APPI-MS / MS, and APPI-(MS) nFragmentation of peptide ions in tandem MS (MS / MS) configurations can be achieved using techniques established in the art, such as collision-induced dissociation (CID). As described herein, detection and quantification of biomarkers by mass spectrometry can include, for example, multiple reaction monitoring (MRM), as described by Kuhn et al., Proteomics 4:1175-86 (2004), among others. Scheduled multiple reaction monitoring (scheduled MRM) mode acquisition during LC-MS / MS analysis enhances the sensitivity and accuracy of peptide quantification. Anderson and Hunter, Molecular and Cellular Proteomics 5(4):573 (2006). As described herein, mass spectrometry-based assays can be advantageously combined with upstream peptide or protein separation or fractionation methods, such as chromatography and other methods described herein below. As further described herein, shotgun quantitative proteomics can be combined with SRM / MRM-based assays for high-throughput identification and validation of prognostic biomarkers for preterm birth.

[0239] Those skilled in the art will appreciate that several methods can be used to determine the amount of a biomarker (including mass spectrometry approaches, such as MS / MS, LC-MS / MS, multiple reaction monitoring (MRM), or SRM, and product ion monitoring (PIM), as well as antibody-based methods, such as immunoassays, such as Western blot, enzyme-linked immunosorbent assay (ELISA), immunoprecipitation, immunohistochemistry, immunofluorescence, radioimmunoassay, dot blotting, and FACS). Thus, in some embodiments, determining the level of at least one biomarker comprises using an immunoassay and / or mass spectrometry. In additional embodiments, the mass spectrometry method is selected from MS, MS / MS, LC-MS / MS, SRM, PIM, and other such methods known in the art. In other embodiments, the LC-MS / MS further comprises 1D LC-MS / MS, 2D LC-MS / MS, or 3D LC-MS / MS. Immunoassay techniques and protocols are generally known to those skilled in the art (Price and Newman, Principles and Practice of Immunoassay, 2nd ed., Grove's Dictionaries, 1997, and Gosling, Immunoassays: A Practical Approach, Oxford University Press, 2000). A variety of immunoassay techniques can be used, including competitive and non-competitive immunoassays (Self et al., Curr. Opin. Biotechnol., 7:60-65 (1996)).

[0240] In further embodiments, the immunoassay is selected from Western blot, ELISA, immunoprecipitation, immunohistochemistry, immunofluorescence, radioimmunoassay (RIA), dot blotting, and FACS. In certain embodiments, the immunoassay is ELISA. In still further embodiments, the ELISA is direct ELISA (enzyme-linked immunosorbent assay), indirect ELISA, sandwich ELISA, competitive ELISA, multiplex ELISA, ELISPOT technology, and other similar techniques known in the art. The principles of these immunoassay methods are known in the art (e.g., John R. Crowther, The ELISA Guidebook, 1st Edition, Humana Press 2000, ISBN 0896037282). Typically, ELISAs are performed using antibodies, but they can be performed using any capture agent that can specifically bind to and detect one or more biomarkers of the present invention. Multiplexed ELISAs allow for the simultaneous detection of two or more analytes in a single compartment (e.g., a microplate well), usually at multiple array addresses (Nielsen and Geierstanger 2004, J Immunol Methods 290:107-20 (2004) and Ling et al., 2007, Expert Rev Mol Diagn 7:87-98 (2007)).

[0241] In some embodiments, radioimmunoassays (RIAs) can be used to detect one or more biomarkers in the methods of the invention. RIAs are competition-based assays that are well known in the art and utilize radioactive labels (e.g., 125 I or 131Radioimmunoassay involves mixing a known amount of target analyte (labeled) with an antibody specific for the analyte, then adding unlabeled analyte from the sample, and measuring the amount of labeled analyte displaced (see, e.g., An Introduction to Radioimmunoassay and Related Techniques, edited by Chard T, Elsevier Science 1995, ISBN 0444821198 for guidance).

[0242] Detectable labels can be used in the assays described herein for direct or indirect detection of biomarkers in the methods of the present invention. A variety of detectable labels can be used, and the choice of label depends on the required sensitivity, ease of conjugation with the antibody, stability requirements, and available equipment and disposal regulations. Those skilled in the art will be familiar with selecting an appropriate detectable label based on the assay detection of biomarkers in the methods of the present invention. Suitable detectable labels include, but are not limited to, fluorescent dyes (e.g., fluorescein, fluorescein isothiocyanate (FITC), Oregon Green™, rhodamine, Texas Red, tetrarhodimine isothiocyanate (TRITC), Cy3, Cy5, etc.), fluorescent markers (e.g., green fluorescent protein (GFP), phycoerythrin, etc.), enzymes (e.g., luciferase, horseradish peroxidase, alkaline phosphatase, etc.), nanoparticles, biotin, digoxigenin, metals, etc.

[0243] For mass spectrometry-based analysis, differential tagging with isotopic reagents (e.g., isotope-coded affinity tags (ICAT) or isobaric tagging reagents, iTRAQ (Applied Microbiology)) is used. More recent variations using tandem mass tags, TMT (Thermo Scientific, Rockford, IL) followed by multidimensional liquid chromatography (LC) and tandem mass spectrometry (MS / MS) analysis, can provide additional methodologies in carrying out the methods of the present invention.

[0244] Chemiluminescence assays using chemiluminescent antibodies can be used for sensitive, non-radioactive detection of protein levels. Antibodies labeled with fluorescent dyes may also be suitable. Examples of fluorescent dyes include, but are not limited to, DAPI, fluorescein, Hoechst 33258, R-phycocyanin, B-phycoerythrin, R-phycoerythrin, rhodamine, Texas Red, and Lissamine. Indirect labels include various enzymes well known in the art, such as horseradish peroxidase (HRP), alkaline phosphatase (AP), beta-galactosidase, urease, etc. Detection systems using appropriate substrates for horseradish peroxidase, alkaline phosphatase, and beta-galactosidase are well known in the art.

[0245] Signals from direct or indirect labels can be detected using, for example, a spectrophotometer to detect color from a chromogenic substrate; a radiation counter to detect radiation, for example 125 Analysis can be performed using a spectrophotometer, such as a gamma counter for the detection of I; or a fluorometer, which detects fluorescence in the presence of light of a specific wavelength. For detection of enzyme-linked antibodies, quantitative analysis can be performed using a spectrophotometer, such as the EMAX Microplate Reader (Molecular Devices; Menlo Park, Calif.), according to the manufacturer's instructions. If desired, the assays used to practice the present invention can be automated or performed robotically, allowing for simultaneous detection of signals from multiple samples.

[0246] In some embodiments, the methods described herein involve quantifying biomarkers using mass spectrometry (MS). In further embodiments, the mass spectrometry can be liquid chromatography-mass spectrometry (LC-MS), multiple reaction monitoring (MRM), or selected reaction monitoring (SRM). In additional embodiments, MRM or SRM can further include scheduled MRM or scheduled SRM.

[0247] As mentioned above, chromatography can also be used in carrying out the methods of the present invention. Chromatography encompasses methods for separating chemical substances and generally involves a process in which a mixture of analytes is carried by a moving stream of liquid or gas (the "mobile phase") and separated into components as they flow around or over a stationary liquid or solid phase (the "stationary phase"), resulting in differential distribution of the analytes between the mobile phase and the stationary phase. The stationary phase can typically be a finely divided solid, a sheet of filter material, or a thin film of liquid on the surface of a solid. Chromatography is well understood by those skilled in the art as a technique applicable for the separation of biologically derived chemical compounds, such as amino acids, proteins, protein fragments, or peptides.

[0248] The chromatography can be columnar (i.e., the stationary phase is deposited or packed in a column), preferably liquid chromatography, even more preferably high performance liquid chromatography (HPLC), or ultra-high pressure / high pressure liquid chromatography (UHPLC). Details of chromatography are well known in the art (Bidlingmeyer, Practical HPLC Methodology and Applications, John Wiley & Sons Inc., 1993). Exemplary types of chromatography include, but are not limited to, high performance liquid chromatography (HPLC), UHPLC, normal phase HPLC (NP-HPLC), reverse phase HPLC (RP-HPLC), ion exchange chromatography (IEC) (such as cation or anion exchange chromatography), hydrophilic interaction chromatography (HILIC), hydrophobic interaction chromatography (HIC), size exclusion chromatography (SEC) (including gel filtration or gel permeation chromatography), chromatofocusing, affinity chromatography (such as immunoaffinity, immobilized metal affinity chromatography, etc.), and the like. Chromatography, including single, two, or more dimensional chromatography, can be used as a method for further peptide analysis, for example, peptide fractionation in conjunction with downstream mass spectrometry as described elsewhere herein.

[0249] Additional peptide or polypeptide separation, identification, or quantification methods may be used to measure the biomarkers in the present disclosure, optionally in conjunction with any of the analytical methods described above. Such methods include, but are not limited to, chemical extraction partitioning, isoelectric focusing (IEF) (including capillary isoelectric focusing (CIEF), capillary isotachophoresis (CITP), capillary electrochromatography (CEC), etc.), one-dimensional polyacrylamide gel electrophoresis (PAGE), two-dimensional polyacrylamide gel electrophoresis (2D-PAGE), capillary gel electrophoresis (CGE), capillary zone electrophoresis (CZE), micellar electrokinetic chromatography (MEKC), free-flow electrophoresis (FFE), etc.

[0250] In the context of the present invention, the term "capture agent" refers to a compound capable of specifically binding to a target, particularly a biomarker. This term includes antibodies, antibody fragments, nucleic acid-based protein binding reagents (e.g., aptamers, Slow Off-rate Modified Aptamers (SOMAmers™)), protein capture agents, natural ligands (i.e., hormones for their receptors or vice versa), small molecules, or variants thereof.

[0251] The capture agent can be configured to specifically bind to a target, particularly a biomarker. The capture agent can include, but is not limited to, organic molecules such as polypeptides, polynucleotides, and other non-polymeric molecules identifiable by those skilled in the art. In the embodiments disclosed herein, the capture agent includes any agent that can be used to detect, purify, isolate, or enrich a target, particularly a biomarker. Any affinity capture technique known in the art can be used to selectively isolate and enrich / concentrate biomarkers that are components of complex mixtures of biological media for use in the disclosed methods.

[0252] Antibody capture agents that specifically bind to biomarkers can be prepared using any suitable method known in the art. See, for example, Coligan, Current Protocols in Immunology (1991); Harlow & Lane, Antibodies: A Laboratory Manual (1988); and Goding, Monoclonal Antibodies: Principles and Practice (2nd ed., 1986). Antibody capture agents can be any immunoglobulin or derivative thereof, whether natural or wholly or partially synthetically produced. All derivatives thereof that maintain specific binding ability are also included within the term. Antibody capture agents have binding domains that are homologous or largely homologous to immunoglobulin binding domains and can be derived from natural sources or partially or wholly synthetically produced. Antibody capture agents can be monoclonal or polyclonal. In some embodiments, the antibody is a single-chain antibody. Those skilled in the art will understand that antibodies can be provided in any of a variety of forms, including, for example, humanized, partially humanized, chimeric, chimeric-humanized, etc. The antibody capture agent may be an antibody fragment, including, but not limited to, Fab, Fab', F(ab')2, scFv, Fv, dsFv diabody, and Fd fragments. The antibody capture agent can be produced by any means. For example, the antibody capture agent can be enzymatically or chemically produced by fragmentation of an intact antibody and / or it can be recombinantly produced from a gene encoding a partial antibody sequence. The antibody capture agent can comprise a single-chain antibody fragment. Alternatively, or in addition, the antibody capture agent can comprise multiple chains linked together, for example, by disulfide bonds; and any functional fragment obtained from such molecules, wherein such fragment retains the specific binding characteristics of the parent antibody molecule. Due to their smaller size as a functional component of the whole molecule, antibody fragments can offer advantages over intact antibodies for use in certain immunochemical techniques and experimental applications.

[0253] Suitable capture agents useful for implementing the present invention also include aptamers. Aptamers are oligonucleotide sequences that can specifically bind to their targets through their unique three-dimensional (3-D) structure. Aptamers can contain any suitable number of nucleotides, and different aptamers can have the same or different numbers of nucleotides. Aptamers can be DNA or RNA or chemically modified nucleic acids, and can be single-stranded, double-stranded, or contain double-stranded regions and higher-dimensional structures. Aptamers can also be photoaptamers, in which a photoreactive or chemically reactive functional group is included in the aptamer so that it can covalently bind to its corresponding target. The use of aptamer capture agents can include the use of two or more aptamers that specifically bind to the same biomarker. Aptamers can include tags. Aptamers can be identified using any known method, including the SELEX (Systematic Evolution of Ligands by Exponential Enrichment) process. Once identified, aptamers can be prepared or synthesized according to any known method, including chemical and enzymatic synthesis, and used in various applications for biomarker detection. Liu et al., Curr Med Chem. 18(27):4117-25 (2011). Capture agents useful in carrying out the methods of the present invention also include SOMAmers (Slow Off-Rate Modified Aptamers), known in the art, which have improved off-rate characteristics. Brody et al., J Mol Biol. 422(5):595-606 (2012). SOMAmers can be generated using any known method, including the SELEX method.

[0254] Those skilled in the art will understand that biomarkers can be modified before analysis to improve their resolution or determine their identity. For example, biomarkers can be subjected to proteolytic digestion before analysis. Any protease can be used. Proteases that can cleave biomarkers into a number of distinct fragments, such as trypsin, are particularly useful. The fragments resulting from digestion serve as fingerprints for the biomarkers, thereby enabling their indirect detection. This is particularly useful when there are biomarkers with similar molecular weights that may be confused with the biomarker in question. Proteolytic fragmentation is also useful for high-molecular-weight biomarkers, because smaller biomarkers are more easily resolved by mass spectrometry. In another example, biomarkers can be modified to improve detection resolution. For example, neuraminidase can be used to remove terminal sialic acid residues from glycoproteins, improving binding to anionic adsorbents and improving detection resolution. In another example, biomarkers can be modified by attaching specific molecular weight tags that specifically bind to molecular biomarkers, further distinguishing them. Optionally, after detecting such modified biomarkers, the identity of the biomarkers can be further determined by matching the physical and chemical characteristics of the modified biomarkers in a protein database (e.g., SwissProt).

[0255] It is further understood in the art that biomarkers in a sample can be captured on a substrate for detection. Traditional substrates include antibody-coated 96-well plates or nitrocellulose membranes, which are then probed for the presence of proteins. Alternatively, protein-binding molecules attached to microspheres, microparticles, microbeads, beads, or other particles can be used to capture and detect biomarkers. Protein-binding molecules can be antibodies, peptides, peptoids, aptamers, small molecule ligands, or other protein-binding capture agents attached to the surface of particles. Each protein-binding molecule can contain an encoded, unique detectable label, which can be distinguished from other detectable labels attached to other protein-binding molecules and enable detection of biomarkers in multiplexed assays. Examples include, but are not limited to, color-coded microspheres with known fluorescence intensities (see, e.g., microspheres using xMAP technology produced by Luminex (Austin, Tex.)); microspheres containing quantum dot nanocrystals, e.g., with different ratios and combinations of quantum dot colors (see, e.g., Qdot nanocrystals produced by Life Technologies (Carlsbad, Calif.)); glass-coated metal nanoparticles (see, e.g., SERS nanotags produced by Nanoplex Technologies, Inc. (Mountain View, Calif.)); barcode materials (e.g., submicron-sized striped metal rods, such as Nanobarcode produced by Nanoplex Technologies, Inc.), coded microparticles with color barcodes (see, e.g., CellCard produced by Vitra Bioscience, vitrabio.com), glass microparticles with digital holographic code images (see, e.g., CyVera microbeads produced by Illumina (San Diego, Calif.)); chemiluminescent dyes, combinations of dye compounds; and detectable beads of different sizes.

[0256] In another embodiment, biochips can be used for capturing and detecting the biomarkers of the present invention. Many protein biochips are known in the art. These include, for example, protein biochips produced by Packard BioScience Company (Meriden Conn.), Zyomyx (Hayward, Calif.), and Phylos (Lexington, Mass.). Generally, protein biochips include a substrate having a surface. A capture reagent or adsorbent is attached to the surface of the substrate. Frequently, the surface includes multiple addressable locations, each of which has a capture agent bound thereto. The capture agent can be a biological molecule, such as a polypeptide or nucleic acid, that specifically captures other biomarkers. Alternatively, the capture agent can be a chromatographic material, such as an anion exchange material or a hydrophilic material. Examples of protein biochips are well known in the art.

[0257] The present disclosure also provides a method for predicting the probability of preterm birth, comprising measuring the change in the reversal value of a biomarker pair.For example, biological sample can be contacted with a panel comprising one or more polynucleotide binding agents.Then, the expression of one or more detected biomarkers can be evaluated by the method disclosed below, for example, with or without nucleic acid amplification.Those skilled in the art will recognize that the method described herein can automate the measurement of gene expression.For example, a system can be used that can perform multiplexed measurement of gene expression, for example, provide simultaneous digital measurement of the relative abundance of hundreds of mRNA species.

[0258] In some embodiments, nucleic acid amplification methods can be used to detect polynucleotide biomarkers. For example, the oligonucleotide primers and probes of the present invention can be used in combination with a variety of well-known and established methods (e.g., Sambrook et al., Molecular Cloning, A Laboratory Manual, pp. 7.37-7.57 (2nd ed. 1989); Lin et al., Diagnostic Molecular Microbiology, Principles and Applications, pp. 605-616 (Persing (1993); Ausubel et al., Current Protocols in Molecular Biology (20 The present invention can be used in amplification and detection methods that utilize nucleic acid substrates isolated by any of the methods described above. Methods for amplifying nucleic acids include, but are not limited to, polymerase chain reaction (PCR) and reverse transcription PCR (RT-PCR) (see, e.g., U.S. Pat. Nos. 4,683,195; 4,683,202; 4,800,159; and 4,965,188), ligase chain reaction (LCR) (see, e.g., Weiss, Science 254:1292-93 (1991)), strand displacement amplification (SDA) (see, e.g., Walker et al., Proc. Natl. Acad. Sci. USA 89:392-396 (1992); U.S. Pat. Nos. 5,270,184 and 5,455,166), thermophilic SDA (tSDA) (see, e.g., European Patent Nos. 0684 315), and methods described in U.S. Pat. No. 5,130,238; Lizardi et al., BioTechnol. 6:1197-1202 (1988); Kwoh et al., Proc. Natl. Acad. Sci. USA 86:1173-77 (1989); Guatelli et al., Proc. Natl. Acad. Sci. USA 87:1874-78 (1990); U.S. Pat. Nos. 5,480,784; 5,399,491; and U.S. Patent Application Publication No. 2006 / 46265.

[0259] In some embodiments, measuring mRNA in a biological sample can be used as a proxy for detecting the level of the corresponding protein biomarker in the biological sample. Thus, any of the biomarkers, biomarker pairs, or biomarker panels described herein can also be detected by detecting the appropriate RNA. mRNA levels can be measured by reverse transcription quantitative polymerase chain reaction (RT-PCR, followed by qPCR). RT-PCR is used to generate cDNA from mRNA. The cDNA can be used in a qPCR assay to produce fluorescence as the DNA amplification process progresses. By comparison with a standard curve, qPCR can produce absolute measurements, such as the number of mRNA copies per cell. Northern blots, microarrays, Invader assays, and RT-PCR combined with capillary electrophoresis have all been used to measure mRNA expression levels in samples. Gene Expression Profiling: Methods and See Protocols, edited by Richard A. Shimkets, Humana Press, 2004.

[0260] Some embodiments disclosed herein relate to diagnostic and prognostic methods for determining the probability of preterm birth in pregnant women.Detecting the expression level of one or more biomarkers and / or determining the ratio of biomarkers can be used to determine the probability of preterm birth in pregnant women.Such detection methods can be used for, for example, early diagnosis of conditions, determining whether a subject has a predisposition to preterm birth, monitoring the progress of preterm birth or the progress of treatment protocols, assessing the severity of preterm birth, predicting the outcome of preterm birth and / or the prospect of recovery or full-term birth, or helping to determine the appropriate treatment for preterm birth.

[0261] The quantification of biomarkers in biological samples can be determined by, but not limited to, the methods described above and any other methods known in the art.The quantitative data thus obtained is then subjected to analytical classification process.In this process, for example, as described in the examples provided herein, raw data is processed according to the algorithm that is predefined by the training set of data.The training set of data provided herein can be used in the algorithm, or the guideline provided herein can be used to generate the algorithm using different sets of data.

[0262] In some embodiments, analyzing measurable features to determine the probability of preterm birth in pregnant women involves the use of a predictive model. In further embodiments, analyzing measurable features to determine the probability of preterm birth in pregnant women involves comparing the measurable features to reference features. As those skilled in the art will appreciate, such comparisons can be direct comparisons with reference features or indirect comparisons where the reference features are incorporated into a predictive model. In further embodiments, analyzing measurable features to determine the probability of preterm birth in pregnant women involves one or more of a linear discriminant analysis model, a support vector machine classification algorithm, a recursive feature elimination model, a predictive analysis of microarray models, a logistic regression model, a CART algorithm, a flextree algorithm, a LART algorithm, a random forest algorithm, a MART algorithm, a machine learning algorithm, a penalized regression method, or a combination thereof. In certain embodiments, the analysis involves logistic regression.

[0263] In the analytical classification process, any one of a variety of statistical analysis methods can be used to manipulate the quantitative data and provide a classification of the sample. Examples of useful methods include linear discriminant analysis, recursive feature elimination, predictive analysis of microarrays, logistic regression, CART algorithm, FlexTree algorithm, LART algorithm, random forest algorithm, MART algorithm, machine learning algorithm, etc.

[0264] To create a random forest for predicting GAB, one skilled in the art can consider a set of k subjects (pregnant women) whose gestational age at birth (GAB) is known and whose blood samples (N analytes) have been measured in blood samples taken several weeks before delivery. The regression tree starts with a root node containing all subjects. The average GAB for all subjects can be calculated at the root node. The variance of GAB within the root node will be high because there is a mixture of women with different GABs. The root node is then partitioned into two branches, each containing women with similar GABs. The average GAB for subjects in each branch is again calculated. The variance of GAB within each branch will be lower than the root node because the subset of women in each branch has relatively more similar GABs than those in the root node. The two branches are created by selecting an analyte and a threshold for the analyte that creates a branch with similar GABs. Analytes and thresholds are chosen from among all sets of analytes and thresholds, typically with a random subset of analytes at each node. The procedure continues recursively generating branches, creating leaves (terminal nodes) where subjects have very similar GABs. The predicted GAB at each terminal node is the average GAB for the subjects at that terminal node. This procedure creates a single regression tree. A random forest can consist of hundreds or thousands of such trees.

[0265] Classification can be performed according to a predictive model method, which sets a threshold value to determine the probability that a sample belongs to a given class. The probability is preferably at least 50%, or at least 60%, or at least 70%, or at least 80% or higher. Classification can also be performed by determining whether a comparison between the obtained data set and a reference data set results in a statistically significant difference. If so, then the sample from which the data set was obtained is classified as not belonging to the class of the reference data set. Conversely, if such comparison is not statistically significantly different from the reference data set, then the sample from which the data set was obtained is classified as belonging to the class of the reference data set.

[0266] The predictive ability of a model can be evaluated according to its ability to provide a quality metric (e.g., AUROC (area under the receiver operating characteristic curve) or accuracy for a particular value or range of values). Area under the curve measurements are useful for comparing the accuracy of classifiers across the complete data range. A classifier with a larger AUC has a greater ability to accurately classify unknowns between two groups of interest. In some embodiments, the desired quality threshold is a predictive model that classifies samples with an accuracy of at least about 0.5, at least about 0.55, at least about 0.6, at least about 0.7, at least about 0.75, at least about 0.8, at least about 0.85, at least about 0.9, at least about 0.95, or higher. As an alternative measurement, the desired quality threshold can refer to a predictive model that classifies samples with an AUC of at least about 0.7, at least about 0.75, at least about 0.8, at least about 0.85, at least about 0.9, or higher.

[0267] As known in the art, the relative sensitivity and specificity of a predictive model can be adjusted to favor either the selection metric or the sensitivity metric, and the two metrics are inversely related.The constraints in the model described above can be adjusted to provide the selected sensitivity or specificity level, depending on the specific requirements of the test being carried out.One or both of sensitivity and specificity can be at least about 0.7, at least about 0.75, at least about 0.8, at least about 0.85, at least about 0.9 or higher.

[0268] Raw data can be analyzed by first measuring the value for each biomarker, usually three or more times. Data can be manipulated, for example, raw data can be transformed using a standard curve and the average value of three measurements used to calculate the mean and standard deviation for each patient. These values ​​can be transformed (e.g., logarithmic transformation, Box-Cox transformation) before being used in the model (Box and Cox, Royal Stat. Soc., Series B, 26:211-246 (1964)). Data is then input into a predictive model that classifies the sample according to its condition. The resulting information can be communicated to the patient or healthcare provider.

[0269] To generate a predictive model for preterm birth, a robust data set is used in training set, which includes known control samples and samples corresponding to the target preterm birth classification.Sample size can be selected using generally accepted standards.As discussed above, different statistical methods can be used to obtain highly accurate predictive models.An example of this analysis is provided in Example 2.

[0270] In one embodiment, hierarchical clustering is performed in deriving the predictive model, where Pearson correlation is used as the clustering metric. One approach is to consider the preterm birth dataset as a "training sample" in a "supervised learning" problem. CART is a standard in medical applications (Singer, Recursive Partitioning in the Health Sciences, Springer (1999)) and is used to convert arbitrary qualitative features into quantitative features; Hotelling et al. 2 They can be modified by sorting them according to the achieved significance level, assessed by sample reuse methods for statistics; and by appropriate application of the lasso method. Problems in prediction become problems in regression, without loss of predictive acuity, by making proper use of the Gini criterion for classification in assessing the quality of the regression.

[0271] This approach led to what is called FlexTree (Huang, Proc. Nat. Acad. Sci. USA 101:10529-10534 (2004)). FlexTree performs very well in simulations and when applied to multiple forms of data, and is useful for implementing the claimed method. Software has been developed to automate FlexTree. Alternatively, LARTree or LART can be used (Turnbull (2005) Classification Trees with Subset Analysis Selection by the Lasso, Stanford University). This name reflects the implementation of the lasso through a binary tree; the lasso described; and what Efron et al. (2004) Annals of Statistics 32:407-451 (2004) called LARS. See also Huang et al., Proc. Natl. Acad. Sci. USA 101(29):10529-34 (2004). Other analytical methods that can be used include logistic regression. One method of logistic regression: Ruczinski, Journal of Computational and Graphical Statistics 12:475-512 (2003). Logistic regression is similar to CART in that its classifier can be viewed as a binary tree. It differs in that each node has a Boolean description of the features, which is more general than the simple "and" description produced by CART.

[0272] Another approach is the nearest shrinking centroid approach (Tibshirani, Proc. Natl. Acad. Sci. USA 99:6567-72 (2002)). This technique is like k-means, but has the advantage of automatically selecting features by shrinking cluster centers, similar to lassoing, and focusing attention on a small number of informative ones. This approach is available as PAM software and is widely used. Two additional sets of algorithms that can be used are random forests (Breiman, Machine Learning 45:5-32 (2001)) and MART (Hastie, The Elements of Statistical Learning, Springer (2001)). These two methods are known in the art as "committee methods," in which predictors "vote" on the outcome.

[0273] To provide a significant ordering, a false discovery rate (FDR) can be determined. First, a set of null distributions of dissimilarity values ​​is generated. In one embodiment, the observed profile values ​​are permuted to create a distribution of a set of correlation coefficients obtained by chance, thereby creating an appropriate set of null distributions of correlation coefficients (Tusher et al., Proc. Natl. Acad. Sci. USA 98:5116-21 (2001)). The set of null distributions is obtained by permuting the values ​​of each profile of all available profiles; calculating pairwise correlation coefficients for all profiles; calculating the probability density function of the correlation coefficients for this permutation; and repeating this procedure N times (where N is a large number, typically 300). The N distribution is used to calculate an appropriate measure (such as the mean or median) of the count of correlation coefficient values ​​whose values ​​exceed the (similar) value obtained from the distribution of experimentally observed similar values ​​at a given significance level.

[0274] The FDR is the ratio of the expected number of falsely significant correlations (estimated from correlations greater than this selected Pearson correlation in the randomized data set) to the number of correlations greater than this selected Pearson correlation (significant correlation) in the empirical data. This cutoff correlation value can be applied to correlations between experimental profiles. The above distribution is used to choose a level of confidence for significance. This is used to determine the minimum correlation coefficient that exceeds the result that would be obtained by chance. This method is used to obtain a threshold for positive correlation, negative correlation, or both. Using this threshold, the user can filter pairwise correlation coefficient observations and eliminate those that do not exceed the threshold. Additionally, an estimate of the false positive rate can be obtained for a given threshold. For each individual "random correlation" distribution, it can be found how many observations fall outside the threshold range. This procedure provides a series of counts. The mean and standard deviation of the sequence provide the average number of potential false positives and their standard deviation.

[0275] In an alternative analytical approach, the variables selected in the cross-sectional analysis are used separately as predictors in a time-to-event analysis (survival analysis), in which the event is the occurrence of preterm birth and subjects without an event are considered censored at birth. Assuming a specific pregnancy outcome (preterm birth event or no event), a random length of time for each patient to be observed, and the selection of proteomic and other features, a parametric approach to analyzing survival may be better than the widely applied semiparametric Cox model. The Weibull parametric fit of survival allows the hazard ratio to be monotonically increasing, decreasing, or constant, and has a proportional hazards representation (as in the Cox model) and an accelerated failure time representation. All standard tools available for obtaining approximate maximum likelihood estimators of regression coefficients and corresponding functions are available in this model.

[0276] In addition, Cox model can be used.In particular, because the number of covariates is reduced to a manageable size by lasso, analysis is significantly simplified, and allows the possibility of non-parametric or semi-parametric approach to predicting the time to preterm birth.These statistical tools are known in the art and can be applied to all types of proteomics data.It can be easily determined, and provides a set of biomarkers, clinical data and genetic data that are highly informative about the probability of preterm birth in pregnant women and the predicted time to preterm birth.Algorithm also provides information about the probability of preterm birth in pregnant women.

[0277] Therefore, those skilled in the art will understand that the probability of preterm birth according to the present invention can be determined using either quantitative or categorical variables. For example, when carrying out the method of the present invention, the measurable characteristics of each of the N biomarkers can be subjected to categorical data analysis to determine the probability of preterm birth as a binary categorical outcome. Alternatively, the method of the present invention can analyze the measurable characteristics of each of the N biomarkers by first calculating quantitative variables, particularly the predicted gestational age at birth. The predicted gestational age at birth can then be used as the basis for predicting the risk of preterm birth. By first using quantitative variables and then converting them into categorical variables, the method of the present invention takes into account the continuous measurements detected for the measurable characteristics. For example, rather than making a binary prediction of preterm birth versus full-term birth, predicting the gestational age at birth makes it possible to individualize treatment for pregnant women. For example, a predicted gestational age at birth that is earlier will likely result in more intensive prenatal intervention, i.e., monitoring and treatment, than a predicted gestational age that is closer to full term.

[0278] Among women with a predicted GAB of j days plus or minus k days, p(PTB) can be estimated as the proportion of women in the PAPR clinical trial with a predicted GAB of j days plus or minus k days who actually deliver before 37 weeks' gestation (see Example 1). More generally, for women with a predicted GAB of j days plus or minus k days, the probability that the actual gestational age at birth will be less than the specified gestational age, p(actual GAB<specified GAB), was estimated as the proportion of women in the PAPR clinical trial with a predicted GAB of j days plus or minus k days who actually deliver before the specified gestational age.

[0279] In developing a predictive model, it may be desirable to select a subset of markers, i.e., at least 3, at least 4, at least 5, or at least 6, up to a complete set of markers. Typically, a subset of markers is chosen that provides the necessary features for quantitative sample analysis, such as reagent availability and ease of quantification, while maintaining a highly accurate predictive model. The selection of several informative markers for building a classification model requires the definition of a performance metric and a user-defined threshold for producing a model with useful predictive ability based on this metric. For example, the performance metric can be the AUC, the sensitivity and / or specificity of the prediction, and the overall accuracy of the predictive model.

[0280] As will be understood by those skilled in the art, analytical classification process can use any one of a variety of statistical analysis methods to manipulate quantitative data and provide sample classification.Examples of useful methods include, but are not limited to, linear discriminant analysis, recursive feature elimination, microarray predictive analysis, logistic regression, CART algorithm, FlexTree algorithm, LART algorithm, random forest algorithm, MART algorithm, and machine learning algorithm.Various methods are used in training models.Selection of marker subsets can be forward selection or backward selection of marker subsets.The number of markers that optimizes model performance can be selected without using all markers.One method for defining the optimal number of items is to use any combination and number of items used for a given algorithm and select the number of items that produces a model with the desired predictive ability (for example, AUC>0.75, or equivalent measurement of sensitivity / specificity), with a standard error of 1 or less from the maximum value obtained for this metric.

[0281] In yet another aspect, the present invention provides a kit for determining the probability of preterm birth. The kit can include one or more agents for detecting biomarkers, a container for holding a biological sample isolated from a pregnant woman, and printed instructions for reacting the agent with the biological sample or a portion of the biological sample to detect the presence or amount of the isolated biomarker in the biological sample. The agents can be packaged in separate containers. The kit can further include one or more control reference samples and reagents for performing immunoassays.

[0282] The kit can include one or more containers for the compositions contained in the kit. The compositions can be in liquid form or can be lyophilized. Suitable containers for the compositions include, for example, bottles, vials, syringes, and test tubes. The containers can be made of various materials, including glass or plastic. The kit can also include a package insert containing instructions on how to determine the probability of preterm birth.

[0283] From the foregoing, it will be apparent that variations and modifications can be made to the invention described herein to adapt it to various usages and conditions, and such embodiments are within the scope of the following claims.

[0284] The recitation of a list of elements in any definition of a variable herein includes definitions of that variable as any single element or combination (or subcombination) of the listed elements. The recitation of an embodiment herein includes that embodiment as any single embodiment or in combination with any other embodiment or portion thereof.

[0285] All patents and publications mentioned in this specification are herein incorporated by reference to the same extent as if each individual patent or publication was specifically and individually indicated to be incorporated by reference.

[0286] The following examples are offered by way of illustration, but not by way of limitation. [Example]

[0287] Example 1: Development of a sample set for biomarker discovery and validation for preterm birth A standardized protocol was developed to govern the conduct of the Proteomic Assessment of Preterm Risk (PAPR) clinical study. Specimens were obtained from women at 11 Institutional Review Board (IRB)-certified sites across the United States. After providing informed consent, serum and plasma samples were obtained, along with relevant information regarding patient demographics, past medical and reproductive history, current pregnancy, and concomitant medications. Following delivery, data related to maternal and infant status and complications were collected. Serum and plasma samples were processed according to a protocol calling for standard refrigerated centrifugation, sample aliquoting into 2-D barcoded cryovials, and subsequent freezing at -80°C.

[0288] Following delivery, preterm birth cases were individually reviewed and their status determined as either spontaneous or medically indicated preterm birth. Only spontaneous preterm birth cases were used for this analysis. To discover biomarkers of preterm birth, serum samples from 86 preterm birth cases and 172 controls, ranging in gestational age at blood collection (GABD) from 17 weeks, 0 days (17.0) to 28 weeks, 6 days (28.6), were analyzed. A separate sample set was also analyzed for corroboration. This set consisted of serum from 50 preterm birth cases and 100 controls spanning the same gestational age range. For each case, two GABD-matched controls were selected from a panel of several randomly generated controls matched to the distribution of births reported in the 2012 National Vital Statistics Report. A protocol was in place to ensure that laboratory personnel were blinded to the gestational age at delivery and case-control status of the subjects used in both sample sets. Informatics staff were also blinded to the corroborating sample set until analytical analysis of the samples was completed.

[0289] Serum samples were depleted of highly abundant proteins using the Human 14 Multiple Affinity Removal System (MARS14). This removes the 14 most abundant proteins, which are considered uninformative for identifying disease-related changes in the serum proteome. To this end, an equal volume (50 μl) of each clinical pooled human serum sample (HGS) or pooled human pregnant female serum sample (pHGS) was diluted with 150 μl of Agilent column buffer A and filtered through a Captiva filter plate to remove precipitates. The filtered samples were depleted using a MARS-14 column (4.6 × 100 mm, catalog number 5188-6558, Agilent Technologies) according to the manufacturer's protocol. The samples were cooled to 4°C in the autosampler, the depletion column was run at room temperature, and the collected fractions were kept at 4°C until further analysis. The unbound fraction was collected for further analysis.

[0290] Depleted serum samples were reduced with dithiothreitol and alkylated using iodoacetamide. They were then digested with 5.0 μg of Trypsin Gold-Mass Spec Grade (Promega) for 17 hours (±1 hour) at 37°C. After trypsin digestion, a mixture of 187 stable isotope standard (SIS) peptides was added to the samples, and half of each sample was desalted on an Empore C18 96-well solid-phase extraction plate (3M Bioanalytical Technologies). The plate was pretreated according to the manufacturer's protocol. Peptides were washed with 300 μl of 1.5% trifluoroacetic acid, 2% acetonitrile, eluted with 250 μl of 1.5% trifluoroacetic acid, 95% acetonitrile, frozen at -80°C for 30 minutes, and then lyophilized. The lyophilized peptides were reconstituted in 2% acetontile / 0.1% formic acid containing three non-human internal standard (IS) peptides. Peptides were purified using Agilent Poroshell Separation was carried out on a 120EC-C18 column (2.1 x 100 mm, 2.7 µm) using a 30 minute acetonitrile gradient of 400 µl / min at 40°C and injected into an Agilent 6490 triple quadrupole mass spectrometer.

[0291] The depleted and trypsin-digested samples were analyzed using scheduled multiple reaction monitoring (sMRM). The sMRM assay monitored 898 transitions, representing 259 biological peptides and 190 IS peptides (187 SIS + 3 IS), representing 148 proteins. Chromatographic peaks were integrated using Mass Hunter Quantitative Analysis software (Agilent Technologies).

[0292] Data analysis

[0293] Analysis of the discovery and corroboration sample data was performed in two stages. In the first stage, robust biomarkers were identified by selection using the discovery samples and validation using an independent corroboration sample set. In the second stage, the combined discovery and corroboration data were used to identify the best analytes and panels of analytes for developing classifiers.

[0294] Phase I: Blinded analysis

[0295] Initial classifier development focused on gestational ages 17.0-25.6. Using discovery samples, a set of peptides corresponding to 62 proteins was selected based on preanalytical and analytical criteria. Diagnostic performance of analytes was assessed in a series of narrow GABD windows spanning 3 weeks with 2 weeks of overlap between adjacent windows. Based on consistency of diagnostic performance (up- and down-regulation of cases vs. controls across the GABD), a subset of 43 analytes was selected for further analysis.

[0296] For each narrow GABD window, a set of inversions was formed using all combinations of up- and down-regulated analytes within the narrow window. The inversion value refers to the ratio of the relative peak area of ​​the up-regulated analyte to the relative peak area of ​​the down-regulated analyte, and serves both to normalize variability and to amplify the diagnostic signal. From all possible inversions within the narrow window, a subset was selected based on individual univariate performance (AUC >= 0.6).

[0297] For each window, inversion panels of various sizes were formed (sizes 2, 3, 4, 6, and 8). Monte Carlo cross-validation (MCCV) was performed by training and testing a logistic classifier 1,000 times for each panel size within the window on 70% and 30% of the samples, respectively. The size 4 panel, determined to be optimal by average MCCV AUC, was then used to identify inversion candidates that performed well in the panel. Inversion candidates were identified by their frequency of appearance in the top-performing logistic classifiers for size 4 panels in the MCCV analysis. For each window, three sets of inversion frequency tables were created using performance measures of either AUC or partial AUC (pAUC) for sensitivity ranging from 0.7 to 1, or the correlation between the classifier output score and the time-to-birth value (TTB) (the difference in days between GABD and gestational age at birth). From each of these inversion lists, the top 15 inversions were selected for further analysis.

[0298] For each GABD narrow window, inversion panels of sizes 2, 3, and 4 were formed from each of the three lists (AUC, pAUC, and TTB) based on the performance of the MCCV analysis. The top 15 panels for each panel size in each window were selected. These top 15 panels of sizes 2, 3, and 4, along with the top 15 inversions from each of the three lists (AUC, pAUC, and TTB) in each window, were used to train a logistic classifier on the discovery samples, and classification scores for the support samples were generated in a blinded manner.

[0299] The performance of all inversion and classifier panels was assessed by an independent statistician. ROC curve AUC, pAUC, and TTB correlation of classifier scores were reported.

[0300] Phase II: Open-label analysis

[0301] After unblinding, the discovery and confirmation datasets were combined and reanalyzed. Because diagnostic protein expression can vary across pregnancy, we examined protein levels as a function of GABD. A median smoothing window of + / - 10 days was applied to generate kinetic plots. Relative protein levels were expressed as the ratio of the endogenous peptide peak area to its corresponding SIS standard (relative ratio). Examples of proteins that increase with pregnancy but show no differential levels in PTB cases and controls are shown in Figures 3, 4, and 10. Measuring the levels of such proteins can be useful for accurate gestational age determination (e.g., pregnancy "clocks"). Pregnancy clocks predict gestational age from the relative abundance of one or more proteins (transitions). Alternatively, in this same analysis, we identified proteins whose levels vary across the GABD but show differential levels between PTB cases and controls (Figure 5). These proteins are obvious diagnostic candidates for developing PTB classifiers. We also illustrated the impact of inversion formation using the ratio of overexpressed proteins to the ratio of underexpressed proteins (Figures 8 and 21). This clearly resulted in increased separation of PTB cases and controls. Previous analyses have suggested that the levels of some analytes may be affected by pre-pregnancy body mass index (BMI). CLIN. CHEM., Vol. 37 / No. 5, pp. 667-672 (1991); European Journal of Endocrinology (2004), Vol. 150, pp. 161-171. Therefore, we investigated the impact of BMI on separation by displaying the inversion values ​​across pregnancy only for patients whose BMI was less than 35 (Figure 21). This resulted in further improvement in separation.

[0302] Inversion selection and classifier development for the combined discovery and validation datasets were similar to previous studies. We illustrated the analysis by focusing on the third overlapping GABD window (days 133–153). MCCV analysis was performed to identify inversion candidates. To evaluate the performance of the panel, the inversion values ​​were combined with a simple LogSum classifier. The LogSum classifier assigns a score to each sample based on the sum of the logarithms of the relative ratio values ​​of each inversion for that sample. This type of classifier has no coefficients, which helps avoid overfitting problems. Anyone skilled in the art can derive a similar logistic classifier using the same analytes using well-established techniques. The multivariate performance of the panel of three top inversions formed from four proteins is shown in Figure 8 as a histogram of AUC values ​​obtained by cross-validation and as a ROC curve. Previous analysis has suggested that the levels of some analytes may be affected by pre-pregnancy body mass index (BMI).

[0303] The inventors have determined proteins and / or inversions that are strong predictors of time to birth (TTB), exemplified herein by the use of ITIH4 / CSH (FIG. 10). TTB is defined as the difference between GABD and gestational age at birth (GAB). This allows for the prediction of clinical estimates of TTB (or GAB), either individually or in mathematical combinations of such analytes.

[0304] Example 2 Validation of IBP4 / SHBG sPTB predictor This example demonstrates validation of IBP4 / SHBG sPTB predictors identified in a large-scale maternal serum proteomics effort in asymptomatic women in early pregnancy.

[0305] subject

[0306] The Proteomic Assessment of Preterm Birth Risk (PAPR) study was conducted under standard protocols at 11 Institutional Review Board (IRB)-approved sites across the United States (Clinical Trial Government Identification Number: NCT01371019). Subjects aged 17 0 / 7 to 28 6 / 7 weeks gestational age were enrolled. Best clinical estimates of gestational age were prepared using a predefined protocol of menstrual day counts confirmed by early ultrasound biometrics or ultrasound alone. Body mass index (BMI) was derived from height and self-reported prepregnancy weight. Multiple pregnancies and pregnancies with known or suspected significant fetal anomalies were excluded. Relevant information on subject demographics, past medical and reproductive history, current pregnancy history, and concomitant medications was collected and entered into an electronic case report form. Postpartum, maternal and infant outcomes and complications were collected. All deliveries were adjudicated as term (≥37 0 / 7 weeks GA), spontaneous preterm (including preterm premature rupture of membranes), or medically indicated preterm. Discrepancies were resolved by the study site investigator, as indicated. Adjudication was completed and data locked prior to the validation study.

[0307] Sample collection

[0308] Maternal blood was collected and processed as follows: after a 10-minute room temperature clotting period, it was either immediately refrigerated and centrifuged or placed in an ice-water bath at 4–8°C until centrifugation. Blood was centrifuged within 2.5 hours of collection, and 0.5 ml serum aliquots were stored at −80°C until analysis.

[0309] Principles for developing predictors

[0310] Development of the IBP4 / SHBG predictor involved independent and sequential discovery, validation, and validation steps in accordance with the Institute of Medicine (IOM) guidelines for best practices in "omics" research. IOM, Evolution of Translation Omics: Lessons Learned and the Path Forward, (Micheel CM, Nass SJ, Omenn GS, eds.), Washington, DC: The National Academies Press; 2012: pp. 1-355. Analytical validation preceded clinical validation sample analysis. Analytical validation included assessment of inter- and intra-batch precision, carryover, and limits of detection.

[0311] A validation nested case / control analysis was performed on prespecified sPTB case and control samples unrelated to discovery and validation. The sPTB cases included samples from a total of nine sites, with two sites unique to validation. Validation cases and controls underwent 100% on-site document matching with each subject's medical record prior to mass spectrometry (MS) serum analysis. This process ensured that all subjects met the inclusion and exclusion criteria and confirmed medical / pregnancy complications and GA at birth for all subjects at the time of sample collection and delivery. A detailed analysis protocol, including the validation study design, analysis plan, and blinding protocol, was established in advance. Personnel other than the Chief Clinical Officer (DCO) and clinical data manager were blinded to the subject case, control, and GA at birth data assignments. The data analysis plan included a prespecified validation claim and a protocol for dual independent external analysis. Predictor scores, calculated as described below, were determined for all subject samples by a blinded statistician. Case, control, and GA data correlated with predictor scores by the DCO were subjected to independent external statistical analysis. Area under the receiver operating characteristic curve (AUROC) and significance test results were then transferred to the DCO. Data transfer incorporated the use of the SUMPRODUCT function (Microsoft, Microsoft Excel 2013) to ensure data integrity was maintained. Real-time digital time stamping was applied to analysis data, plans, and reports to provide a data audit trail from each subject to validation results.

[0312] Validation study design

[0313] In the primary analysis, sPTB cases were defined as subjects who delivered due to preterm premature rupture of membranes (PPROM) or spontaneous onset of labor at <37 0 / 7 weeks GA. Controls were subjects who delivered at ≥37 0 / 7 weeks GA. Previous findings and supporting analyses included a wide range of gestational ages (17 Forty-four candidate biomarkers were investigated using serum samples collected over a period ranging from 0 / 7 to 25 6 / 7 weeks GA (Supplementary Materials). Discovery and validation identified two proteins (IBP4 and SHBG) used in the optimal narrow GA range of blood collection (19 0 / 7 to 21 6 / 7 weeks) and the ratio (IBP4 / SHBG) as the best predictor of sPTB by AUROC (Supplementary Materials). Discovery and validation also identified subjects with less extreme BMI values ​​as having improved classification performance by IBP4 / SHBG (Supplementary Results). Following discovery and validation analyses, we moved on to analytical and clinical validation.

[0314] Validation sPTB cases were 18 subjects collected at blood collection time GA (GABD) between 19 0 / 7 and 21 6 / 7 weeks from a total of 81 subjects available over GA between 17 0 / 7 and 28 6 / 7 weeks. A set of controls, including two controls per GABD-matched sPTB case, was analyzed using the R statistical program (R 3.0.2) (Team RC., R: a Language and Environment for Statistical Computing., Vienna, Austria; 2014). Year, 2015; Matei A, Tille Y., The R "sampling" package., European Conference on Quality in Survey Statistics., 2006) and randomly selected using chi-square tests, based on the 2012 National Vital Statistics Reports (Martin JA, Hamilton BE, Osterman MJ, Curtin SC, Mathews TJ., Births: Final Data for 2012., National Vital Statistics Reports., 2014;Vol. 63(09):1-86 The results were compared with the distribution of term deliveries outlined in [page 1]. Randomly generated control sets (10 in each group) were examined for sets that yielded p-values ​​close to 1.0.

[0315] The primary objective was to validate the performance of the IBP4 / SHBG ratio as a predictor of sPTB using AUROC (Team RC., R: a Language and Environment for Statistical Computing., Vienna, Austria; 2014, 2015; Sing T, Sander O, Beerenwinkel N, Lengauer T., ROCR: visualizing classifier performance in R. Bioinformatics., 2005; 21(20):7881). Overall To control for the significant multiple testing error rate (α = 0.05), a fixed order method (Dmitrienko A, Tamhane AC, Bretz F, eds., Multiple Testing Problems in Pharmaceutical Statistics. Boca Raton, Florida: CRC Press; 2009: pp. 1–320; Dmitrienko A (Bamber D., The area above the mean mean (GA) of 19 0 / 7 to 21 6 / 7 weeks was used to estimate the mean mean (GA) of 19 0 / 7 to 21 6 / 7 weeks of GA. The Wilcoxon-Mann-Whitney statistic was used to test for equivalence to AUROC=0.5 (by chance). ordinal dominance graph and the area below the receiver operating characteristic graph., Journal of mathematical psychology., 1975; Vol. 12 (Issue 4): pp. 387-415, doi:10.1016 / 0022-2496(75)90001-2; Mason SJ, Graham NE., Areas beneath the relative operating characteristics (ROC) and relative operating levels (ROL) curves: Statistical significance and interpretation., QJR Meteorol Soc., 2002; vol. 128 (issue 584): 2145~2 166 pages, doi:10.1256 / 003590002320603584.) GA boundaries other than <37 0 / 7 vs. ≥37 0 / 7 weeks GA (e.g., <36 0 / 7 vs. ≥36 0 / 7, <35 0 / 7 vs. ≥35 To determine classification performance at 0 / 7), cases and controls were redefined as all subjects below and at / above a specified boundary, respectively.

[0316] Laboratory Methods

[0317] A systems biology approach was used to generate a highly multiplexed multiple reaction monitoring (MRM) MS assay (Supplementary Methods and Results). The validation assay quantified proteotypic peptides specific for the predictor proteins IBP4 and SHBG, as well as other controls. Samples were processed in 32 batches, consisting of 24 clinical subjects, three pooled serum standards derived from healthy non-pregnant donors (HGS), three pooled serum standards derived from healthy pregnant donors (pHGS), and two phosphate-buffered saline solutions, which served as treatment controls. For all analyses, serum samples were first depleted of high-abundance, non-diagnostic proteins using a MARS-14 immunodepletion column (Agilent Technologies), reduced with dithiothreitol, alkylated with iodoacetamide, and digested with trypsin. Heavy-labeled stable isotope standard (SIS) peptides were then added to the samples, which were then desalted and analyzed by reversed-phase liquid chromatography (LC) / MRM-MS. SIS peptides were used for normalization by generating a response ratio (RR) by dividing the peak area of ​​the peptide fragment ion (i.e., transition) measured in serum by the peak area of ​​the corresponding SIS transition spiked into the same serum sample.

[0318] IBP4 / SHBG predictors

[0319] The predictor score was defined as the natural logarithm of the ratio of the IBP4 peptide transition response ratio and the SHBG peptide transition response ratio:

number

[0320] result

[0321] Figure 23 summarizes the distribution of PAPR study subjects. Between March 2011 and August 2013, 5,501 subjects were enrolled. 410 (6.7%) subjects received progestogen therapy after the first trimester and were therefore excluded from analysis, as prespecified in the protocol. An additional 120 (2.2%) subjects were excluded due to early discontinuation, and 146 (2.7%) were lost to follow-up. A total of 4,825 subjects were available for analysis. There were 533 cases of PTB: 248 (4.7%) spontaneous preterm births and 285 (5.9%) medically indicated preterm births. Subjects with sPTB were more likely to have one or more prior PTBs and to experience bleeding after 12 weeks of gestation in the study pregnancy compared with subjects who delivered at term (Table 1). Characteristics of sPTB cases and term controls selected for validation did not differ significantly from each other, except for significantly more Hispanic controls (47.5% vs. 33.3%, p = 0.035). Similarly, subjects selected for validation were broadly representative of the study cohort as a whole, with the exception of the ethnicity of term controls (Table 1).

[0322] Verification Analysis

[0323] Discovery and supportive analyses identified the IBP4 / SHBG ratio and the GA interval of 19 0 / 7 to 21 6 / 7 weeks as the best-performing predictors of sPTB, based on AUROC and GA interval, respectively (Supplementary Results, below). For validation, IBP4 / SHBG predictors were tested with and without BMI stratification using a predefined fixed-order approach. Optimal performance was identified for the GA interval of 19 1 / 7 to 20 6 / 7 weeks. When BMI was not considered, the validated performance was AUROC = 0.67 (p = 0.02) (Supplementary Results). However, as expected, the GA intervals of >22 and ≤37 kg / m 2BMI stratification improved performance. This corresponded to an AUROC of 0.75 (p = 0.016, 95% CI 0.56-0.91) (Figure 24). A more detailed characterization of BMI stratification can be found in the Supplementary Results. Performance measures of sensitivity, specificity, AUROC, and odds ratio (OR) were determined at various case-to-control boundaries (Table 2). For sPTB versus term birth (<37 0 / 7 vs. ≥37 0 / 7 weeks), sensitivity and specificity were 0.75 and 0.74, respectively, and the odds ratio (OR) was 5.04 (95% CI 1.4-18). Results at other boundaries are summarized in Table 2. Test accuracy improved at lower GA boundaries.

[0324] The prevalence-adjusted positive predictive value (PPV), a measure of clinical risk, is shown as a function of predictor score in Figure 25. The PPV was calculated based on the background value (population sPTB rate of singleton births in the US is 7.3%) (Martin et al., Births: final data for 2013. Natl Vital Stat Rep., 2015; 64(1): 1-65 Martin JA, Hamilton BE, Osterman MJ, Curtin SC, Matthews TJ., Births: final data for 2013., Natl. Vital Stat Rep., 2015;64(1):1-65) 2×(14. Stratification of subjects with increasing predictor scores occurred with increasing relative risk to 6%) and 3× (21.9%) (dotted lines) and above (FIG. 25). The distribution of IBP4 / SHBG predictor score values ​​for subjects color-coded by GA category at delivery is shown in the box plots in FIG. 25. The earliest sPTB cases (<35 0 / 7 weeks GA) had higher predictor scores than late term controls (≧39 0 / 7 weeks GA), whereas late sPTB cases (≧35 0 / 7 to <37 0 / 7 weeks GA) had higher predictor scores than early term controls (≧37 The risk overlapped with those of 0 / 7 to <39 0 / 7 weeks GA (Figure 25). Subjects were identified as high or low risk according to the predictor score cutoff corresponding to 2× relative risk (PPV of 14.6%). The birth rates of the high-risk and low-risk groups were then displayed as events in a Kaplan-Meier test (Figure 26). According to this analysis, those classified as high risk generally delivered earlier than those classified as low risk (p=0.0004).

[0325] Post-validation analysis

[0326] Predictor performance was measured using a combination of blinded corroboration subjects and validation analyses within optimal BMI and GA intervals (Supplementary Data, below). The ROC curve for the combined sample set is shown, corresponding to an AUROC of 0.72 (p=0.013) (Figure 27).

[0327] Using omics approaches, we found that the BMI intervals >22 and ≤37 kg / m 2 developed a maternal serum predictor consisting of the ratio of IBP4 / SHBG levels at 19–20 weeks of pregnancy, which identified 75% of women destined for sPTB. A previous history of sPTB (Goldenberg et al., Epidemiology and causes of preterm birth., Lancet., 2008;3 71(9606):75-84, doi:10.1016 / S0140-6736(08)60074-4; Petrini et al., Estimated effect of 17 alpha-hydroxyprogesterone caproate on preterm birth in the United States., Obstet Gynecol., 2005;105(2):267-272) and cervical length measurements (Iams et al., The length of the cervix and the risk of spontaneous premature delivery., National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network., N Engl J Med., 1996;334(9):567-72; Hassan et al., Vaginal progesterone reduces the rate of preterm birth in women with a sonographic short cervix: a multicenter, randomized, double-blind, placebo-controlled trial. Ultrasound Obstet Gynecol. 2011;38(1):18-31) is currently considered the best indicator of clinical risk, but either individually or in combination fails to predict the majority of sPTB.

[0328] An ideal sPTB predictive tool would be minimally invasive, performed early in pregnancy to coincide with routine obstetric visits, and accurately identify those most at risk. This omics study suggests that disruptions in the physiological state of pregnancy can be detected by maternal serum analytes measured in sPTB subjects. Omics discovery studies of PTB include proteomics approaches (Gravett et al., Proteomic analysis of cervical-vaginal fluid: identification of novel biomarkers for detection of intra-amniotic infection). , J Proteome Res., 2007;Vol. 6(1):pp. 89-96; Goldenberg et al., The preterm prediction study: the value of new vs. standard risk factors in predicting early and all spontaneous preterm births. NICHD MFMU Network., Am J Public Health., 1998;Vol. 88(2):pp. 233-8; Gravett et al., Diagnosis of intra-amniotic infection by proteomic profiling and identification of novel biomarkers., JAMA., 2004;292(4):462-469; Pereira et al., Insights into the multifactorial nature of preterm birth: Proteomic profiling of the maternal serum glycoproteome and maternal serum peptidome among women in preterm labor., Am J Obstet Gynecol., 2010;202(6):555.e1-10;32. Pereira et al., Identification of novel protein biomarkers of preterm birth in human cervical-vaginal fluid., J Proteome Res., 2007;6(4):1269-76 Dasari et al., Comprehensive proteomic analysis of human cervical-vaginal fluid. J Proteome Res., 2007;6(4):1258-1268; Esplin et al., Proteomic identification of serum peptides predicting subsequent spontaneous preterm birth. Am J Obstet Gynecol., 2010;204(5): 391.e1-8), transcriptome techniques (Weiner et al., Human effector / initiator gene sets that regulate myometrial contractility during term and preterm labor. Am J Obstet Gynecol., 2010;202(5):474.e 1-20; Chim et al., Systematic identification of spontaneous preterm birth-associated RNA transcripts in maternal plasma., PLoS ONE., 2012; Vol. 7 (4):e34328, Enquobahrie et al., Early pregnancy peripheral blood gene expression and risk of preterm delivery: a nested case-control study., BMC Pregnancy Childbirth., 2009;9(1):56), genomic methods (Bezold et al., The genomics of preterm birth: from animal models to human studies., Genome Med., 2013;5(4):34; Romero et al., Identification of fetal and maternal single nucleotide polymorphisms in candidate genes that predispose to spontaneous preterm labor with intact membranes., Am J Obstet Gynecol., 2010; 202(5):431.e1~34 ; Swaggart et al., Genomics of preterm birth., Cold Spring Harb Perspect Med. , 2015; Vol. 5(2): a023127; Haataja et al., Mapping a new spontaneous preterm birth susceptibility gene, IGF1R, using linkage haplotype sharing, and association analysis.,PLoS Genet.,2011;7(2) :e1001293; McElroy et al., Maternal coding variants in complement receptor 1 and spontaneous idiopathic preterm birth., Hum Genet., 2013;132(8):935-42.), and metabolomics techniques (Menon et al., Amniotic fluid metabolomic analysis in spontaneous preterm birth., Reprod Sci. , 2014;21(6):791-803). However, to date, none of these approaches have resulted in a validated test to reliably predict the risk of sPTB in asymptomatic women.

[0329] This invention is the result of a large-scale prospective, concurrent clinical study that allowed for independent discovery, confirmation, and validation analyses while adhering to IOM guidelines for omics test development. It involved the construction of a large-scale standardized multiplexed proteomic assay to investigate biological pathways related to pregnancy. Furthermore, the study size and relatively wide blood collection window (17 The GA interval (0 / 7–28 6 / 7 weeks) allowed for the identification of GA intervals where there were significant changes in protein concentrations between sPTB cases and term controls. The use of a less complex predictor model (i.e., ratio of two proteins) limited the pitfalls of overfitting.

[0330] The application of proteomic assays and model building led to the identification of a key protein pair (IBP4 and SHBG) with consistently good predictive performance for sPTB. Despite the challenges of building a classifier for conditions with multiple etiologies, this predictor performed well with a cutoff of <37 0 / 7 vs. ≥37 0 / 7 weeks GA and an AUROC of 0.75. Importantly, the accuracy of the predictor improved with earlier sPTB (e.g., <35 0 / 7 weeks GA), enabling the detection of sPTB with the highest potential for morbidity. Subjects identified as at high risk for sPTB using the IBP4 / SHBG predictor delivered significantly earlier than subjects identified as at low risk. Our findings suggest that IBP4 and SHBG may play important functions related to the pathogenesis of sPTB and / or act as a convergence point for related biological pathways.

[0331] Universal transvaginal ultrasound (TVU) measurement of cervical length (CL) was not routinely performed at the majority of our study centers and was available in less than one-third of study subjects. It would be interesting for future studies to assess whether CL measurements improve on proteomic predictors or, alternatively, whether risk stratification with an IBP4 / SHBG classifier identifies women who would most benefit from serial CL measurements. Finally, it would be of great interest to investigate the performance of molecular predictors together with BMI variables or perhaps in combination with other medical / reproductive history and sociodemographic characteristics.

[0332] In conclusion, a routine predictive test for sPTB based on serum measurements of IBP4 and SHBG in asymptomatic parous and nulliparous women was validated in a fully independent set of subjects. Further functional studies of these proteins, their genetic regulation, and associated pathways can help elucidate the molecular and physiological basis of sPTB. Application of this predictor should enable early and sensitive detection of women at risk for sPTB. This could not only improve pregnancy outcomes through increased clinical surveillance but also accelerate the development of clinical interventions for PTB prevention.

[0333] Supplementary Materials and Methods

[0334] Discovery and Support

[0335] Discovery and validation subjects were subjects from the PAPR study described above in this example.

[0336] Discovery and Supporting Principles

[0337] sPTB was defined as described above in this example. Predictor discovery and validation was performed according to the guidelines of best practices in "omics" research. (IOM (Institute of Medicine), Evolution of Translation Omics: Lessons Learned and the Path Forward. (Micheel CM, Nass SJ, Omenn GS, eds.), Washington, DC: The National Academies Press.; 2012: pp. 1-355). Nested case / control analyses used completely independent sample sets. Cases and controls selected for discovery and corroboration underwent central review for within-subject data discrepancies. No source document verification (SDV) with medical records was performed. All discovery and corroboration sPTB cases and controls were individually adjudicated by the lead investigator, and discrepancies were resolved by the clinical site PI. A detailed analysis protocol was pre-established, including the study design, analysis plan, and corroboration blinding protocol. Laboratory and data analysis personnel were blinded to the case, control, and GA data allocation for corroboration subjects. A blinded internal statistician assigned predictor scores, calculated as described below, to all subject samples. Case, control, and GA data correlated with predictor scores by the DCO were provided to an independent external statistician for analysis. AUROC results were then transferred to the DCO. Data transfer used the SUMPRODUCT function in Excel (Microsoft Excel 2013) to ensure data integrity was maintained. Digital time stamps were applied to analysis data, plans, and reports to provide an audit trail from the data of interest to the supporting results.

[0338] Discovery and Confirmatory Research Design

[0339] Discovery and corroboration sPTB cases totaled 86 and 50 subjects, respectively, collected over the course of blood collection from 17 0 / 7 to 28 6 / 7 weeks GA (GABD). Subjects used for discovery and corroboration were completely independent of each other and those used for validation. Matched controls for discovery and corroboration sPTB cases were identified as described above in this example.

[0340] Prevalence analysis

[0341] After the discovery, corroboration, and validation analyses, additional full-term controls not used in previous studies were selected from the PAPR database and processed in the laboratory using the MRM-MS assay applied for validation as described above in this example. R statistical software (version 3.0.3) (Team RC. R: a Language and Environment for Statistical Computing., Vienna, Austria; 2014, 2015; Matei A, Tille Y., The R "sampling" Using the sampling package from the European Conference on Quality in Survey Statistics (ECS) (2006), a set of 187 subjects was randomly selected from validated GA blood collection intervals and compared with gestational age at birth (GAB) data from the 2012 National Vital Statistics Report (NVSR) by univariate statistical analysis (chi-square test). Martin et al.: Final Data for 2012. National Vital Statistics Reports., 2014;63(09):1-86. The set of controls that most closely approximated the 2012 NVSR parity distribution based on a p-value (close to 1.0, with a minimum acceptable value of 0.950) was then selected and compared, as a whole, to the BMI distribution from the PAPR study. Univariate statistical analysis (chi-square test) on BMI data from the PAPR research database was used to select a control set that most closely approximated the distribution of BMI (close to 1.0, minimum acceptable value 0.950) and delivery timing in the NVSR, and compared it to the validated blood collection sample GABD. The set that most closely approximated all three distributions was selected as the subject set for the prevalence study. Corroboration, validation, and prevalence predictor score values ​​within the validation GABD interval and BMI limits totaled 150 subjects. This combined data set was used to obtain the best estimate of the confidence interval for the PPV curve in Figure 25. Confidence intervals for PPV were calculated using a normal approximation of the error for a binomial proportion. Brown et al., Interval estimation for a binomial proportion, Statistical science, 2001; vol. 16 (issue 2): pp. 101-133.

[0342] Laboratory Methods

[0343] A systems biology approach was used to generate a highly multiplexed multiple reaction monitoring (MRM) mass spectrometry (MS) assay through literature curation, targeted and untargeted proteomic discovery, and iterative application of small-volume MRM-MS analysis of target samples. A mature MRM-MS assay measuring 147 proteins was applied to discovery and validation studies. For all analyses, serum samples were processed in the laboratory as described above in this example. An aliquot of the pooled serum control (pHGS) was used to calculate the batch-to-batch analytical coefficient of variation (CV) for IBP4 and SHBG.

[0344] Basic predictor development strategies

[0345] A strategy was developed to avoid overfitting and to overcome the dilution of biomarker performance that is expected over a wide gestational age range due to the dynamic nature of protein expression during pregnancy. The ratio of upregulated to downregulated analyte intensities was used for predictor development. Such "inversion" is similar to top-scoring pair and two-gene classifier strategies. (Geman et al., Classifying gene expression profiles from pair wise mRNA comparisons, Stat Appl Genet Mol Biol, 2004;3(1 ):Article19;Price et al., Highly accurate two-gene classifier for differentiating gastrointestinal stromal tumors and leiomyosarcomas, Proc Natl Acad Sci USA 2007;104(9):3414-9. This approach allowed for amplification and self-normalization of diagnostic signals because both proteins in the "inversion" underwent the same pre-analytical and analytical processing steps. Inversion as a strategy for normalizing peptide intensity measurements in complex proteomics workflows is also similar to a recently introduced technique called "endogenous protein normalization (EPN)." (Li et al., An integrated quantification method to increase the precision, robustness, and resolution of protein measurement in human plasma samples, Clin Proteomics, 2015;12(1):3; Li et al., A blood-based proteomic classifier for the molecular characterization of pulmonary nodules, Sci Transl Med, 2013;5(207):207ra142. The number of candidate analytes used for model construction was reduced by analytical criteria. Analytical filters included cutoffs for analytical precision, strength, evidence of interference, sample processing order dependency, and preanalytical stability. The total number of analytes for any one predictor was limited to a single inversion, thereby avoiding complex mathematical models. The predictor score was defined as the natural logarithm of the single inversion value, and the inversion itself was the response ratio (defined above in this example). Finally, predictive performance was examined across narrowly overlapping 3-week gestational intervals.

[0346] Receiver operating characteristic curve

[0347] AUROC values ​​and associated p-values ​​were calculated to determine reversal, as described above in this example. The distribution and mean values ​​of predictor AUROCs in the discovery and validation combination sets were calculated using bootstrap sampling, performed iteratively by selecting a random set of samples with replacement. Efron B, Tibshirani RJ., An Introduction to the Bootstrap, Boca Raton, Florida: Chapman and Hall / CRC Press; 1994. The total number of selected samples in each iteration corresponded to the total number available in the starting pool.

[0348] Supplementary results

[0349] The discovery, corroboration, and validation subject characteristics are summarized in Table 3. The percentage of subjects with one or more prior sPTB histories was higher among discovery sPTB cases than among corroboration or validation. Other characteristics were generally consistent across studies.

[0350] Discovery and supporting analysis

[0351] Forty-four proteins were up- or down-regulated in the overlapping 3-week GA intervals and passed the analytical filter (Figure 28). Inversions were formed from the ratio of up-regulated to down-regulated proteins, and their predictive performance was tested in samples from each overlapping 3-week GA interval. The performance of a subset of inversions showing representative patterns is shown in Figure 29. Fluctuations in performance were evident: IBP4 / SHBG and APOH / SHBG inversions had better AUROC values ​​in the early window, while ITIH4 / BGH3 and PSG2 / BGH3 peaked later in pregnancy (Figure 24). Some inversions showed consistent but moderate performance across the entire gestational age range (PSG2 / PRG2) (Figure 29). Overall, the top-performing inversion, IBP4 / SHBG, had an AUROC = 0.74 in the interval from 19 0 / 7 to 21 6 / 7 (Figure 29). Subjects were excluded if they had a pre-pregnancy BMI < 35 (kg / m 2When stratified by GA, the AUROC performance of the IBP4 / SHBG predictor increased to 0.79 (Table 4). The IBP4 / SHBG predictor was selected for supportive analyses because of its consistently strong performance in early pregnancy (i.e., 17 0 / 7–22 6 / 7 weeks GA) (Figure 29) and potential desirable clinical utility.

[0352] The blinded IBP4 / SHBG AUROC performance in the corroboration samples was 0.77 and 0.79 for all subjects and BMI-stratified subjects, respectively. This was in good agreement with the performance obtained in discovery (Table 5). After blind corroboration, the discovery and corroboration samples were combined to determine bootstrap performance. A mean AUROC of 0.76 was obtained from 2,000 bootstrap replicates (Figure 30).

[0353] BMI Validation Analysis

[0354] The performance of the IBP4 / SHBG predictor was evaluated using several BMI cutoffs in the validation sample (Table 5). Performance, as measured by AUROC, was very high (e.g., >37 kg / m 2 ) or low (e.g., ≦22 kg / m 2 ) was slightly improved by excluding BMI. Stratification by the combination of these two cutoffs resulted in an AUROC of 0.75 (Table 5).

[0355] Example 3 Correlation of mass spectrometry and immunoassay data This example shows (1) the results of Myriad RBM screening to identify IBP4 and other individual biomarkers of sPTB in early, mid, and late trimester collection windows, (2) the correlation of MS and immunoassay results for SHBG / IBP4, and (3) clinical data on SHBG as a biomarker for sPTB.

[0356] RBM data

[0357] Briefly, 40 cases and 40 controls with PAPR were assayed for RBM (20 / 20 from the early window, 10 / 10 from the mid-term window, and 10 / 10 from the late window). For RBM, Human Discovery MAP 250+ v2.0 (Myriad RBM, Austin, TX) was used. The goal of these analyses was to develop a multivariate model for predicting PTB using multiple analytes. We used four modeling methods: random forest (rf), boosting, lasso, and logistic (logit). We performed a first round of variable selection, in which each method independently selected its 15 best variables. From these 15, the best analytes were independently selected by each of the four modeling methods using backward variable selection and estimation of the area under the receiver operating characteristic curve (AUC) using out-of-bag bootstrap samples. Table 6 shows the top hits from several multivariate models. Table 7 shows the ranking of analytes in the early window (GABD weeks 17-22) by various multivariate models. Table 8 shows the ranking of analytes in the mid-term window (GABD weeks 23-25) by various multivariate models. Table 9 shows the ranking of analytes in the late window (GABD weeks 26-28) by various multivariate models.

[0358] Identifying commercially available ELISA kits that correlate with mass spectrometry data

[0359] Briefly, the ELISA vs. MS comparison included multiple studies using PAPR samples, ranging in size from 30 to 40 subjects. Each ELISA was performed according to the manufacturer's protocol. The predicted concentrations of each analyte by ELISA were then compared to the relative ratios from the same samples derived by MS. Pearson's r (Person's) correlation values ​​were then generated for comparison. The ELISA vs. MS comparison included multiple studies using PAPR samples, ranging in size from 30 to 40 subjects. Each ELISA was performed according to the manufacturer's protocol. The predicted concentrations of each analyte by ELISA were then compared to the relative ratios from the same samples derived by MS. Pearson's r (Person's) correlation values ​​were then generated for comparison. Table 10 provides epitope and clonality information for kits tested for the analytes IBP4_HUMAN and SHBG_HUMAN. Table 11 shows that not all ELISA kits correlate with MS, even for proteins for which correlation exists. See, for example, IBP4, CHL1, ANGT, and PAPP1.

[0360] One hundred twenty previously frozen serum samples from PAPR studies with known outcomes were selected for comparison between the ELISA and MS assays. These samples had a gestational age at blood collection (GABD) of 119 to 180 days. Samples were not excluded based on maternal BMI. ELISA was performed with commercially available kits for IBP4 (AL-126, ANSCH Labs Webster, TX) and SHBG (DSHBG0B, R&D Systems, Minneapolis, MN). Assays were performed according to the manufacturer's protocol. An internal standard was used for interplate normalization. Scores were calculated as ELISA concentration values ​​by LN ([IBP4] / [SHBG]) and LN (IBP4 RR / SHBG RR ) MS analysis. RR refers to the relative ratio of endogenous peptides to SIS peptide peak areas. The scores derived from the two methods were compared for case vs. control separations (p-values ​​were derived by unpaired t-test assuming equal standard deviations) (Figure 31).

[0361] Fifty-seven previously frozen serum samples (19 sPTB cases, 38 term controls) from the PAPR study with known outcomes were selected for comparison between the ELISA and MS assays. These samples had a gestational age at blood collection (GABD) of 133 to 148 days. ELISA was performed with commercially available kits for IBP4 (AL-126, ANSCH Labs Webster, TX) and SHBG (DSHBG0B, R&D Systems, Minneapolis, MN). The assays were performed according to the manufacturer's protocol. An internal standard was used to normalize samples run on different plates. Scores were calculated based on the ELISA concentration values ​​by LN ([IBP4] / [SHBG]) and LN (IBP4 RR / SHBG RR ) was calculated from MS. RR, where RR refers to the relative ratio of endogenous peptides to SIS peptide peak areas. The performance of the immunoassay was then determined by the area under the receiver operating characteristic curve (AUC) and compared with the MS-derived AUC for the same sample set (Figure 32). AUC values ​​were also determined after applying BMI stratification to samples (BMI > 22 < 37) to select a total of 34 samples (13 sPTB cases, 21 term controls) (Figure 33).

[0362] Sixty previously frozen serum samples from PAPR studies with known outcomes were analyzed by ELISA and MS assays. These samples have an expected gestational age at blood collection (GABD) of 133 to 146 days. Correlation analyses were performed on all BMI samples (Figure 34, right panel) or on a subset of samples with a BMI of >22 or ≤37 (Figure 34, left panel). ELISAs were performed with commercially available kits for IBP4 (AL-126, ANSCH Labs Webster, TX) and SHBG (DSHBG0B, R&D Systems, Minneapolis, MN). Assays were performed according to the manufacturer's protocol. An internal standard was used for interplate normalization. Scores were calculated based on ELISA concentration values ​​by LN ([IBP4] / [SHBG]) and LN (IBP4 RR / SHBG RR) MS analysis. RR refers to the relative ratio of endogenous peptides to SIS peptide peak areas. The scores derived from the two methods were compared by correlation and for case vs. control isolations (p-values ​​were derived by unpaired t-test assuming equal standard deviations). Table 12 shows the IBP4 and SHBG ELISA kits showing sPTB vs. control isolations (univariate).

[0363] Comparison of SHBG measurements by mass spectrometry and clinical analyzers

[0364] Thirty-five samples from individual subjects and serum pools from pregnant and non-pregnant women were analyzed simultaneously by Sera Prognostics and two independent reference laboratories, ARUP Laboratories and Intermountain Laboratory Services. Aliquots were shipped refrigerated to each laboratory, and shipments were coordinated so that testing would begin at all three laboratories on the same day. ARUP used a Roche cobas e602 analyzer, while Intermountain used an Abbott Architect CMIA. Both were semi-automated immunoassay instruments. Sera Prognostics used a proprietary proteomic analysis method that included immunodepletion and enzymatic digestion of samples and analysis on an Agilent 6490 mass spectrometer. Results from both ARUP and IHC were reported in nmol / L, while Sera uses the relative ratio (RR) of heavy and light peptide surrogates. Data from ARUP and Intermountain were compared to each other to determine accuracy (Figure 39). Linearity and precision were in good agreement across a wide range of results. The linear slope was 1.032 and r 2 The value was 0.990. The data from each reference laboratory were then compared to the Sera RR and linear regression plots (Figures 37 and 38). The data were in good agreement with the Sera results. ARUP is the r 2 The value is 0.937, which is the same as Intermountain 2 The value was 0.934.

[0365] Example 4 SNPs, insertions and deletions, and structural variants within PreTRM IBP4 and SHBG peptides This example shows known SNPs, insertions and deletions (indels), and structural variants within the PreTRM IBP4 and SHBG peptides.

[0366] Tables 13 and 14 detail known SNPs, insertions and deletions (indels), and structural variants within the PreTRM IBP4 and SHBG peptides. This information is from the Single Nucleotide Polymorphism Database (dbSNP) build 146. The single missense mutation (G>C) A179P in SHBG (dbSNP id: rs115336700) has the highest overall allele frequency of 0.0048. Although this allele frequency is low, several subpopulations studied in the 1000 Genomes Project showed significantly higher frequencies. These populations (allele frequencies) were: Americans of African ancestry from the southwestern United States (0.0492); Afro-Carribbean from Barbados (0.0313); Yoruba from Ibadan, Nigeria (0.0278); Ruiya from Webuye, Kenya (0.0101); Esan, Nigeria (0.0101); Colombians from Medellin, Colombia (0.0053); and Gambians from western Gambia (0.0044). All other subpopulations studied did not have the mutation at this nucleotide position. Table headings include cluster id - (dbSNP rs number), heterozygosity - average heterozygosity, validation - validation method (or blank without validation), MAF - minor allele frequency, function - functional feature of the polymorphism, dbSNP allele - allelic nucleotide identity, protein residue - residue resulting from the allele, codon position - codon position, NP_001031.2 amino acid position - amino acid position in reference sequence NP_001031.2, and NM_001040.2 mRNA position - nucleotide position in reference sequence NM_001040.2.

[0367] Example 5 IBP4 / SHBG reversal amplifies the diagnostic signal and reduces analytical variability in sPTB This example demonstrates the amplification and reduced variability of diagnostic signals obtained by using the IBP4 / SHBG reversal strategy.

[0368] The levels of IBP4 and SHBG determined separately by MS over the indicated gestational age ranges for sPTB cases and full-term controls are shown (Figures 44 and 45). Curves were generated by mean smoothing the peptide relative ratios (endogenous peptide peak area to the corresponding SIS peak area). Case vs. control signals correspond to approximately a 10% maximum difference in IBP4 and SHBG. When the scores calculated as ln(IBP4RR / SHBGRR) are plotted, signal amplification is evident (approximately a 20% maximum difference) (Figure 46). These data demonstrate the diagnostic signal amplification obtained using the IBP4 / SHBG inversion strategy.

[0369] Because each protein undergoes the same analytical and pre-analytical processing steps, forming a ratio of the levels of the two proteins can reduce variability. To examine the impact on variability, the CVs of the individual proteins (IBP4 and SHBG RR) and the CV of the IBP4 RR / SHBG RR ratio were determined in pooled control serum samples (pHGS) from pregnant donors. Pooled control samples with no biological variability were analyzed in multiple batches over several days. The inverse variability is less than the variability associated with the individual proteins. (Figure 48)

[0370] To investigate whether the formation of inversions commonly amplifies diagnostic signals, we examined the ROC performance (AUC) of high-performing inversions (AUC>0.6) formed by ratios of multiple proteins. The range of AUC values ​​(sPTB cases vs. term controls) using datasets derived from samples collected at 19 / 0 to 21 / 6 weeks of gestation is shown in the top panel of Figure 47. The adjacent boxplots show the range of ROC performance for the individual up- and down-regulated proteins used to form the associated inversions. Similarly, p-values ​​derived from Wilcoxon tests (sPTB cases vs. term controls) for inversions are more significant than the corresponding individual proteins (Figure 47, bottom).

[0371] To investigate whether the formation of inversions reduces variability more generally, we examined the analytical variability (i.e., the ratio of relative peak areas to the analytical variability of individual proteins, including inversions, in pooled control serum samples (pHGS) from pregnant donors) of 72 different inversion values. Pooled control samples with no biological variability were analyzed in multiple batches over several days. The inversion variability is less than the variability associated with individual proteins (Figure 49).

[0372] Generalizability of reversal strategies to reduce analytical variability.

[0373] Figure 48 reports the calculated CVs for pHGS samples (pooled pregnancy samples) analyzed in the laboratory across several batches, days, and multiple instruments. The CVs were calculated using pHGS samples that lack biological variability and therefore represent a measure of analytical variability introduced in laboratory processing of the samples. The analytical variability associated with the ratioed values ​​of the 72 inversions is lower than the analytical variability of the relative peak areas of the individual up- and down-regulated proteins used to form the inversions (Figure 49).

[0374] Example 6 Medically Indicated PTB Analysis This example confirms that the classifier is sensitive to components of medically indicated PTB based on conditions such as pre-eclampsia or gestational diabetes.

[0375] The PreTRM™ was developed and validated as a predictor of spontaneous PTB. Approximately 75% of all PTB in the United States are spontaneous, with the remainder being medically indicated due to some maternal or fetal complication (e.g., preeclampsia, intrauterine growth restriction, infection). Forty-one medically indicated PTB samples from the PAPR Biobank were analyzed in the laboratory and a PreTRM score was calculated. The PreTRM™ score was compared between subjects annotated as medically indicated for preeclampsia and subjects annotated for other procedures. Subjects with preterm delivery medically indicated for preeclampsia had significantly higher scores than others (Figure 50).

[0376] Figure 52 shows an inversion intensity heatmap for diabetes annotation. Red arrows indicate diabetic subjects. Samples are listed at the bottom, with PTB cases on the right side of the screen and full-term births on the left. Diabetic patients are clustered on the right. This demonstrates that inversions that stratify gestational diabetes can be identified, and therefore, diagnostic tests can be constructed from biomarkers that predict gestational diabetes.

[0377] Example 7 Other transitions and peptides Table 16 shows a comparison of IBP4 peptide MS data and transition MS data. * +10 Daltons) exemplify various transitions and their relative intensities that can be monitored for quantification of IBP4. One skilled in the art could potentially select any of these peptides or transitions, or others not exemplified, for quantification of IBP4.

[0378] Table 17 shows a comparison of IBP4 peptide MS data and transition MS data. IBP4 tryptic peptides derived from recombinant protein were analyzed by MRM-MS to identify potential surrogate peptides and their transitions. One skilled in the art could potentially select any of these peptides or transitions, or others not exemplified, for quantification of IBP4. IBP4 was identified by RBM (above), and then synthetic peptides were ordered to construct the assay.

[0379] Table 18 shows a comparison of SHBG peptide MS data and transition MS data. SHBG tryptic peptides derived from recombinant protein and pooled pregnancy serum were analyzed by MRM-MS to identify potential surrogate peptides and their transitions. One skilled in the art could potentially select any of these peptides or transitions, or others not exemplified, for quantification of SHBG. Isoform-specific peptides identified in serum are also shown.

[0380] Table 19 shows proteins whose serum levels change across 17-25 weeks GA in PTB samples. * Additional proteins restricted to 19-21 weeks GA in PTB. LC-MS (MRM) assays of 148 proteins from multiple pathways were used to analyze serum samples from 312 women (104 sPTB cases, 208 term controls) at 17-25 weeks gestational age (GA). MRM peak area data were analyzed by hierarchical clustering, t-tests, and correlation with GA. After analytical filtering, 25 proteins showed significant differences (p<0.05) between sPTB and term subjects (Table 1). Levels of 14 proteins were higher and three were lower in sPTB samples across the entire GA range. Other proteins were found to be dynamically regulated at subintervals of GA. For example, at 19-21 weeks GA, seven additional proteins were elevated in sPTB and one was lower.

[0381] Table 20 lists the 44 proteins that met the analytical filter that were up- or down-regulated in sPTB versus full-term controls.

[0382] Example 8 Mechanistic insights into serum proteomic biomarkers predicting spontaneous preterm birth This example demonstrates that biomarker performance varies considerably across GA, as specific protein expression changes dynamically throughout pregnancy. Differentially expressed proteins have functions related to steroid metabolism, placental development, immune tolerance, angiogenesis, and pregnancy maintenance. Figure 55, Figures 57-59. These differences in protein profiles seen in sPTB reflect disturbed developmental transitions within the fetal / placental compartment during the second trimester.

[0383] Briefly, the aim of the study described in this example was to gain insight into the physiological basis of biomarker associations with spontaneous preterm birth (sPTB) prediction.

[0384] research design

[0385] Inflammatory, infectious, and hemorrhagic pathways have been implicated in the pathogenesis of preterm birth. However, little is known about which proteins are measurable in the blood and at what stage during pregnancy they are disrupted. To address these questions, we developed LC-MS (MRM) assays for 148 proteins from multiple pathways and analyzed serum samples from 312 women (104 sPTB cases and 208 full-term controls) at 17-25 weeks of gestational age (GA).

[0386] Briefly, serum samples were depleted of high-abundance proteins, digested with trypsin, and enriched with heavily labeled stable isotope standard (SIS) peptides. SIS peptides were used for normalization by dividing the peak area of ​​the peptide fragment ion (i.e., transition) measured in serum by the peak area of ​​the corresponding SIS transition to generate a response ratio. Response ratios of the MRM peak area data were analyzed by hierarchical clustering, t-tests, and correlation with GA.

[0387] As shown in Figure 53, multiple peptides correlate well to the same protein. Different branches (grouped by color) correspond to identifiable functional categories such as acute phase proteins, apolipoproteins, and known pregnancy-specific proteins. Protein complexes important to reproductive biology such as PAPP1:PRG2, INHBE:INHBC, and IGF2:IBP3:ALS are revealed. This quality assessment and relationship highlighting validates the highly multiplexed MRM-MS assay described in this application for use in investigating pregnancy biology and discovering analytes predictive of sPTB.

[0388] Figure 54 shows differentially expressed proteins that function in cell-extracellular matrix interactions. TENX activates latent TGF-b at the transition point of cytotrophoblast differentiation and is localized in fetal and maternal stroma. Alcaraz, L. et al., 2014, J. Cell Biol., 2 Vol. 05 (No. 3), pp. 409-428; Damsky, C. et al., 1992, J. Clin. Invest. 89(1), pp. 210-222. Reduced serum TENX levels in sPTB indicate placental vascular abnormalities or reduced TGF-β activity. NCAM1 (CD56) is highly expressed on neurons and natural killer cells. NCAM1 is also expressed by endovascular trophoblasts, but is reduced or absent in PE placentas. Red-Horse, K. et al., 2004, J. Clin. Invest., 114:744-754. Reversed serum NCAM1 levels in sPTB cases may reflect defective spiral artery remodeling and / or immune dysregulation. CHL1 is homologous to NCAM1 and directs integrin-mediated cell migration. BGH3 (TGF-1) is a cell adhesion molecule expressed on vascular endothelial cells and inhibits angiogenesis through specific interaction with αv / β3 integrin. Son, HN. et al., 2013, Biochimica et Biophysica Acta, 1833(10), 2378-2388. Elevated TGFBI in sPTB cases may indicate reduced placental vascularization.

[0389] Figure 55 shows a kinetic plot of differentially expressed proteins functioning in the IGF-2 pathway, showing maximum separation at 18 weeks. IGF2 stimulates proliferation, differentiation, and endometrial invasion by extravillous trophoblasts during early pregnancy. IBP4 binds to IGF2 at the maternal-fetal interface and regulates IGF2 bioavailability. Elevated IBP4 and reduced IGF2 during the first trimester correlate with IUGR and SGA, respectively. Qiu, Q. et al., 2012, J. Clin. Endocrino.l Metab., 97(8):E1429-39; Demetriou, C. et al., 2014, PLOS, 9(1):E85454. PAPP1 is a placenta-specific protease that cleaves IBP4, releasing active IGF2. Low serum PAPP1 levels during early pregnancy are associated with IUGR, PE, and PTB. Huynh, L. et al., 2014, Canadian Family Physician, Vol. 60 (No. 10), pp. 899-903. PRG2( PRG2:proMBP) is expressed in the placenta and covalently binds to and inactivates PAPP1. The PRG2:PAPP1 inactive complex circulates in maternal serum. Huynh, L. et al., 2014, Canadian Family Physician, 60(10), 899-903. Pathway Research Disruption of the node is consistent with impaired IGF2 activity in sPTB cases, which may result in abnormal placentation. Figure 56A provides a schematic representation of the dynamic regulation and bioavailability of the aforementioned proteins during sPTB.

[0390] Figure 56B shows a schematic of the intracellular signals preferentially activated by insulin binding to IR-B, and by insulin and IGF binding to either IR-A or IGF1R. Belfiore and Malaguarnera, Endocrine-Related Cancer (2011) 18:R125-R147. Insulin and IG Activation of IR-A and IGF1R by F leads to the dominance of growth and proliferation signals through the phosphorylation of IRS1 / 2 and Shc proteins. Shc activation leads to the recruitment of the Grb2 / Sos complex, which subsequently activates Ras / Raf / MEK1 and Erk1 / 2. This latter kinase translocates to the nucleus and induces the transcription of several genes involved in cell proliferation and survival. Phosphorylation of IRS1 / 2 induces activation of the PI3K / PDK1 / AKT pathway. In addition to its role in metabolic effects, AKT is linked to the activation of effectors involved in the control of apoptosis and survival (BAD, Mdm2, FKHR, NFkB, and JNK) and protein synthesis and cell proliferation (mTOR).

[0391] Figure 57 shows a kinetic plot of differentially expressed proteins with functions related to metabolic hormone balance. The placental protein sex hormone-binding globulin (SHBG) increases during pregnancy and determines the bioavailability and metabolism of sex steroid hormones. Decreased SHBG levels result in higher free androgen and estrogen levels. Free androgens can be converted to estrogen by placental aromatase activity. The anti-progestational activity of estrogen accelerates pregnancy / labor. Thyroxine-binding globulin (THBG) is induced by estrogen and increases approximately 2.5-fold by mid-pregnancy. Elevated serum THBG levels in sPTB cases may result in reduced free thyroid hormones. Hypothyroidism during pregnancy is associated with an increased risk of miscarriage and premature birth. Stagnaro-Green A. and Pearce E., 2012, Nat. Rev. Endocrinol. , Vol. 8(11): pp. 650-658. Angiotensinogen is increased approximately threefold by estrogen by mid-pregnancy, stimulating an approximately 40% increase in plasma volume. Upregulation of ANGT may be linked to pregnancy-induced hypertension, a condition associated with an increased risk of sPTB.

[0392] Figure 58 shows a kinetic plot of differentially expressed proteins with functions in angiogenesis. TIE1 is an inhibitory co-receptor for the TIE2 angiopoietin receptor and blocks the ability of Ang-2 to stimulate angiogenesis. Seegar, T. et al., 2010, Mol. Cell., Volume 37 (No. 5): pp. 643-655. Pigment epithelium-derived factor (PEDF) It is an anti-angiogenic factor expressed in the ovarian tissue and stimulates the cleavage and inactivation of VEGFR-1 by gamma-secretase. 10 Cathepsin D (CATD) cleaves prolactin to produce vasoinhibin, which inhibits angiogenesis. Elevated serum CATD and vasoinhibin levels are associated with preeclampsia. Nakajima, R. et al., 2015, Hypertension Research, 38, pp. 899-901. Leucine-rich alpha-2-glycoprotein (LRG1 / A2GL) promotes TGF-β signaling by binding to its co-receptor, endoglin. TGF-β activates endothelial cell mitogenesis and angiogenesis via the Smad1 / 5 / 8 signaling pathway. Wang, X. et al., 2013, Nature, vol. 499 ( (No. 7458). PSG3 induces anti-inflammatory cytokines from monocytes and macrophages and stimulates angiogenesis by binding to TGF-β. Low levels of PSG are associated with IUGR. Moore, T. and Dveksler, G., 2014, Int. J. Dev. Biol., 58:273-280. ENPP2 (autotaxin) is an extracellular enzyme with lysophospholipase D activity that produces lysophosphatidic acid (LPA). LPA acts on placental receptors to stimulate angiogenesis and chemotaxis of NK cells and monocytes. Autotaxin levels are reduced in cases of PIH and early-onset PE. Chen, SU, et al. , 2010, Endocrinology, vol. 151(1): pp. 369-379.

[0393] Figure 59 shows a kinetic plot of differentially expressed proteins with functions related to innate immunity. LBP is a bacterial LPS toll-like receptor-4 that induces inflammatory responses in the innate immune pathway through its co-receptor CD14. Fetuin-A (alpha-2-HS-glycoprotein) is a carrier protein for blood fatty acids, and the FetA-FA complex can bind to and activate the TLR4 receptor. Pal, D. et al., 2012, Nature Med., 18(8):1279-85.

[0394] Figure 60 shows kinetic plots of differentially expressed proteins with functions related to coagulation.

[0395] Figure 61 shows the kinetic plots of the differentially expressed serum / secreted proteins.

[0396] Figure 62 shows the kinetic plots of differentially expressed PSG / IBP.

[0397] Figure 63 shows the kinetic plots of differentially expressed ECM / cell surface proteins.

[0398] Figure 64 shows the kinetic plots of differentially expressed complement / acute phase protein-1.

[0399] Figure 65 shows the kinetic plots of differentially expressed complement / acute phase protein-2.

[0400] Figure 66 shows the kinetic plots of differentially expressed complement / acute phase protein-3.

[0401] Figure 67 shows the kinetic plots of differentially expressed complement / acute phase protein-4.

[0402] Example 9 SDT4 / SV4 kinetic analysis This example provides a kinetic analysis of all analytes first exemplified in Example 1 above, using data from 17 weeks, day 0 to 28 weeks, day 6.

[0403] In Figures 68-85, the mean relative ratio of each peptide transition was plotted against GABD using the R ggplot2 package with a mean smoothing function (window = + / - 10 days). Graphs feature separate plots of cases versus controls using two different gestational age at delivery cutoffs (<37 0 / 7 vs. >=37 0 / 7 weeks, and <35 0 / 7 vs. >=35 0 / 7 weeks). Plot titles include the protein abbreviation, underlined, and peptide sequence. Analyte sequences may be abbreviated to fit the title to the plot.

[0404] The kinetic analyses exemplified herein are useful for several purposes. They indicate whether and in what direction analyte levels change during pregnancy, whether the changes differ between cases and controls, and the diagnostic differences as a function of gestational age. In some cases, the diagnostic signal falls within a narrow gestational age range and increases or decreases with time. Additionally, the shape of the kinetic plot provides a visual guide for selecting pairs of proteins with good inversion.

[0405] Analytes found to show significant case-control segregation in the early window, e.g., with samples collected between 18 and 20 weeks of gestation, include, for example, AFAM, B2M, CATD, CAH1, C1QB, C1S, F13A, GELS, FETUA, HEMO, LBP, PEDF, PEPD, PLMN, PRG2, SHBG, TENX, THRB, and VCAM1. Analytes found to show significant case-control segregation in the late window, e.g., with samples collected between 26 and 28 weeks of gestation, include, for example, ITIH4, HEP2, IBP3, IGF2, KNG1, PSG11, PZP, VASN, and VTDB. Separation of cases versus controls was improved using cutoffs of less than 35 0 / 7 weeks versus 35 0 / 7 weeks or greater versus less than 37 0 / 7 weeks versus 37 0 / 7 weeks or greater, as seen for analytes including, for example, AFAM, APOH, CAH1, CATD, CD14, CLUS, CRIS3, F13B, IBP6, ITIH4, LYAM1, PGRP2, PRDX, PSG2, PTGDS, SHBG, and SPRL1. Numerous inflammatory and immunomodulatory molecules were found to show improved separation using lower gestational age at delivery cutoffs. Those skilled in the art will appreciate that any analyte showing significant separation between cases and controls for a given time window, as shown in the accompanying figures, is a candidate for use in the inversion pairs of the present invention, either as a single biomarker or as part of a biomarker panel of analytes.

[0406] Finally, analyte kinetic plots that lack case-control differences but show changes in analyte intensity throughout pregnancy are useful in pregnancy clocks according to the methods of the present invention. These analytes, also referred to herein as "clock proteins," can be used to determine gestational age without or in conjunction with other dating methods (e.g., date of last menstrual period, ultrasound dating). Table 60 provides a list of clock proteins useful in the pregnancy clocks of the present invention.

[0407] Example 10 Finding 2: Analysis of sPTB cases This example describes the analysis of all previously analyzed sPTB cases described in the examples above, their matched controls (two per case), and two new controls. The analysis described in this example extends the commercial blood collection window beyond 19 and 20 weeks, generates additional data on the prediction of sPTB <35 weeks based on multiple samples from all previous cases, leads to the discovery of new analytes and reversals, defines molecular clock proteins, elucidates risk thresholds, and forms precise validation claims for future clinical studies.

[0408] Sample processing method

[0409] A standardized protocol was developed to govern the conduct of the Proteomic Assessment of Preterm Birth Risk (PAPR) clinical study. This protocol specified that samples and clinical information could also be used to study other pregnancy complications. Specimens were obtained from women at 11 Institutional Review Board (IRB)-approved sites across the United States. After providing informed consent, serum and plasma samples were obtained, along with relevant information regarding patient demographics, past medical and reproductive history, current pregnancy history, and concomitant medications. Postdelivery, data were collected regarding maternal and infant status and complications. Serum and plasma samples were processed according to a protocol calling for standardized refrigerated centrifugation, sample aliquoting into 0.5 ml 2D-barcoded cryovials, and subsequent freezing at -80°C.

[0410] After delivery, preterm birth cases were individually reviewed to determine their status as either spontaneous preterm or medically indicated preterm. Only spontaneous preterm birth cases were used in this analysis. For preterm biomarker discovery, LC-MS data were generated for 413 samples (82 sPTB cases and 331 full-term controls) spanning gestational ages 17 0 / 7 to 21 6 / 7 weeks. Each preterm sample was matched to four full-term controls for gestational age at blood collection. Every gestational age between 17 0 / 7 and 21 6 / 7 weeks included at least one sPTB case (and matched full-term control) on every day except for one. Four full-term controls were selected based on blood collection from that day. One full-term control in this study failed laboratory analysis but was not reanalyzed.

[0411] Serum samples were subsequently depleted of highly abundant proteins using the Human 14 Multiple Affinity Removal System (MARS-14), which removes 14 of the most abundant proteins. Equal volumes of each clinical sample or duplicates of two quality control serum pools were diluted with column buffer and filtered to remove precipitates. The filtered samples were depleted using a MARS-14 column (4.6 x 100 mm, Agilent Technologies). The samples were cooled to 4°C in an autosampler, the depletion column was run at room temperature, and the collected fractions were kept at 4°C until further analysis. The unbound fraction was collected for further analysis.

[0412] Depleted serum samples were reduced with dithiothreitol, alkylated using iodoacetamide, and then digested with trypsin. After trypsin digestion, samples were fortified with a pool of stable isotope standards at concentrations that approximated those of the surrogate peptide analytes. SIS-fortified samples were mixed and divided into two equal volumes. Each aliquot was placed in a -80°C storage container until ready to continue the workflow. One frozen aliquot of each sample was removed from the -80°C storage container, thawed, and desalted on a C18 solid-phase extraction plate (Empore, 3M). Eluted peptides were dried by lyophilization. The lyophilized samples were reconstituted in a reconstitution solution containing only an internal standard (IS Recon) to monitor the quality of the LC-MS step.

[0413] The fully processed samples were analyzed using dynamic multiple reaction monitoring (dMRM). Peptides were separated on a 2.1 x 100 mm Poroshell EC-C18, 2.7 μ particle size column using an Agilent 1290 UPLC at a flow rate of 0.4 mL / min and eluted using an acetonitrile gradient into an Agilent 6490 triple quadrupole mass spectrometer with an electrospray source operated in positive ion mode. The dMRM assay measured 442 transitions representing 119 peptides and 77 proteins, serving both diagnostic and qualitative purposes. Chromatographic peaks were integrated using MassHunter quantitative analysis software (Agilent Technologies). The ratio of the chromatographic peak area of ​​the surrogate peptide analyte to the corresponding SIS chromatographic peak area was reported.

[0414] A summary of the proteins, peptides, and transitions of serum analytes, SIS transitions, and IS Recon standards measured by the dMRM method is shown in Table 21. MARS-14 depleted proteins identify the analytes targeted by the MARS-14 immunodepletion column and are measured for quality control purposes. Quantitative transitions are used for relative response ratios, and qualitative transitions serve for quality control purposes. Asterisks (* ) indicates a name change. CSH indicates that this peptide corresponds to both CSH1 and CSH2. HLAG is referred to here as HLACI because this peptide is conserved in several HLA type I isotypes. LYAM3 is referred to here as LYAM1 because this peptide sequence is present in each but is only derived from LYAM1 by trypsin cleavage. SOM2 is referred to here as SOM2. CSH as a peptide is specific for both SOM2 and CSH.

[0415] Selection of significant proteins and inversions

[0416] For each analyte, fold-change values ​​were calculated indicating whether the mean of the SPTB case samples was higher or lower than the mean of the TERM control samples using two SPTB definitions (37 / 37 and 35 / 35) with and without BMI restrictions in each of the two- and three-week overlapping windows. Tables 22 and 23 show protein / transition AUROCs for two-week gestational age windows overlapping by one week (e.g., 119-132 refers to gestational days 119-132, which is equivalent to 17 and 18 weeks of gestation). Performance within each two-week window is reported for two different case-to-control cutoffs (<37 0 / 7 vs. >=37 0 / 7, <35 0 / 7 vs. >=35 0 / 7) and with (rBMI) and without (aBMI) BMI stratification. Tables 24 and 25 show protein / transition AUROCs for 3-week gestational age windows (shown in days, e.g., "119-139" refers to gestational days 119-139, which equate to 17, 18, and 19 weeks of gestation) overlapping by 2 weeks. Performance within each 3-week window was evaluated using two different case vs. control cutoffs (<37 0 / 7 vs. >=37 0 / 7, <35 0 / 7 vs. >=35 0 / 7) and with (rBMI) and without (aBMI) BMI stratification.

[0417] Figures 86-95 show kinetic plots of various peptide transitions for cases versus controls using a gestational age at term cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks. Figures 96-105 show kinetic plots of various peptide transitions for cases versus controls using a gestational age at term cutoff of <35 0 / 7 vs. >=35 0 / 7 weeks. Briefly, the mean relative ratio of each peptide transition is plotted against GABD using the R ggplot2 package with a mean smoothing function (window = + / - 10 days). The graphs feature separate plots of cases versus controls using two different gestational age at term cutoffs (<37 0 / 7 vs. >=37 0 / 7 weeks, and <35 0 / 7 vs. >=35 0 / 7 weeks). Plot titles include the protein abbreviation, underline, and peptide sequence. Analyte sequences may be abbreviated to fit the title to the plot.

[0418] Based on the fold change value indicating whether the mean of the SPTB case samples was higher or lower than the mean of the TERM control samples, each analyte was marked as up- or down-regulated for each combination (i.e., overlapping 2- or 3-week windows, BMI limits, and SPTB definitions), and if the majority of combinations were marked as up-regulated, the analyte was referred to as an overall up-regulated analyte. Vice versa. This is shown in Table 26.

[0419] Based on these up- and down-regulated assignments (55 up-regulated and 30 down-regulated), inversions were generated by dividing each up-regulated analyte relative ratio value by the down-regulated analyte relative ratio value and taking the natural logarithm of the result. This resulted in 1650 inversions (55 x 30 = 1650). For each inversion, the area under the receiver operating characteristic curve (AUCROC) indicating SPTB and TERM separation was calculated, along with a p-value indicating whether the AUCROC value was significantly different from AUCROC = 0.5 (i.e., SPTB and TERM separation was not significant). For inversions with an AUCROC > 0.6 and a p-value < 0.05, the performance of each inversion under different conditions (e.g., with and without gestational window, BMI restriction, and two sPTB cutoffs) was tabulated. Tables 27-42 show inversion classification performance at 17 and 18 weeks of gestation. Tables 47-58 show reversal classification performance at 17, 18, and 19 weeks of gestation. Tables 43-46 show reversal classification performance at 17-21 weeks of gestation. Additional potentially significant reversals are shown in Table 59.

[0420] Improved performance of predictors formed from more than one inversion (17-21 weeks) was also demonstrated. Briefly, inversions that showed strong predictive performance either early (e.g., 17-19 weeks) or later (e.g., 19-21 weeks) in this gestational age range were combined, and the performance of predictors formed from the combination of multiple inversions (SumLog) across the entire blood draw range was evaluated. This is shown in Table 61. Predictor scores were derived from the sum of the logarithmic values ​​of the individual inversions (SumLog); however, one skilled in the art could select other models (e.g., logistic regression). It is also contemplated that this multiple inversion approach could be applied to combinations of inversions specific to preterm rupture of membranes (PPROM) versus preterm labor without PPROM (PTL), fetal gender, and gestational age. It is further contemplated that predictors could include indicator variables that select the subset of inversions used if blood draw period, fetal gender, or gestational age are known.

[0421] Figure 110 shows the relationship between predictor score (ln IBP4 / SHBG) and prevalence-adjusted relative risk (positive predictive value) for sPTB using a cutoff of <37 0 / 7 gestational weeks versus >=37 0 / 7 gestational weeks. Samples were collected between 19 1 / 7 weeks and 20 6 / 7 weeks for BMIs >22 and <=37. The relative risk increases with increasing predictor score from a background rate of 7.3% (average population risk of sPTB in singleton pregnancies) to approximately 50%. Screening positivity curves for all score thresholds are overlaid. Confidence intervals (gray shading) were calculated assuming binomial observations and approximating the error distribution using a normal distribution. Sample distribution by classifier score is shown by a bar graph according to the color coding in the figure legend.

[0422] Figure 111 shows the relationship between predictor score (ln IBP4 / SHBG) and prevalence-adjusted relative risk (positive predictive value) for sPTB using a cutoff of <35 0 / 7 gestational weeks versus >=35 0 / 7 gestational weeks. Samples were collected between 19 1 / 7 and 20 6 / 7 weeks. The relative risk increases with increasing predictor score from a background rate of 4.4% (average population risk of sPTB (<35) in singleton pregnancies) to approximately 50%. Screening positivity curves for all score thresholds are overlaid. Confidence intervals (gray shading) were calculated assuming binomial observations and approximating the error distribution using a normal distribution. Sample distribution by classifier score is shown by a bar graph according to the color coding in the figure legend.

[0423] Clinical Observations: sPTB, PPROM, and PTL

[0424] Reversal performance (GABD 17-21 weeks) was evaluated independently for two distinct phenotypes: sPTB, PPROM, and PTL. PPROM occurs more often early and is associated with infection or inflammation. PTL can occur later and is generally considered a less severe phenotype. There were more significant reversals, and performance was higher for PPROM, dominated by proteins known to be involved in inflammation and infection. Reversals were selected to construct independent assays for PPROM and PTL, or to maximize performance overall using a combination of more than one reversal in a single predictor. For the analyses shown in Tables 61-64, an AUC of >0.65 and p<0.05 were required for either PPROM or PTL.

[0425] Table 61 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation, without BMI stratification and using a case-to-control cutoff of <37 0 / 7 vs. ≥37 0 / 7 weeks, for PPROM and PTL, separately. Table 62 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation, with BMI stratification (>22, <=37), using a case-to-control cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks, for PPROM and PTL, separately. Table 63 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation, without BMI stratification and using a case-to-control cutoff of <35 0 / 7 vs. >=35 0 / 7 weeks, for PPROM and PTL, separately. Table 64 shows the inverted AUROCs from 17 0 / 7 to 21 6 / 7 weeks of gestation, stratified by BMI (>22, <=37) and using a case vs. control cutoff of <35 0 / 7 vs. >=35 0 / 7 weeks, for PPROM and PTL separately.

[0426] Additionally, the best-performing analytes for PTL and PPROM were determined for GABD 19-20 weeks. Several inversions were constructed from the most robustly performing. IBP4 was identified as performing well for both PTL and PPROM, enabling its general utility for sPTB. Table 76 lists the transition AUROCs for PTL from 19 1 / 7 to 20 6 / 7 weeks of gestation using a case-control cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks without BMI stratification. Table 77 lists the transition AUROCs for PPROM from 19 1 / 7 to 20 6 / 7 weeks of gestation using a case-control cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks without BMI stratification. Figure 108 illustrates inversions with good performance at 19-20 weeks for PTL. Figure 109 illustrates reversal in PPROM with good performance at 19-20 weeks.

[0427] Clinical observations: Primigravida and multiparous women

[0428] Inversion performance (17-21 weeks) was further evaluated independently for two different phenotypes of sPTB: primigravida and multiparous women. In Tables 65-68, the top-performing inversions (17-21 weeks) are presented separately for primigravida (first-time mothers) and multiparous subjects. First-time mothers are most in need of a test to predict the probability of PTB, as physicians do not have a pregnancy history to determine / estimate risk. These results allow tests to be performed independently for these two groups or to combine high-performing inversions in a single classifier to predict both risks. The analyses presented in Tables 65-68 required an AUC > 0.65 and p < 0.05 for either primigravida or multiparous women.

[0429] Table 65 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation without BMI stratification and using a case-control cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks, separately for primiparas and multiparas. Table 66 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation with BMI stratification (>22, <=37) and using a case-control cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks, separately for primiparas and multiparas. Table 67 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation without BMI stratification and using a case-control cutoff of <35 0 / 7 vs. >=35 0 / 7 weeks, separately for primiparas and multiparas. Table 68 shows the inverted AUROCs from 17 0 / 7 to 21 6 / 7 weeks of gestation, stratified by BMI (>22, <=37) and using a case-to-control cutoff of <35 0 / 7 vs. >=35 0 / 7 weeks, separately for primiparas and multiparas.

[0430] Clinical Observation: Fetal Sex

[0431] Inversion performance (17-21 weeks) was further evaluated independently for subjects carrying male fetuses versus subjects carrying female fetuses. Several inversions were found to have fetal sex-specific predictive performance. Figure 106 shows the fetal sex-specific differences in IBP4 and SHBG analytes and scores (IBP4 / SHBG). IBP4 was significantly higher in subjects carrying male fetuses. Inversion performance remained comparable for gestational ages 19-21 weeks without BMI stratification (Figure 106). Additionally, the PAPR found that male fetuses were at increased risk for sPTB, with a p-value of 0.0002 and an odds ratio of 1.6. Therefore, fetal sex could be incorporated into the predictors (e.g., inversion score + fetal sex). The analyses shown in Tables 69-72 required an AUC of >0.65 and p<0.05 for either male or female fetuses.

[0432] Table 69 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation without BMI stratification and using a case-control cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks, separately by fetal sex. Table 70 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation with BMI stratification (>22, <=37) and using a case-control cutoff of <37 0 / 7 vs. >=37 0 / 7 weeks, separately by fetal sex. Table 71 shows the inverted AUROCs for 17 0 / 7 to 21 6 / 7 weeks of gestation without BMI stratification and using a case-control cutoff of <35 0 / 7 vs. >=35 0 / 7 weeks, separately by fetal sex. Table 72 shows the inverted AUROCs from 17 0 / 7 to 21 6 / 7 weeks of gestation, separately by fetal sex, stratified by BMI (>22, <=37) and using a case-to-control cutoff of <35 0 / 7 vs. >=35 0 / 7 weeks.

[0433] Example 11 Correlation of mass spectrometry data with immunoassay data This example demonstrates performance of immunoassays using the MSD platform (eg, MSD data correlates with data from commercial ELISAs and MS data for IBP4 and SHBG).

[0434] material

[0435] The following antibodies were used: sex hormone-binding globulin (Biospacific catalog numbers 6002-100051 and 6001-100050; R&D Systems catalog numbers MAB2656 and AF2656), IGFBP-4 (Ansh catalog numbers AB-308-AI039 and AB-308-AI042). SHBG proteins from Origene (catalog number TP328307), Biospacific (catalog number J65200), NIBSC (code: 95 / 560), and R&D Systems (available only as part of the ELISA SHBG kit) were tested as calibrators. Recombinant human IGFBP-4 (Ansh, catalog number AG-308-AI050) was used as a calibrator.

[0436] Generation of individual U-PLEX-coupled antibody solutions

[0437] Each biotinylated antibody was diluted to 10 μg / mL in Diluent 100 to a final volume of ≥ 200 μL. The biotinylated antibody was then added to 300 μL of the corresponding U-PLEX linker (a different linker was used for each biotinylated antibody). The samples were vortexed and incubated for 30 minutes at room temperature. Stop solution (200 μL) was added to each tube. The tubes were vortexed and incubated for 30 minutes at room temperature.

[0438] Preparation of multi-coating solutions

[0439] Each U-PLEX coupled antibody (600 μL) solution was combined in a single tube and vortexed to mix. When combining fewer than 10 antibodies, the solution volume was increased to 6 mL with stop solution to a final 1× concentration. Note that in these experiments, only a single antibody was present per well.

[0440] Coating of U-PLEX plates.

[0441] Multicoating solution (50 μL) was added to each well. The plate was sealed with an adhesive plate seal and incubated at room temperature for 1 hour or at 2-8°C overnight with shaking at approximately 700 rpm. After washing three times with at least 150 μL of 1x MSD wash buffer, the plate was ready for use.

[0442] Sample analysis

[0443] A 50 μl aliquot of sample or calibrator was added to each well. The plate was sealed and incubated at room temperature with shaking at approximately 700 rpm for 1 hour. The plate was then washed with at least 150 μL of 1× MSD wash buffer. *The plates were washed three times with 150 μL of 1× MSD wash buffer. 50 μL of detection antibody solution was added to each well. After sealing, the plates were incubated at room temperature for 1 hour with shaking at approximately 700 rpm. The plates were washed three times with at least 150 μL of 1× MSD wash buffer. 150 μL of 2× measurement buffer was added to each well, and the plates were immediately measured on the MSD instrument.

[0444] SHBG Antibody and Calibrator Screening

[0445] All antibodies were tested in all pairwise combinations in both capture detector orientations. Capture antibodies were prepared at 10 μg / mL, coupled to the U-PLEX linker, and coated onto the U-PLEX plate. SHBG R&D Systems calibrators were diluted in Diluent 43 to generate a 7-point standard curve along with an assay diluent blank. Samples were diluted in Diluent 43 and tested in the assay as follows: serum SHBG "high" and "low" samples: 100- and 500-fold dilutions, and serum pregnancy pool: 100-, 200-, 400-, and 800-fold dilutions. Detector antibodies were tested at 1 μg / mL in Diluent 3. Binding to the standard curve and native analyte in serum was assessed. The top analyte pairs were then tested with NIBSC and Biospacific calibrators diluted as described above.

[0446] IGFBP-4 antibody and calibrator screening.

[0447] Two antibodies were tested in both capture detector orientations. Capture antibodies were prepared at 10 μg / mL, coupled to the U-PLEX linker, and coated onto the U-PLEX plate. IGFBP-4 calibrators were diluted in diluent 12 to generate a seven-point standard curve along with an assay diluent blank. Samples were diluted in diluent 12 and tested in the assay as follows: serum IGFBP-4 "high" and "low" samples: 5x; serum pregnancy pool: 2x dilutions from 2x to 64x; and two individual human serum samples (MSD samples): 2x, 4x, 8x, and 16x. The detection antibody was tested at 1 μg / mL in diluent 12. Binding to the standard curve and native analyte in serum was assessed.

[0448] SHBG and IGFBP-4 testing using 60 serum samples.

[0449] Twelve antibody pairs were selected to measure SHBG in duplicate in 60 plasma samples derived from serum. For IGFBP-4, pair 2 was selected. For SHBG and IGFBP-4, plasma samples were diluted 1:1000 and 1:20, respectively. MSD ELISA results were compared with commercial ELISA kits and MS-MRM data.

[0450] result:

[0451] SHBG antibody screening

[0452] Only antibody pair 1 (R&D Mono Capture, Poly Detect) produced a strong signal with the Origene calibrator. This suggested that this calibrator could represent a subpopulation of endogenous SHBG analytes. Therefore, additional calibrators were tested in subsequent studies to identify calibrators that would work across all pairs. However, all antibody pairs recognized the native analyte in the high serum sample, low serum sample, and pregnancy pooled sample. R&D Poly AF2656 and Biospacific Mono 6001-100050 showed similar performance. Pairs 2, 3, and 12 showed nearly linear titration with sample dilution (Table 73). The top four antibody pairs were then tested for performance with three additional calibrators. Good calibrator curves were achieved for the top four pairs across the three calibrators (Table 74). The differences in signal may be due, in part, to differences in assigned concentrations.

[0453] The bottom panel shows that the NIBSC or Biospacific signal for the R&D calibrators varied depending on the antibody pair. Pairs 3 and 10 (same antibody with reversed capture detection orientation) had similar profiles. Pair 2 showed lower signals with NIBSC and Biospacific (same capture compared to pair 3). Pair 12 showed higher signals with Biospacific and over three-fold higher signals with the NIBSC standard.

[0454] IGFBP-4 antibody screening

[0455] The antibody pair 2 standard curve showed 4-6 fold higher specific calibrator signal and background compared to pair 1 (Table 75). Serum sample signals were within the linear range at most dilutions tested. The pregnancy pool approached background at 32- and 64-fold dilutions. Pair 2 showed approximately 12-fold higher signal in the samples, resulting in a 2- to 4-fold difference in quantity. Signal CVs were generally <5% for both pairs.

[0456] Measurement of SHBG and IGFBP-4 in 60 serum samples

[0457] For SHBG, the dilution was 1:1000, and the samples fell between standard calibrators 1 and 3. The median measured concentration was 58.4 μg / mL. The CV for replicates was low, with a median CV of 2.4%. The median measured concentration for IGFBP-4 was 234 ng / mL. The median CV for replicates was 2.2%. As shown in Figure 107, the MSD assay showed good correlation with both proteins compared to the commercial ELISA kit and the MS-MRM assay.

[0458] From the foregoing, it will be apparent that variations and modifications can be made to the invention described herein to adapt it to various usages and conditions, and such embodiments are within the scope of the following claims.

[0459] The recitation of a list of elements in any definition of a variable herein includes definitions of that variable as any single element or combination (or subcombination) of the listed elements. The recitation of an embodiment herein includes that embodiment as any single embodiment or in combination with any other embodiment or portion thereof.

[0460] All patents and publications mentioned in this specification are herein incorporated by reference to the same extent as if each individual patent or publication was specifically and individually indicated to be incorporated by reference. [Table 1-1] [Table 1-2] [Table 1-3] [Table 1-4] [Table 2] [Table 3-1] [Table 3-2] [Table 3-3] [Table 4] [Table 5] [Table 6-1] [Table 6-2] Table 6-3 Table 7-1 Table 7-2 Table 8-1 Table 8-2 Table 9-1 Table 9-2 Table 10 Table 11 Table 12 Table 13 Table 14-1 Table 14-2 Table 15-1 Table 15-2 Table 16-1 Table 16-2 Table 16-3 Table 16-4 Table 17-1 Table 17-2 Table 17-3 Table 17-4 Table 17-5 Table 18-1 Table 18-2 Table 18-3 Table 18-4 Table 18-5 Table 18-6 Table 18-7 Table 18-8 Table 18-9 Table 18-10 Table 18-11 Table 18-12 Table 18-13 Table 18-14 Table 18-15 Table 18-16 Table 18-17 Table 18-18 Table 18-19 Table 18-20 Table 18-21 Table 18-22 Table 18-23 Table 19 Table 20-1 Table 20-2 Table 21-1 Table 21-2 Table 21-3 Table 21-4 Table 21-5 Table 21-6 Table 21-7 Table 21-8 Table 22-1 Table 22-2 Table 22-3 Table 23-1 Table 23-2 Table 23-3 Table 24-1 Table 24-2 Table 24-3 Table 25-1 Table 25-2 Table 25-3 Table 26-1 Table 26-2 Table 27-1 Table 27-2 Table 27-3 Table 27-4 Table 27-5 Table 27-6 Table 27-7 Table 28-1 Table 28-2 Table 28-3 Table 29-1 Table 29-2 Table 30-1 Table 30-2 Table 31-1 Table 31-2 Table 31-3 Table 31-4 Table 31-5 Table 31-6 Table 32-1 Table 32-2 Table 32-3 Table 32-4 Table 32-5 Table 32-6 Table 32-7 Table 32-8 Table 32-9 Table 33-1 Table 33-2 Table 33-3 Table 33-4 Table 34-1 Table 34-2 Table 34-3 Table 35-1 Table 35-2 Table 35-3 Table 35-4 Table 35-5 Table 35-6 Table 35-7 Table 36-1 Table 36-2 Table 36-3 Table 36-4 Table 36-5 Table 37-1 Table 37-2 Table 37-3 Table 37-4 Table 37-5 Table 37-6 Table 37-7 Table 37-8 Table 38-1 Table 38-2 Table 38-3 Table 38-4 Table 38-5 Table 39-1 Table 39-2 Table 39-3 Table 39-4 Table 39-5 Table 39-6 Table 39-7 Table 39-8 Table 39-9 Table 39-10 Table 39-11 Table 40-1 Table 40-2 Table 40-3 Table 40-4 Table 40-5 Table 40-6 Table 40-7 Table 40-8 Table 40-9 Table 40-10 Table 41-1 Table 41-2 Table 41-3 Table 41-4 Table 41-5 Table 42-1 Table 42-2 Table 42-3 Table 42-4 Table 42-5 Table 42-6 Table 43-1 Table 43-2 Table 43-3 Table 43-4 Table 43-5 Table 43-6 Table 44-1 Table 44-2 Table 44-3 Table 44-4 Table 44-5 Table 44-6 Table 44-7 Table 45-1 Table 45-2 Table 45-3 Table 45-4 Table 45-5 Table 45-6 Table 46-1 Table 46-2 Table 46-3 Table 46-4 Table 46-5 Table 47-1 Table 47-2 Table 47-3 Table 47-4 Table 48-1 Table 48-2 Table 48-3 Table 48-4 Table 48-5 Table 49-1 Table 49-2 Table 49-3 Table 49-4 Table 50-1 Table 50-2 Table 50-3 Table 51-1 Table 51-2 Table 51-3 Table 51-4 Table 51-5 Table 51-6 Table 51-7 Table 51-8 Table 52-1 Table 52-2 Table 52-3 Table 52-4 Table 52-5 Table 52-6 Table 52-7 Table 52-8 Table 52-9 Table 52-10 Table 52-11 Table 52-12 Table 52-13 Table 53-1 Table 53-2 Table 53-3 Table 53-4 Table 53-5 Table 53-6 Table 53-7 Table 54-1 Table 54-2 Table 54-3 Table 54-4 Table 55-1 Table 55-2 Table 55-3 Table 55-4 Table 55-5 Table 55-6 Table 55-7 Table 55-8 Table 55-9 Table 56-1 Table 56-2 Table 56-3 Table 56-4 Table 56-5 Table 56-6 Table 56-7 Table 56-8 Table 56-9 Table 56-10 Table 56-11 Table 57-1 Table 57-2 Table 57-3 Table 57-4 Table 57-5 Table 57-6 Table 57-7 Table 58-1 Table 58-2 Table 58-3 Table 58-4 Table 58-5 Table 58-6 Table 59-1 Table 59-2 Table 59-3 Table 59-4 Table 60 Table 61-1 Table 61-2 Table 61-3 Table 61-4 Table 61-5 Table 62-1 Table 62-2 Table 62-3 Table 62-4 Table 63-1 Table 63-2 Table 63-3 Table 63-4 Table 63-5 Table 63-6 Table 64-1 Table 64-2 Table 64-3 Table 64-4 Table 65-1 Table 65-2 Table 65-3 Table 66-1 Table 66-2 Table 66-3 Table 66-4 Table 67-1 Table 67-2 Table 67-3 Table 67-4 Table 67-5 Table 68-1 Table 68-2 Table 68-3 Table 68-4 Table 69-1 Table 69-2 Table 69-3 Table 70-1 Table 70-2 Table 70-3 Table 71-1 Table 71-2 Table 71-3 Table 71-4 Table 72-1 Table 72-2 Table 72-3 Table 73 Table 74 Table 75 Table 76 Table 77-1 Table 77-2

Claims

1. 1. A composition comprising a progesterone or an antenatal corticosteroid for use in a method for providing preventative treatment of preterm birth in a pregnant human patient, said method comprising: (i) measuring a panel of isolated biomarkers comprising IBP4 and SHBG in a biological sample obtained from said patient, wherein measuring comprises subjecting said biological sample to a proteomics workflow consisting of mass spectrometry (MS); (ii) calculating a risk score using the inverted value of IBP4 / SHBG; wherein said risk score being greater than a reference risk score is an indication for administering said composition to said patient.

2. The composition described in claim 1, wherein the administration of the composition further comprises subjecting the patient to a treatment regimen including cervical cerclage, serial cervical length measurements, or a cervical pessary.

3. 3. The composition of claim 2, wherein the treatment regimen further comprises an enhanced monitoring and clinical management regimen including one or more of: (a) more frequent prenatal visits, (b) increased education regarding the signs and symptoms of early preterm labor, or (c) changes in treatment for diabetes and / or hypertension, compared to a pregnant human patient not at risk of preterm birth.

4. 10. The composition of claim 1, wherein the administration of the composition comprising progesterone comprises a progestogen, 17-alpha hydroxyprogesterone caproate, or vaginal progesterone.

5. 5. The composition of claim 4, wherein the administration of the composition comprising 17-alpha hydroxyprogesterone caproate is by injection and the administration of the composition comprising vaginal progesterone is in gel form.

6. 10. The composition of claim 1, wherein the method further comprises an initial step of detecting one or more measurable risk indicator features selected from the group consisting of age, previous pregnancies, history of previous low birth weight or preterm birth, multiple second trimester spontaneous abortions, previous first trimester induced abortions, familial and intergenerational factors, history of infertility, nulliparity, placental abnormalities, cervical and uterine abnormalities, pregnancy bleeding, intrauterine growth restriction, intrauterine diethylstilbestrol exposure, multiple pregnancy, infant gender, short stature, low pre-pregnancy weight / low body mass index (BMI), number of pregnancies, fetal gender, diabetes, hypertension, hypothyroidism, asthma, educational level, smoking, and urogenital infections.

7. the risk indicator is BMI, and the BMI is 22 kg / m ２ Higher and 37 kg / m ２ or less than 37 kg / m ２ The composition of claim 6, wherein the composition is equal to

8. 10. The composition of claim 1, wherein the biological sample is selected from the group consisting of whole blood, plasma, serum, amniotic fluid, vaginal secretions, saliva, and urine, and wherein the biological sample is obtained at 19-21 weeks gestation.

9. 10. The composition of claim 1, wherein the proteomics workflow comprises quantification of stable isotope labeled (SIS) standard peptides of the biomarkers.

10. The MS, a) matrix-assisted laser desorption / ionization time-of-flight mass spectrometry (MALDI-TOF MS); b) matrix-assisted laser desorption / ionization time-of-flight post-source decay (MALDI-TOF post-source decay (PSD)); c) matrix-assisted laser desorption / ionization time-of-flight / time-of-flight (MALDI-TOF / TOF); d) Surface-enhanced laser desorption / ionization time-of-flight mass spectrometry (SELDI-TOF MS); e) electrospray ionization mass spectrometry (ESI-MS); f) electrospray ionization mass spectrometry / mass spectrometry (ESI-MS / MS); g) Electrospray Ionization Mass Spectrometry / Mass Spectrometry ｎ (ESI-MS / (MS)n (n is an integer greater than zero)); h) electrospray ionization 3D (ESI 3D) or linear 2D ion trap mass spectrometry (ESI linear (2D) ion trap MS); i) electrospray ionization triple quadrupole mass spectrometry (ESI triple quadrupole MS); j) Electrospray ionization quadrupole orthogonal time-of-flight (ESI Q-TOF); k) Electrospray ionization Fourier transform mass spectrometry system (ESI Fourier transform MS system); l) Desorption / Ionization on Silicon (DIOS); m) Secondary ion mass spectrometry (SIMS); n) atmospheric pressure chemical ionization mass spectrometry (APCI-MS); o) Atmospheric pressure chemical ionization mass spectrometry / mass spectrometry (APCI-MS / MS); p) Atmospheric Pressure Chemical Ionization Mass Spectrometry ｎ (APCI-(MS)n) (n is an integer greater than zero); q) Ion Mobility Spectrometry (IMS); r) inductively coupled plasma mass spectrometry (ICP-MS); s) Atmospheric pressure photoionization mass spectrometry (APPI-MS); t) atmospheric pressure photoionization mass spectrometry / mass spectrometry (APPI-MS / MS); and u) Atmospheric Pressure Photoionization Mass Spectrometry ｎ (APPI-(MS)n) (n is an integer greater than zero) 10. The composition of claim 1, selected from the group consisting of:

11. 2. The composition of claim 1, wherein the MS comprises co-immunoprecipitation-mass spectrometry (co-IP MS), liquid chromatography-mass spectrometry (LC-MS), multiple reaction monitoring (MRM), or selected reaction monitoring (SRM).

12. 2. The composition of claim 1, wherein the biomarker for IBP4 comprises a peptide fragment of IBP4 comprising an amino acid sequence selected from the group consisting of QCHPALDGQR, LPGGLEPK, THEDLYIIPNCDR, and EDARPVPQGSCQSELHR, and the biomarker for SHBG comprises a peptide fragment of SHBG comprising an amino acid sequence selected from the group consisting of IALGGLLLFPASNLR, GEDSSTSFCLNGLWAQGQR, DDWFMLGLR, SCDVESNPGIFLPPGTQAEFNLR, TWDPEGVIFYGDTNPK, VVLSSGSGPGLDLPLVLGLPLQLK, and ALALPPLGLAPLLNLWAKPQGR.

13. 13. The composition of claim 12, wherein the peptide fragment of IBP4 comprises QCHPALDGQR and the peptide fragment of SHBG comprises IALGGLLLFPASNLR.

14. A method for analyzing inverted values ​​of pairs of biomarkers as an indicator of whether a pregnant human patient should be provided with prophylactic treatment for preterm birth, said method comprising: (i) measuring a panel of isolated biomarkers comprising IBP4 and SHBG in a biological sample obtained from said patient, wherein measuring comprises subjecting said biological sample to a proteomics workflow consisting of mass spectrometry (MS); (ii) calculating a risk score using the inverted value of IBP4 / SHBG; wherein if the risk score is greater than a reference risk score, the patient is indicated to receive a treatment regimen comprising progesterone treatment, cervical cerclage, serial cervical length measurement, cervical pessary, or antenatal corticosteroids.

15. 15. The method of claim 14, wherein the treatment regimen further comprises an enhanced monitoring and clinical management regimen including one or more of: (a) more frequent prenatal visits, (b) increased education regarding the signs and symptoms of early preterm labor, or (c) altered treatment for diabetes and / or hypertension, compared to a pregnant human patient not at risk of preterm birth.

16. 15. The method of claim 14, wherein the progesterone treatment comprises administration of a progestogen, 17-alpha hydroxyprogesterone caproate, or vaginal progesterone.

17. 17. The method of claim 16, wherein the administration of hydroxyprogesterone 17-alpha caproate is by injection and the administration of vaginal progesterone is in gel form.

18. 15. The method of claim 14, further comprising the initial step of detecting one or more measurable risk indicator features selected from the group consisting of age, previous pregnancies, history of previous low birth weight or preterm birth, multiple second trimester spontaneous abortions, previous first trimester induced abortions, familial and intergenerational factors, history of infertility, nulliparity, placental abnormalities, cervical and uterine abnormalities, pregnancy bleeding, intrauterine growth restriction, intrauterine diethylstilbestrol exposure, multiple pregnancy, infant gender, short stature, low pre-pregnancy weight / low body mass index (BMI), number of pregnancies, fetal gender, diabetes, hypertension, hypothyroidism, asthma, learning level, smoking, and genitourinary infections.

19. the risk indicator is BMI, and the BMI is 22 kg / m ２ Higher and 37 kg / m ２ or less than 37 kg / m ２ 20. The method of claim 18, wherein:

20. 15. The method of claim 14, wherein the biological sample is selected from the group consisting of whole blood, plasma, serum, amniotic fluid, vaginal secretions, saliva, and urine, and wherein the biological sample is obtained at 19-21 weeks gestation.

21. 15. The method of claim 14, wherein the proteomics workflow comprises quantification of stable isotope labeled (SIS) standard peptides of the biomarkers.

22. The MS, a) matrix-assisted laser desorption / ionization time-of-flight mass spectrometry (MALDI-TOF MS); b) matrix-assisted laser desorption / ionization time-of-flight post-source decay (MALDI-TOF post-source decay (PSD)); c) matrix-assisted laser desorption / ionization time-of-flight / time-of-flight (MALDI-TOF / TOF); d) Surface-enhanced laser desorption / ionization time-of-flight mass spectrometry (SELDI-TOF MS); e) electrospray ionization mass spectrometry (ESI-MS); f) electrospray ionization mass spectrometry / mass spectrometry (ESI-MS / MS); g) Electrospray Ionization Mass Spectrometry / Mass Spectrometry ｎ (ESI-MS / (MS)n (n is an integer greater than zero)); h) electrospray ionization 3D (ESI 3D) or linear 2D ion trap mass spectrometry (ESI linear (2D) ion trap MS); i) electrospray ionization triple quadrupole mass spectrometry (ESI triple quadrupole MS); j) Electrospray ionization quadrupole orthogonal time-of-flight (ESI Q-TOF); k) Electrospray ionization Fourier transform mass spectrometry system (ESI Fourier transform MS system); l) Desorption / Ionization on Silicon (DIOS); m) Secondary ion mass spectrometry (SIMS); n) atmospheric pressure chemical ionization mass spectrometry (APCI-MS); o) Atmospheric pressure chemical ionization mass spectrometry / mass spectrometry (APCI-MS / MS); p) Atmospheric Pressure Chemical Ionization Mass Spectrometry ｎ (APCI-(MS)n) (n is an integer greater than zero); q) Ion Mobility Spectrometry (IMS); r) inductively coupled plasma mass spectrometry (ICP-MS); s) Atmospheric pressure photoionization mass spectrometry (APPI-MS); t) atmospheric pressure photoionization mass spectrometry / mass spectrometry (APPI-MS / MS); and u) Atmospheric Pressure Photoionization Mass Spectrometry ｎ (APPI-(MS)n) (n is an integer greater than zero) 15. The method of claim 14, selected from the group consisting of:

23. 15. The method of claim 14, wherein the MS comprises co-immunoprecipitation-mass spectrometry (co-IP MS), liquid chromatography-mass spectrometry (LC-MS), multiple reaction monitoring (MRM), or selected reaction monitoring (SRM).

24. 15. The method of claim 14, wherein the biomarker for IBP4 comprises a peptide fragment of IBP4 comprising an amino acid sequence selected from the group consisting of QCHPALDGQR, LPGGLEPK, THEDLYIIPNCDR, and EDARPVPQGSCQSELHR, and the biomarker for SHBG comprises a peptide fragment of SHBG comprising an amino acid sequence selected from the group consisting of IALGGLLLFPASNLR, GEDSSTSFCLNGLWAQGQR, DDWFMLGLR, SCDVESNPGIFLPPGTQAEFNLR, TWDPEGVIFYGDTNPK, VVLSSGSGPGLDLPLVLGLPLQLK, and ALALPPLGLAPLLNLWAKPQGR.

25. 25. The method of claim 24, wherein the peptide fragment of IBP4 comprises QCHPALDGQR and the peptide fragment of SHBG comprises IALGGLLLFPASNLR.

Images