Biomarker pairs 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, ISBN9789241503433 (2012). An estimated one 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 life 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 nevertheless 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 helped with feasible, 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 for death and a variety of health and development problems than infants born at term. Complications include acute respiratory, gastrointestinal, immune, central nervous system, hearing, and visual problems, as well as long-term motor, cognitive, visual, hearing, behavioral, social-emotional, health, and growth problems. The birth of a preterm infant also results in significant emotional and economic costs to 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 ages. However, infants born closer to term represent the majority of infants born preterm and experience more complications than infants born at term.

[0006] To prevent premature 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. For women less than 34 weeks pregnant and in active premature labor, hospitalization and administration of medications to temporarily stop premature labor and / or promote fetal lung development may be required. If a pregnant woman is determined to be at risk for premature birth, health care providers can implement a variety of clinical strategies, which may include preventive medications, such as hydroxyprogesterone caproate (Makena) injections and / or vaginal progesterone gel, cervical pessaries, restrictions on sexual activity and / or other physical activity, and changes in treatment for chronic conditions that increase the risk of premature labor, such as diabetes and hypertension.

[0007] There is a high need to identify women at risk for preterm birth and provide them with appropriate prenatal care. In women identified as high risk, more intensive prenatal surveillance and preventive interventions can be planned. Current strategies for risk assessment are based on obstetric and medical history and clinical examination, but these strategies can only identify a small percentage of women at risk for preterm birth. Currently, a prior history of spontaneous PTB (sPTB) is the strongest single predictor of subsequent PTB. If a woman has had one previous sPTB, the chance of a second PTB is 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 women who ultimately deliver preterm have abnormalities in multiple molecular pathways. Reliable early identification of risk for preterm birth would allow for planning appropriate monitoring and clinical management 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 critical to the cost-effective allocation of monitoring resources. Despite the intense research being conducted to identify at-risk women, PTB prediction algorithms based solely on clinical and demographic factors or using serum or vaginal biomarkers measured have not yet been developed into clinically useful tests.To allow clinical intervention, more accurate methods are needed to identify at-risk women during first 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, ISBN9789241503433(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 a pregnant woman.

[0010] 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, 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 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 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, 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, which 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, which 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, which biomarker pairs show 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 pairs of biomarkers shows an inverse value change 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, where the pair of biomarkers show inverted 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.

[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 said pair of biomarkers exhibits inverted 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.

[0020] In certain embodiments, the invention provides an isolated biomarker pair IBP4 / SHBG that exhibits an inverse value change among pregnant women at risk for preterm birth compared to full-term controls, hi further embodiments, the invention provides an isolated biomarker pair IBP4 / SHBG that exhibits a higher ratio in pregnant women at risk for 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, where 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 a stable isotope labeled standard peptide (SIS peptide) of 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 said pairs exhibiting inverted 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 said 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 of said pairs exhibiting inverted 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 said biomarkers.

[0024] In additional embodiments, the invention provides a panel of at least two pairs of surrogate peptides, each of said pairs 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, each of said pairs exhibiting inverted 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) of each of the surrogate peptides.

[0025] In additional embodiments, the invention provides a panel of at least two pairs of surrogate peptides, each of said 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, each of said pairs showing inverted 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) of each of the surrogate peptides.

[0026] In a further embodiment, the invention provides a panel of at least two pairs of surrogate peptides, each of said pairs 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 at least one of said pairs shows an inverted value change 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) of each of the surrogate peptides.

[0027] In a further embodiment, the invention provides a panel of at least two pairs of surrogate peptides, each of said 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 at least one of said pairs shows an inverted value change 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) of each of the surrogate peptides.

[0028] In additional embodiments, the invention provides a panel of at least two pairs of surrogate peptides, each of said pairs 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 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) of each of the surrogate peptides.

[0029] In additional embodiments, the invention provides a panel of at least two pairs of surrogate peptides, each of said 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 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) of 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 a 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. In some embodiments, the pregnant woman is 22 kg / m 2 and 37kg / 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 a 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 37kg / 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 the SIS surrogate peptides of each of the biomarkers.

[0032] In some embodiments, determining the probability of preterm birth in a pregnant woman involves an initial step that includes 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. 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 with documented gestational age at birth. In some embodiments, determining the probability of preterm birth in a pregnant woman includes measuring the ratio of IBP4 / SHBG and comparing the value to the index to derive the risk of preterm birth using the same isolation and measurement techniques that 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 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. In some embodiments, the pregnant woman is 22 kg / m 2 and 37kg / 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 the SIS surrogate peptides of 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 the reversal 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. 2 and 37kg / 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 the SIS surrogate peptides of 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 a change in reversal 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 another embodiment, the 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, 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 that there is a change in the relative intensity of the individual biomarkers between the pregnant woman and a full-term control, and are indicative of 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 37kg / 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 the SIS surrogate peptides of 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 the 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 37kg / 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 the SIS surrogate peptides of each of the biomarkers.

[0037] In one embodiment, the 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 is 22 kg / m 2 and 37kg / 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 the SIS surrogate peptides of 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 inverted values ​​of the biomarker pairs 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 37kg / 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 the SIS surrogate peptides of 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 to a first member of the pair and a second capture agent that specifically binds to a 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 the steps of: a. obtaining a biological sample from the pregnant woman; 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 of IBP4 to the capture agent and SHBG to the capture agent. In one embodiment, the method comprises measuring a reversal value of the pair of biomarkers. In a further embodiment, the presence of a change in the 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 an additional 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 that there is a change in the relative intensity of the individual biomarkers between the pregnant women and full-term controls, and are indicative of the pregnant women'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 women. 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 individual pregnancies to the index to derive the risk of preterm birth, using the same isolation and measurement techniques 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 shows an inverse value change 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 shows inverse value changes between pregnant women at risk of preterm birth and full-term controls. (Item 3) 3. The composition according to item 2, further comprising a stable isotope-labeled standard peptide (SIS peptide) 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 said 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 said pairs of biomarkers shows inverted value changes between pregnant women at risk of preterm birth and full-term controls. (Item 6) 6. The panel of claim 5, further comprising a SIS peptide corresponding to each of the surrogate peptides. (Item 7) 1. A method for determining a 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 a probability of preterm birth in a pregnant woman, comprising measuring a change 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 to determine 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 said pregnant woman and a full-term control is indicative of said probability of preterm birth in said 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) 11. The method of claim 10, wherein the measuring further comprises measuring a stable isotope-labeled standard peptide (SIS peptide) 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) 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 claim 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 according to item 7 or 8, further comprising an initial step of detecting one or more measurable characteristics of a risk indication. (Item 21) 18. The method of claim 17, wherein the risk indicia is selected from the group consisting of body mass index (BMI), parity, and fetal sex. (Item 22) 22. The method of claim 21, wherein the risk indicator is BMI. (Item 23) 9. The method according to item 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 37kg / m 2 23. The method according to item 22, wherein the subject has a body mass index (BMI) less than or equal to: (Item 25) 26. A method for determining a probability of preterm birth in a pregnant woman, 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 probability of preterm birth in said pregnant woman. (Item 26) 1. A method for determining a pregnant woman's likelihood of premature birth, comprising measuring inverted values ​​of the biomarker pair IBP4 and SHBG in a biological sample obtained from said pregnant woman to determine said pregnant woman's likelihood of premature birth. (Item 27) 27. The method of claim 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 female. (Item 29) 27. The method of claim 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 claim 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 features indicative of risk. (Item 38) 38. The method of claim 37, wherein the risk indicia is 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 37kg / 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 pregnant women compared to full-term controls 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 to a first member of the pair and a second capture agent that specifically binds to a 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. The method includes: (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 inversion value between the pregnant woman and a full-term control is indicative of the probability of preterm birth in 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 claim 47, wherein the biological sample is serum. (Item 49) 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, carried out by an assay selected from the group consisting of enzyme immunoassay (EIA), enzyme-linked immunosorbent assay (ELISA), and radioimmunoassay (RIA). (Item 52) 45. The method of claim 44, further comprising detecting one or more measurable characteristics of the risk indicator. (Item 53) 53. The method of claim 52, wherein the risk indicia is selected from the group consisting of body mass index (BMI), parity, and fetal sex. (Item 54) 54. The method of claim 53, wherein the risk indicator is BMI. (Item 55) The pregnant woman is 22 kg / m 2 and 37kg / m 2 55. The method according to item 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 said capture agent, and between SHBG and said capture agent. The method includes: (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 biomarker pair is present in said biological sample comprising subjecting said sample to a proteomic 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. The method includes: (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 inversion value between said pregnant woman and a full-term control is indicative of said probability of preterm birth in said 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 claim 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 proteomic workflow comprises quantifying SIS peptides. (Item 67) 60. The method of claim 58 or 59, further comprising detecting one or more measurable characteristics indicative of risk. (Item 68) 68. The method of claim 67, wherein the risk indicia is selected from the group consisting of body mass index (BMI), parity, and fetal sex. (Item 69) 70. The method of claim 68, wherein the risk indicia is body mass index (BMI). (Item 70) The pregnant woman is 22 kg / m 2 and 37kg / m 2 70. The method of claim 69, wherein the subject has a body mass index (BMI) less than or equal to: [Brief description 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] [Diagram 2] Discovery, supporting, and verifying cases of GABD.

[0047] [Diagram 3]Protein Expression During Pregnancy. Various proteins can be analyzed based on known protein behavior and knowing which proteins / pathways 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] [Diagram 5] Protein pathology during pregnancy. Insulin-like growth factor binding protein 4 (IBP4) was overexpressed by at least 10% in the blood collection window 19-21 weeks. Sex hormone-binding globulin (SHBG) was underexpressed by at least 10% in the blood collection window 19-21 weeks.

[0050] [Figure 6] Supporting Selection Criteria. Figure 6 describes the criteria for implementing a high performance preterm 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 taken 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 determined biochemically using multiple analytes that increase during pregnancy but do not differ in PTB cases and controls. Biochemical dating can be useful to confirm dating by date of 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 proteomic 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] Highlighted sex hormone binding globulin (SHBG) biology. 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] Interaction of 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 indicate suppression of IGF2 activity. PTB cases suppress the levels of PAPPA, PRG2, and increase the levels of IBP4.

[0063] [Figure 19] Insulin-like growth factor binding protein 4 (IBP4). IBP4 is upregulated in PTB cases. 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 control of IGF2 activity at the maternal-fetal interface. 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 across two orthogonal platforms ensures reliability of analyte measurements.

[0065] [Figure 21] PTB classification by IBP4 / SHBG in discovery samples from GABD at 19-21 weeks. Discovery samples from 19-21 weeks of gestation were classified according to 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 that increase at an accelerating rate throughout pregnancy.

[0067] [Diagram 23] Figure 23 summarizes the distribution of PAPR studies.

[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] [Diagram 25] FIG. 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 for any unknown subject to be determined. The top line (purple) under the risk curve graph corresponds to GAB < 35 0 / 7 weeks, the second line (red) from the top corresponds to GAB 35 0 / 7-37 0 / 7 / week, the third line (green) corresponds to GAB 37 0 / 7-39 0 / 7 / week, and the fourth line (blue) from the top corresponds to GAB 39 0 / 7 weeks ≦ GAB.

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

[0071] [Figure 27] Figure 27 shows the ROC curves corresponding to predictor performance using a combination of subjects from 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 analysis filters.

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

[0074] [Diagram 30] Figure 30 shows that the average AUROC obtained from 2,000 bootstrap replicates was 0.76. The blinded IBP4 / SHBG AUROC performance of the corroboration 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 corroboration, the discovery and corroboration samples were combined for bootstrap performance determination.

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

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

[0077] [Diagram 33] FIG. 33 shows the immunoassay vs. MS ROC analysis for BMI >22 and <37 or equal.

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

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

[0080] [Diagram 36] FIG. 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 a proteomic analysis method from Sera Prognostics that involves immunodepletion of samples, enzymatic digestion, and analysis on an Agilent 6490 mass spectrometer.

[0082] [Figure 38] FIG. 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]FIG. 39 shows a comparison of SHBG measurements from Abbott Architect CMIA and Roche cobas e602 analyzers, both semi-automated immunoassay machines.

[0084] [Diagram 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 is located within the thyroglobulin type 1 domain. IBP4 has a single N-linked glycosylation site at residue 125.

[0085] [Diagram 41] FIG. 41 highlights the location of the QCHPALDGQR peptide in the two IBP4 isoforms.

[0086] [Diagram 42] Figure 42 shows the domains and structural features of the longest isoform of the SHBG protein (Uniprot: P04278). The SHBG IALGGLLFPASNLR (aa, 170-183) peptide is located in the first lamin G-like domain. SHBG has three glycosylation sites; two N-linked sites at residues 380 and 396; one O-linked site at residue 36.

[0087] [Diagram 43] FIG. 43 highlights the location of the IALGGLLFPASNLR peptide in exon 4 within the seven isoforms of SHBG.

[0088] [Diagram 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 was analyzed by smoothing over a sliding 10-day window. Case vs. control signals correspond approximately to a maximum of 10% difference.

[0089] [Diagram 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 was analyzed by smoothing over a sliding 10-day window. Case vs. control signals correspond approximately to a maximum 10% difference.

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

[0091] [Figure 47] FIG. 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 investigated the diagnostic performance of inversions formed in multiple different proteins by ROC analysis. The range of AUC values ​​(sPTB cases vs. term controls) using a dataset derived from samples collected at 19 / 0-21 / 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, the bottom panel shows that the p-values ​​derived from the Wilcoxon test (sPTB cases vs. term controls) of inversions are more significant than those of 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 reversal score calculations. 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 the individual proteins. These data show 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 CV of multiple inversions and their individual up- and downregulated proteins. To investigate whether the amplification of diagnostic signals by 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 a dataset derived from samples collected at 19 / 0-21 / 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 downregulated 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 ratios 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 for diabetes annotation. 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] [Diagram 53] FIG. 53 shows hierarchical clustering of analyte response ratios.

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

[0099] [Figure 55] FIG. 55 shows a kinetic plot of differentially expressed proteins functioning 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 on metabolic hormone balance.

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

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

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

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

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

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

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

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

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

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

[0112] [Figure 68] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 82 shows kinetic plots of the analytes designated in Panels AI using data from 17 weeks 0 to 28 weeks 6.

[0127] [Figure 83] FIG. 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] FIG. 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] FIG. 85 shows kinetic plots of the peptide transitions designated in panels A-G using GABD data from 17 weeks 0 to 28 weeks 6.

[0130] [Figure 86] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] FIG. 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] [Fig. 106] FIG. 106 shows IBP4 and SHBG levels and IBP4 / SHBG reversal values ​​for sPTB cases and controls separately.

[0151] [Figure 107] FIG. 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 weeks preterm labor (PTL) without PPROM.

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

[0154] [Figure 110] Figure 110 is a risk curve showing 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 is a risk curve showing 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 from the top (red) corresponds to sPTB (35≦GAB<37 weeks), the third line (green) corresponds to TERM (37≦GAB<39 weeks), and the fourth line (blue) corresponds to TERM (39 weeks≦GAB). DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0156] (Detailed Description) 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 show increased risk of preterm birth compared to controls.The present disclosure is more particularly based in part on the unexpected discovery that the inversion value 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 inversion 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 inversion pair, or by panel of biomarkers / inversion pairs.The inversion value is the ratio of the relative peak area of ​​the upregulated biomarker to the relative peak area of ​​the downregulated biomarker, and serves as both variability normalization and diagnostic signal amplification. The present invention resides, in part, in the selection of particular biomarkers that, when paired together, can predict the probability of preterm birth based on the inversion value. Thus, underlying the present invention is human ingenuity in the selection of particular 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 ​​the upregulated analyte to the relative peak area of ​​the downregulated analyte, and serves both to normalize variability and to amplify diagnostic signals. Among 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 ​​the upregulated biomarker to the relative peak area of ​​the downregulated biomarker, 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. Thus, the present invention is based in part on identifying biomarker pairs whose relative expression is inverted between PTB and non-PTB, showing a change in inversion value. The use of biomarker ratios in the methods disclosed herein allows for correction of variability resulting from human manipulations after removal of biological samples from pregnant women. Such variability may be introduced, for example, during sample collection, processing, depletion, digestion, or any other step of the method used to measure the biomarkers present in the sample, and is independent of how the biomarkers behave 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 variability and to amplify the diagnostic signal, although it is also contemplated that the biomarker pairs of the present invention can be measured by any other means, such as by subtraction, addition, or multiplication of the relative peak areas. The methods disclosed herein encompass the measurement of biomarker pairs by such other means.

[0159] This method is advantageous because it does not rely on data standardization, helps to avoid overfitting, and provides the simplest possible classifier, resulting in a very simple experimental test that is easily implemented in the clinic. The use of marker pairs based on inversion value changes that do not rely on data standardization has enabled the development of clinically relevant biomarkers as disclosed herein. Since the quantification of any single protein is subject to uncertainties 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 allows for a robust method of personalized diagnosis and prognosis.

[0160] The present disclosure provides a biomarker reversal pair and a related panel 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, and appropriate monitoring and clinical management can be initiated accordingly to prevent preterm birth.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 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 change in reversal 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 a 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 having 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 a number of art-known proteins in the context of the present invention and provides exemplary accession numbers associated with one or more public databases as well as exemplary references to published journal articles related to these art-known proteins. However, one 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), 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 including the pairs listed in any of the accompanying figures, including FIG. 1, and tables.

[0165] In some embodiments, the present invention provides a method for determining a probability of preterm birth in a pregnant woman, comprising measuring in a biological sample obtained from the pregnant woman at least one pair of reversal values ​​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 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 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 for 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 for preterm birth compared to full-term controls.

[0170] In some embodiments, the present invention provides a method for determining a 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, and CATD / SHBG in a biological sample obtained from the pregnant woman to determine the probability 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 a method for determining a probability of preterm birth in a pregnant woman, comprising measuring a reversal level of IBP4 / SHBG in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in said 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 probability of preterm birth in a pregnant woman, comprising measuring inversion 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 said pregnant woman, wherein the presence of a change in inversion values ​​between the pregnant woman and a full-term control determines the probability 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] Included within the embodiments of the present invention is an iterative method for determining the probability of preterm birth in a pregnant woman, comprising measuring inversion 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 said pregnant woman, wherein the presence of a change in inversion values ​​between the pregnant woman and a full-term control determines the probability of preterm birth in the pregnant woman. Iterative implementation 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 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] The additional markers may 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 may include, for example, previous low birth weight or early 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 measurements, pregnancy bleeding, intrauterine growth restriction, intrauterine diethylstilbestrol exposure, multiple pregnancy, infant sex, short stature, low pre-pregnancy weight, low or high body mass index, diabetes, hypertension, genitourinary infections (i.e., urinary tract infections), asthma, anxiety and depression, asthma, hypertension, hypothyroidism. Demographic risk indicators for preterm birth may 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 (In: Premature Birth: Causes, Consequences, and Prevention, Institute of Medicine (US) Committee on Understanding Premature Birth and Assuring 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 of skill in the art and are 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" are inclusive of plural referents and are used interchangeably with "at least one" and "one or more," unless the content clearly dictates otherwise.

[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 may 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 has therefore 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 change and / or detection of which may correlate 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 that include 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, e.g., peptide fragments of proteins or polypeptides comprising at least 5 contiguous amino acid residues, at least 6 contiguous amino acid residues, at least 7 contiguous amino acid residues, at least 8 contiguous amino acid residues, at least 9 contiguous amino acid residues, at least 10 contiguous amino acid residues, at least 11 contiguous amino acid residues, at least 12 contiguous amino acid residues, at least 13 contiguous amino acid residues, at least 14 contiguous amino acid residues, at least 15 contiguous amino acid residues, at least 5 contiguous amino acid residues, at least 16 contiguous amino acid residues, at least 17 contiguous amino acid residues, at least 18 contiguous amino acid residues, at least 19 contiguous amino acid residues, at least 20 contiguous amino acid residues, at least 21 contiguous amino acid residues, at least 22 contiguous amino acid residues, at least 23 contiguous amino acid residues, at least 24 contiguous amino acid residues, at least 25 contiguous amino acid residues, or more contiguous amino acid residues.

[0183] As used herein, the term "surrogate peptide" refers to a peptide that is 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 stable isotope-labeled standard surrogate peptides ("SIS surrogate peptides" or "SIS peptides") in conjunction with MRM detection techniques. The surrogate peptide may be synthetic. For example, a SIS surrogate peptide can be synthesized in which arginine or lysine is heavily labeled, or any other amino acid at the C-terminus of the peptide that serves as an internal standard for the MRM assay. A SIS surrogate peptide is not a naturally occurring peptide, and has significantly different structures and properties compared to its naturally occurring counterpart.

[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, HPX / PTGDS in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman, and the presence of a change in the ratio between the pregnant woman and a full-term control determines the probability of preterm birth in the pregnant woman. In some embodiments, the ratio may include upregulated proteins in the numerator and downregulated proteins in the denominator or both. For example, as exemplified herein, IBP4 / SHBG is the ratio of upregulated proteins in the numerator and downregulated proteins in the denominator, defined herein as "inversion." When the ratio contains an upregulated protein in the numerator or a downregulated protein in the denominator, the unregulated protein will play a normalization role (e.g., reduce preanalysis or analysis 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 the ratio of biomarkers.

[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 classes being compared. Detecting inversion of protein concentration or gene expression levels eliminates the need for data standardization or establishing a threshold value across a population. In some embodiments, the inversion pair is an isolated biomarker pair IBP4 / SHBG that shows an inversion value change between pregnant women at risk of preterm birth compared to full-term controls. In further embodiments, the inversion pair IBP4 / SHBG shows a higher ratio in pregnant women at risk of preterm birth compared to full-term controls. Within the definition of any inversion pair are included corresponding inversion pairs in which the individual biomarkers are swapped between the numerator and 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 ​​upregulated analytes to the relative peak area of ​​downregulated analytes, and serves both as variability normalization and diagnostic signal amplification. Among 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 ​​upregulated biomarkers to the relative peak area of ​​downregulated 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 to avoid overfitting, and provides the simplest possible classifier, resulting in a very simple experimental test that is easily implemented in clinics.The use of biomarker pairs based on inversion that does not rely on data standardization as described herein has a very high potential as a method for identifying clinically relevant PTB biomarkers.Since 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 show inverse value changes between pregnant women at risk for 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 trophoblast 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 villi 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 to term pregnancies. (Qiu et al. (above, 2012)

[0192] SHBG controls 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), and 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 laboratories., Cell Mol Immunol., 2014;11(6) ):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 destined for sPTB may be the result of infection and / or inflammation. Thus, SHBG may be important in controlling 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, each of said pairs showing inverted value changes 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 said pairs 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, each of said pairs showing inverted 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 that there is 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 a surrogate peptide of the biomarker in a biological sample obtained from the pregnant woman.

[0199] In one embodiment, the present invention provides a method for determining a probability of preterm birth in a pregnant woman comprising measuring in a biological sample obtained from the pregnant woman the inverted 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 to determine the probability of preterm birth in the pregnant woman.

[0200] In an additional embodiment, the present invention provides a method for determining a 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 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 probability of preterm birth in the pregnant woman.

[0201] In a further 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 the biomarkers listed in Table 26 of the pregnant woman in a biological sample obtained from the pregnant woman to determine the probability of preterm birth in the pregnant woman.

[0202] In another embodiment, the invention provides a method for determining a probability of preterm birth in a pregnant woman comprising measuring a change in inversion values ​​of a panel of at least two pairs of biomarkers selected from the group consisting of biomarker pairs specified in any of Tables 1-77 and Figures 1-111 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 that there is 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.

[0203] In another embodiment, the invention provides a method for determining a 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 biomarker pairs specified in Tables 27-59, 61-72, 76, and 77 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 that there is 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.

[0204] In another embodiment, the invention provides a method for determining a 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 the biomarkers specified in Table 26 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 that there is 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.

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

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

[0207] Additionally, while the present disclosure has been described and exemplified with reference to a method for determining the probability of preterm birth in a pregnant woman, it is equally applicable to other conditions, such as abnormal glucose testing, gestational diabetes, hypertension, preeclampsia, intrauterine growth retardation, stillbirth, fetal growth retardation, HELLP syndrome, oligohydramnios, chorioamnionitis, placenta previa, placental acreta, abruption, abruption, and the like. Applicable methods for predicting placenta, placental hemorrhage, preterm rupture of membranes, preterm labor, adverse cervix, post-term pregnancy, cholelithiasis, uterine over distention, 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 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 discovery allows prediction of TTB or GAB, either individually or mathematical combinations of such analytes. Analytes that lack case vs. 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 of other disorders. Such pregnancy clocks are useful to confirm age determination by another measurement (e.g., last menstrual period and / or ultrasound age determination) or to predict sPTB, GAB, or TTB alone after the fact and more accurately, for example. Such analytes, also referred to herein as "clock proteins," can be used without or 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 of determining the probability of preterm birth in a pregnant woman further comprises detecting a measurable feature for one or more risk indicators associated with preterm birth. In additional embodiments, the risk indicators are selected from the group consisting of previous low birth weight or early birth, multiple second trimester spontaneous abortions, previous first trimester induced abortions, familial and intergenerational factors, infertility history, nulliparity, primipara, multiparous, 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. The term further encompasses any characteristic, feature, or aspect that can be determined and correlated in the context of 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, the presence or amount of an altered structure, such as a post-translational modification, such as oxidation at one or more positions on the amino acid sequence of the biomarker, or the presence of an altered conformation, such as 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 may further include risk indicators including, for example, maternal characteristics, age, race, ethnicity, medical history, past pregnancy history, obstetric history, etc. For risk indicators, measurable characteristics may include, for example, previous low birth weight or early 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 measurements, 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, low academic achievement, smoking, drug use, and alcohol consumption.

[0212] In some embodiments, the methods of the 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 method of determining the probability of preterm birth in a pregnant woman includes communicating the probability to a health care provider. The disclosed method of predicting GAB, predicting full-term birth, determining the probability of full-term birth in a pregnant woman, and determining the time to birth in a pregnant woman also includes communicating the probability to a health care provider. Although described and illustrated above with reference to determining the probability of preterm birth in a pregnant woman, all embodiments described throughout this disclosure are equally applicable to the methods of predicting GAB, predicting full-term birth, determining the probability of full-term birth in a pregnant woman, and determining the time to birth in a pregnant woman. In particular, the biomarkers and panels listed throughout this application that explicitly refer to the methods for preterm birth can also be used in the methods of predicting GAB, predicting full-term birth, determining the probability of full-term birth in a pregnant woman, and determining the time to birth in a pregnant woman. It will be apparent to one skilled in the art that each of the above methods has specific and substantial utility and benefit with respect to maternal-fetal health considerations.

[0216] In additional embodiments, the communication informs subsequent treatment decisions for the pregnant woman. In some embodiments, the methods of determining the probability for preterm birth in a pregnant woman include 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 full-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 isolated biomarkers using the same measurement method used in the initial step of generating the probability / risk index, and comparing the ratios of the measured risk indexes 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 full-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 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 the risk score can be expressed based on various data transformations as well as as the ratio itself. Furthermore, those skilled in the art will understand that any ratio, especially with respect to inverted pairs, will provide similarly useful information with the numerator and denominator biomarkers swapped, or a related data transformation (e.g., subtraction) is applied. Since the level of the biomarkers may not be static throughout pregnancy, the standard score or reference score needs to be obtained for the pregnancy time point that corresponds to that of the pregnant woman when the sample is taken. The standard score or reference score can be pre-determined and incorporated into the predictive model, so that the comparison is indirect, rather than being actually 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 a standard or reference risk score, the pregnant woman may have an increased likelihood of preterm birth. In some embodiments, the magnitude of the pregnant woman's risk score, or the 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 is defined as the natural logarithm of the SIS normalized intensities of the IBP4 peptide transitions (QCHPALDGQR_394.5_475.2) and the SHBG peptide transitions (IALGGLLFPASNLR_481.3_657.4). Score = ln(P 1 n / P 2 n ), where P 1 n and P 2 n are the SIS normalized peak area values ​​of 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 association 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] While the PreTRM™ classifier is defined as the natural logarithm of the SIS normalized intensities of the IBP4 peptide transition (QCHPALDGQR_394.5_475.2) and SHBG peptide transition (IALGGLLFPASNLR_481.3_657.4), the present invention also includes 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 show strong predictive performance, for example, for separate GABD windows, preterm rupture of membranes before term (PPROM) vs. preterm labor without PPROM (PTL), fetal gender, primipara vs. multipara. This embodiment is illustrated in Example 10 and Table 61 for any 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 a combination of multiple inversions (SumLog) was evaluated across the entire blood draw range, and predictor scores were derived from the sum of the logarithmic values ​​of the individual inversions (SumLog). Those skilled in the art can choose other models (e.g., logistic regression) to construct predictors formed from more than one inversion.

[0221] The method of the present invention further includes 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, number of gestational years, and any discriminatory patient characteristics and / or risk factors described throughout this application. This embodiment is illustrated in Example 10, Tables 61-64, which independently illustrate the reversal performance (17-21 weeks) of two different phenotypes of sPTB, PPROM, and PTL. Similarly, this embodiment is illustrated in Example 10, Tables 76 and 77, and Figures 108 and 109, which independently illustrate the reversal performance (19-21 weeks) of two different phenotypes of sPTB, preterm rupture of membranes (PPROM) and preterm labor without PPROM (PTL). Thus, the method of the present invention includes selecting reversals to build independent predictors of PPROM and PTL, or to maximize performance overall using a combination of reversals of more than one of the single predictors as described above. This embodiment is further illustrated in Example 10, Tables 65-68, which independently illustrate the reversal performance of two different types of sPTB, primipara and multipara (17-21 weeks). This embodiment is further illustrated in Example 10, Tables 69-72 and Figure 106, which independently illustrate the reversal performance of two different types of sPTB based on fetal gender (17-21 weeks). The method of the present invention is illustrated with respect to PPROM and PTL, parity, and fetal gender, but includes a classifier that includes an indicator variable that selects one or a subset of reversals based on GABD, or any known clinical / risk factor described herein or otherwise known to one of skill in the art. As an alternative to having a classifier that includes an indicator variable, the present invention further provides separate classifiers that fit subsets of pregnant women based on GABD, or any known clinical / risk factor described herein or otherwise known to one 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 The BMI stratification can be improved by BMI stratification equal to or less than 100%. Thus, 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 the weight in kilograms of an individual divided by the square of the height in meters. Although BMI is not a direct measurement of body fat, there are studies that show that BMI correlates with more direct measurements of body fat obtained from skinfold thickness measurement, bioelectrical impedance analysis, densitometry (underwater weight measurement), dual energy X-ray absorptiometry (DXA), and other methods. Furthermore, BMI is a more direct measure of body adiposity, and therefore is believed to correlate strongly 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 more to 24.9 are considered normal weight, individuals with a BMI of 25.0 or more to 29.9 are considered overweight, and individuals with a BMI of 30.0 or more are considered obese. In some embodiments, the predictive performance of the claimed method can be improved by BMI stratification of 18 or more, 19 or more, 20 or more, 21 or more, 22 or more, 23 or more, 24 or more, 25 or more, 26 or more, 27 or more, 28 or more, 29 or more, or 30 or more. 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" includes any sample taken 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 secretion, 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, which delineate 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 that delineate preterm birth and full-term birth, as well as the cutoffs that delineate subcategories of preterm birth, can be adjusted when implementing the methods disclosed herein, for example, to maximize certain health benefits. In various embodiments of the invention, the cutoffs defining preterm birth include, 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 the artisan and are encompassed within the scope of the invention disclosed herein. Gestational age is a surrogate for the degree of fetal development and readiness of the fetus 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 aid in estimating gestational age. Premature birth is generally classified into two separate subgroups. One is spontaneous premature birth, which occurs following the spontaneous onset of premature labor or premature rupture of membranes, regardless of subsequent labor augmentation or Caesarean section. Two is medically indicated premature birth, which occurs following induction or Caesarean 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 premature birth or medically indicated premature birth. In some embodiments, the method disclosed herein is related to determining the probability of spontaneous premature birth or medically indicated premature 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 gestational age.

[0226] In some embodiments, the pregnant female is between 17-28 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 female is between 16-29 weeks gestation, between 17-28 weeks gestation, between 18-27 weeks gestation, between 19-26 weeks gestation, between 20-25 weeks gestation, between 21-24 weeks gestation, or between 22-23 weeks gestation at the time the biological sample is collected. In further embodiments, the pregnant female is between about 17-22 weeks gestation, between about 16-22 weeks gestation, between about 22-25 weeks gestation, between about 13-25 weeks gestation, between about 26-28 weeks gestation, or between about 26-29 weeks gestation at the time the biological sample is collected. Thus, the gestational age of the pregnant female at the time the biological sample is collected may 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 gestation. In certain embodiments, the biological sample is collected at 19-22 weeks gestation. In certain embodiments, the biological sample is collected at 19-21 weeks gestation. In certain embodiments, the biological sample is collected at 19-22 weeks gestation. In certain embodiments, the biological sample is collected at 18 weeks gestation. In further embodiments, the best performing inversions over consecutive or overlapping time windows can be combined with a single classifier to predict the probability of sPTB over a wider window of gestational age at 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, comprising the steps of: a. obtaining a biological sample from a 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 a first member of the pair and a second capture agent that specifically binds to a second member of the pair; and detecting binding between the first biomarker of the pair and the first capture agent 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 specified in Tables 27-59, 61-72, 76, and 77 in a pregnant woman, comprising the steps of: 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 a first member of the pair and a second capture agent that specifically binds to a second member of the pair; and detecting binding between the first biomarker of the pair and the first capture agent 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, comprising the steps of: 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 a first member of the pair and a second capture agent that specifically binds to a second member of the pair; and detecting binding between the first biomarker of the pair and the first capture agent and the second member of the pair and the second capture agent. In one embodiment, the present invention provides a method for detecting IBP4 and SHBG in a pregnant woman, comprising the steps of: a. obtaining a biological sample from the pregnant woman; 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 of IBP4 to the capture agent and SHBG to the capture agent. In one embodiment, the method comprises measuring a reversal value of the pair of biomarkers. In a further embodiment, the presence of a change in the 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 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 an additional 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, comprising the steps of: a. obtaining a biological sample from the pregnant woman; and b. subjecting the biological sample to a proteomic workflow consisting of mass spectrometry quantification to detect whether said isolated pair of biomarkers is present in the sample.

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

[0233] A "proteomics workflow" generally includes one or more of the following steps: thawing serum samples and depleting the 14 most abundant proteins by immunoaffinity chromatography; digesting the depleted serum with a protease, e.g., trypsin, to generate peptides; enriching the digest with a mixture of SIS peptides, desalting, and subjecting it to LC-MS / MS with a triple quadrapole instrument operated in MRM mode; forming a response ratio from the area ratio of the endogenous peptide peaks and the corresponding SIS peptide equivalent peaks. Those skilled in the art will recognize that other types of MS, e.g., MALDI-TOF or ESI-TOF, can be used in the methods of the invention. For example, in addition, those skilled in the art can modify the proteomics workflow by selecting certain 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, and 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 vs. undetectable amount) and / or amount (e.g., readout is absolute 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 that utilizes a capture agent. In further embodiments, the capture agent is an antibody, an antibody fragment, a nucleic acid-based protein binding reagent, a small molecule, or a variant thereof. In additional embodiments, the assay is an enzyme immunoassay (EIA), an enzyme-linked immunosorbent assay (ELISA), and a 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 that can volatilize / ionize analytes, form gas phase ions, and determine 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 matched with various other instruments that 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 on peptide mass, and preferably also on 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 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 carrying out quantitative MS to measure one or more biomarkers. Such quantitative methods can be performed in an automated (Villanueva et al., Nature Protocols (2006) vol. 1(2): 880-891) or semi-automated format. In certain embodiments, the MS can be operatively 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 tags (ICAT), tandem mass tags (TMT), or stable isotope labeling with amino acids in cell culture (SILAC), followed by chromatography and MS / MS.

[0237] As used herein, the term "multiple reaction monitoring (MRM)" or "selected reaction monitoring (SRM)" refers to an MS-based quantification method that is particularly useful for quantifying analytes that are in low abundance. In an SRM experiment, a predefined precursor ion and one or more of its fragments are selected by two mass filters of 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 the 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 term "schedule" or "dynamic" in reference to MRM or SRM refers to a variation of the assay where transitions for a particular analyte are acquired only in a time window around 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, since the retention time is a feature that depends on the physical properties of the analyte. A single analyte can also be monitored with more than one transition. Finally, a standard that corresponds to the analyte of interest (e.g., the same amino acid sequence) but differs 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 a precise level and used to quantify the corresponding unknown analyte. An additional level of specificity is contributed by the co-elution of the unknown analyte and its corresponding SIS as well as the characteristics of their transitions (e.g., similarity in the ratio of the levels of two transitions of the unknown and the ratio of two transitions of its 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 TOF (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 accomplished using modalities established in the art, such as collision-induced dissociation (CID). As described herein, detection and quantification of biomarkers by mass spectrometry can include multiple reaction monitoring (MRM), as described, for example, 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] One 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 blots, enzyme-linked immunosorbent assays (ELISAs), immunoprecipitation, immunohistochemistry, immunofluorescence, radioimmunoassays, 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 of skill 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 noncompetitive 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 yet 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 with antibodies, but they can be performed with any capture agent that can specifically bind 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 use radioactive labels (e.g., 125 I or 131Radioimmunoassay (RI-labeling) involves mixing a known amount of target analyte with an antibody specific for the analyte, then adding unlabeled analyte from the sample, and measuring the amount of labeled analyte displaced (see, for example, for guidance, An Introduction to Radioimmunoassay and Related Techniques, edited by Chard T, Elsevier Science 1995, ISBN 0444821198).

[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 antibody, stability requirements, and available equipment and disposal regulations. Those skilled in the art are familiar with the selection of suitable detectable labels 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 isothiocynate (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 More recent variations using chromatographic techniques, such as the use of chromatographic techniques (Microbial and Nanomaterials Biosystems, Foster City, Calif.), or 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 invention.

[0244] Chemiluminescence assays using chemiluminescent antibodies can be used for sensitive non-radioactive detection of protein levels. Antibodies labeled with fluorescent dyes can 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 suitable 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, spectrophotometers to detect color from chromogenic substrates; radiation counters to detect radiation, for example 125 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 invention can be automated or performed robotically, and signals from multiple samples can be detected simultaneously.

[0246] In some embodiments, the methods described herein include quantification of 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, the MRM or SRM can further include scheduled MRM or scheduled SRM.

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

[0248] 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 performance / 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 further method of peptide analysis, for example, as a peptide fractionation method 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 this 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, in particular 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 that can be identified 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 a complex mixture 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); Goding, Monoclonal Antibodies: Principles and Practice (2nd ed., 1986). The antibody capture agent 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 in the term. The antibody capture agent has a binding domain that is homologous or largely homologous to an immunoglobulin binding domain, and can be derived from natural sources or can be partially or wholly synthetically produced. The antibody capture agent can be a monoclonal or polyclonal antibody. In some embodiments, the antibody is a single chain antibody. Those skilled in the art will appreciate that the antibody can be provided in any of a variety of forms, including, for example, humanized, partially humanized, chimeric, chimeric humanized, and the like. 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 may be produced by any means. For example, the antibody capture agent may be enzymatically or chemically produced by fragmentation of an intact antibody, and / or it may be recombinantly produced from a gene encoding a partial antibody sequence. The antibody capture agent may include a single chain antibody fragment. Alternatively, or in addition, the antibody capture agent may include multiple chains linked together, for example, by disulfide bonds; and any functional fragment obtained from such molecules, in which such fragments retain the specific binding characteristics of the parent antibody molecule. Due to their smaller size as a functional component of the whole molecule, antibody fragments may offer advantages over intact antibodies for use in certain immunochemical techniques and experimental applications.

[0253] Suitable capture agents useful for the implementation of the present invention also include aptamers. Aptamers are oligonucleotide sequences that can specifically bind to their targets through a unique three-dimensional (3-D) structure. Aptamers can include any suitable number of nucleotides, and different aptamers can have the same or different number of nucleotides. Aptamers can be DNA or RNA or chemically modified nucleic acids, and can be single-stranded, double-stranded, or can include 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, and 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] It will be understood by those skilled in the art that biomarkers can be modified prior to analysis to improve their resolution or to determine their identity. For example, biomarkers can be subjected to proteolytic digestion prior to analysis. Any protease can be used. Proteases that can cleave biomarkers into a distinct number of fragments, such as trypsin, are particularly useful. The fragments resulting from digestion serve as fingerprints for the biomarkers, thereby indirectly allowing their detection. This is particularly useful when there are biomarkers with similar molecular weights that can be confused for the biomarker in question. Proteolytic fragmentation is also useful for biomarkers with high molecular weights, since 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 to improve binding to anionic adsorbents and improve detection resolution. In another example, biomarkers can be modified by attachment of tags of specific molecular weights 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 that are then probed for the presence of proteins. Alternatively, protein-binding molecules attached to microspheres, microparticles, microbeads, beads, or other particles can be used for capturing and detecting biomarkers. Protein-binding molecules can be antibodies, peptides, peptoids, aptamers, small molecule ligands, or other protein-binding capture agents attached to the surface of a particle. Each protein-binding molecule can contain a unique detectable label that is encoded, allowing it to be distinguished from other detectable labels attached to other protein-binding molecules and allowing the detection of biomarkers in multiplex 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., striped metal rods of submicron size, 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 the capture and detection of 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 substrate's surface. Frequently, the surface includes a plurality of addressable locations, each of which has a capture agent bound thereto. The capture agent can be a biological molecule, such as a polypeptide or a nucleic acid, which 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, a biological sample may be contacted with a panel comprising one or more polynucleotide binding agents.The expression of one or more of the detected biomarkers may then be evaluated by the methods disclosed below, for example, with or without nucleic acid amplification.Those skilled in the art will recognize that the methods described herein can automate the measurement of gene expression.For example, a system can be used that can perform multiplexed measurements of gene expression, for example, providing simultaneous digital measurements 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 invention can be used to detect polynucleotide biomarkers using 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. 605-16 (Persing et al., 2001)). (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 disclosed herein (E. G., et al., 2001 and subsequent latest editions). Methods for amplifying nucleic acids include, but are not limited to, the following: 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. 0 684 315), 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, the measurement of mRNA in a biological sample can be used as a surrogate for detecting the level of the corresponding protein biomarker in the biological sample. Thus, any of the biomarkers, biomarker pairs, or biomarker inversion panels described herein can also be detected by detecting the appropriate RNA. The mRNA level can be measured by reverse transcription quantitative polymerase chain reaction (RT-PCR, followed by qPCR). RT-PCR is used to generate cDNA from the mRNA. The cDNA can be used in a qPCR assay to produce fluorescence as the DNA amplification process proceeds. By comparison to 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 in combination with capillary electrophoresis have all been used to measure the expression level of mRNA in a sample. 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 premature birth in pregnant women.Detection of the expression level of one or more biomarkers and / or determination of the ratio of biomarkers can be used to determine the probability of premature birth in pregnant women.Such detection methods can be used, for example, for early diagnosis of conditions, to determine whether a subject has a predisposition to premature birth, to monitor the progress of premature birth or the progress of treatment protocols, to evaluate the severity of premature birth, to predict the outcome of premature birth and / or the prospect of recovery or full-term birth, or to help determine the appropriate treatment for premature birth.

[0261] The quantification of biomarkers in biological samples can be determined by the methods described above as well as any other methods known in the art, without being limited thereto.The quantitative data thus obtained is then subjected to an analytical classification process.In such a process, raw data is manipulated according to an algorithm that is predefined by a training set of data, for example as described in the examples provided herein.The algorithm can utilize the training set of data provided herein, or can use the guidelines provided herein to generate the algorithm using different sets of data.

[0262] In some embodiments, analyzing the measurable features to determine the probability of preterm birth in pregnant women comprises using a predictive model. In further embodiments, analyzing the measurable features to determine the probability of preterm birth in pregnant women comprises comparing the measurable features to reference features. As those skilled in the art can understand, 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 the measurable features to determine the probability of preterm birth in pregnant women comprises 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 comprises 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 to provide a classification of the sample. Examples of useful methods include linear discriminant analysis, recursive feature elimination, predictive analysis of microarrays, logistic regression, CART algorithms, FlexTree algorithms, LART algorithms, random forest algorithms, MART algorithms, machine learning algorithms, and the like.

[0264] To create a random forest for prediction of GAB, one skilled in the art can consider a set of k subjects (pregnant women) whose gestational age at birth (GAB) is known and N analytes (transitions) have been measured in blood samples taken a few weeks before delivery. The regression tree starts with a root node that includes all subjects. The average GAB for all subjects can be calculated at the root node. The variance of GAB in the root node will be high because there is a mix 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 calculated again. The variance of GAB in each branch will be lower than the root node because the subset of women in each branch have relatively more similar GABs than those in the root node. The two branches are created by selecting the analytes and a threshold for the analytes that creates the branches with similar GABs. The analytes and thresholds are chosen among the set of all analytes and thresholds, typically with a random subset of analytes at each node. The procedure continues to recursively generate branches, creating leaves (terminal nodes) where the 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 is obtained is classified as not belonging to the class of the reference data set. Conversely, if such a comparison is not statistically significantly different from the reference data set, then the sample from which the data set is 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 ROC curve) or accuracy at a particular value or range of values). Area under the curve measurements are useful for comparing the accuracy of classifiers across a complete data range. Classifiers with larger AUCs have 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 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 accuracy. As an alternative measurement, the desired quality threshold may refer to a predictive model that classifies samples with 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 AUC.

[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, where the two metrics have an inverse relationship.The limitations in the models described above can be adjusted to provide a selected level of sensitivity or specificity, 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] The raw data can be analyzed initially by measuring values ​​for each biomarker, usually in triplicate or multiple triplicates. The data can be manipulated, for example, the raw data can be transformed using standard curves and the averages of the triplicate 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, vol. 26:211-246 (1964)). The data is then input into a predictive model that classifies the sample according to condition. The resulting information can be communicated to the patient or health care provider.

[0269] To generate a prediction 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 criteria.As discussed above, different statistical methods can be used to obtain highly accurate prediction models.An example of such analysis is provided in Example 2.

[0270] In one embodiment, hierarchical clustering is performed in the derivation of 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 any qualitative feature into a quantitative feature; Hotelling T 2 They can be modified by sorting them by the achieved significance level, assessed by the sample reuse method for the statistics; and by appropriate application of the lasso method. Problems in prediction become problems in regression, without loss of vision in prediction, by making practical 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 methods claimed. 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). The name reflects the implementation of the Lasso through a binary tree; the lasso described; and what is called LARS by Efron et al. (2004) Annals of Statistics 32:407-451 (2004). 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 shrink 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, as in the case of lasso, and focusing attention on the few that are informative. This approach is available as PAM software and is widely used. Two additional sets of algorithms that can be used are Random Forest (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" that involve predictors that "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 set of distributions of correlation coefficients that would be 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 (mean, median, etc.) 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 false significant correlations (estimated from correlations in the randomized data set that are greater than this selected Pearson correlation) to the number of correlations in the empirical data that are greater than this selected Pearson correlation (significant correlation). This cutoff correlation value can be applied to the correlation between experimental profiles. Using the above distribution, a level of confidence is chosen for significance. This is used to determine the minimum value of the correlation coefficient that exceeds the result that would have been 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 the 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 of the individual "random correlation" distributions, it can be found how many observations are 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), where 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 each patient is observed, and the selection of proteomic and other features, a parametric approach to analyzing survival may be better than the widely applied semi-parametric Cox model. The Weibull parametric fit of survival allows the hazard ratio to be monotonically increasing, decreasing, or constant, and also has a proportional hazards representation (as in the Cox model) and an accelerated failure time representation. All standard tools available in 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 reduction of the number of covariates to a manageable size with lasso significantly simplifies the analysis, 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 are applicable to all types of proteomics data.A set of biomarkers, clinical data, and genetic data that can be easily determined and is highly informative regarding the probability of preterm birth in said pregnant women and the predicted time to preterm birth event is provided.Algorithm also provides information regarding the probability of preterm birth in pregnant women.

[0277] Therefore, those skilled in the art will understand that the probability of premature 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 features of each of the N biomarkers can be subjected to categorical data analysis to determine the probability of premature birth as a binary categorical outcome. Alternatively, the method of the present invention can analyze the measurable features of each of the N biomarkers by first calculating the quantitative variables, in particular the predicted gestational age at birth. The predicted gestational age at birth can then be used as the basis for predicting the risk of premature birth. By first using quantitative variables and then converting them into categorical variables, the method of the present invention takes into account the continuity of measurements detected for the measurable features. For example, rather than making a binary prediction of premature birth versus full-term birth, predicting the gestational age at birth allows treatment to be individualized for pregnant women. For example, an earlier predicted gestational age at birth will result in more intensive prenatal intervention, i.e., monitoring and treatment, than a predicted gestational age approaching 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 is 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, at least 6, up to a complete set of markers. Typically, a subset of markers is chosen that provides what is needed for quantitative sample analysis, e.g., availability of reagents, ease of quantification, etc., while maintaining a highly accurate predictive model. Selection of some informative markers for building a classification model requires the definition of a performance metric and a user-defined threshold to produce a model with useful predictive ability based on this metric. For example, the performance metric may 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, the analytical classification process can use any one of a variety of statistical analysis methods to manipulate quantitative data and provide classification of samples. Examples of useful methods include, but are not limited to, linear discriminant analysis, recursive feature elimination, predictive analysis of microarrays, logistic regression, CART algorithm, FlexTree algorithm, LART algorithm, random forest algorithm, MART algorithm, and machine learning algorithm. Various methods are used in training models. The selection of a subset of markers can be about forward selection or backward selection of marker subsets. The number of markers that optimizes the performance of the model without using all markers can be selected. One way to define the optimal number of items is to choose the number of items that produces a model with the desired predictive ability (e.g., AUC>0.75, or equivalent measurement of sensitivity / specificity) with 1 or less standard error from the maximum value obtained for this metric, using any combination and number of items used for a given algorithm.

[0281] In yet another embodiment, 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 a printed instruction 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 composition contained in the kit. The composition can be in liquid form or can be lyophilized. Suitable containers for the composition include, for example, bottles, vials, syringes, and test tubes. The container can be made of various materials, including glass or plastic. The kit can also include a package insert that includes instructions on how to determine the probability of premature birth.

[0283] From the foregoing description, it will be apparent that variations and modifications can be made to the invention described herein to adapt it to various usages and conditions, such embodiments also falling 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 and publication was specifically and individually indicated to be incorporated by reference.

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

[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, as well as relevant information regarding patient demographic characteristics, past medical and reproductive history, current pregnancy history, 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, aliquoting of samples into 2-D barcoded cryovials, and subsequent freezing at -80°C.

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

[0289] Serum samples were depleted of high abundance proteins using the Human 14 Multiple Affinity Removal System (MARS14). This removes 14 of the most abundant proteins that are treated as not providing useful information regarding the identification of disease-related changes in the serum proteome. For this purpose, equal volumes (50 μl) of each clinical pooled human serum sample (HGS) or pooled human pregnant female serum sample (pHGS) were 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, Cat. No. 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 h (±1 h) 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 Empore C18 96-well solid-phase extraction plates (3M Bioanalytical Technologies). The plates were 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 for 30 min at −80°C, and then dried by lyophilization. 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 performed on a 120EC-C18 column (2.1×100 mm, 2.7 μm) using a 30 min 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 the scheduled Multiple Reaction Monitoring method (sMRM). The sMRM assay monitored 898 transitions in which 259 biological peptides and 190 IS peptides (187 SIS + 3 IS) were measured, representing 148 proteins. Chromatographic peaks were integrated using Mass Hunter Quantitative Analysis software (Agilent Technologies).

[0292] Data analysis

[0293] Analysis of the discovery and corroborating 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 corroborating sample set. In the second stage, the combined discovery and corroborating 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. Among 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, 8). For each panel size within the window, Monte Carlo cross-validation (MCCV) was performed by training and testing a logistic classifier for 1,000 iterations on 70% and 30% of the samples, respectively. The size 4 panel, determined to be optimal by the 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 of 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) of sensitivity ranging from 0.7 to 1, or correlation between classifier output scores and time to birth values ​​(TTB) (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 were selected for each panel size for each window. 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) for each window, were used to train a logistic classifier on the discovery samples and generate classification scores for the corroborating samples in a blinded manner.

[0299] The performance of all reversal 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 corroborating datasets were combined and reanalyzed. Because the expression of diagnostic proteins may 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. The relative levels of proteins were expressed as the ratio of endogenous peptide peak area to its corresponding SIS standard (relative ratio). Examples of proteins that increase with pregnancy but show non-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 differences between PTB cases and controls (Figure 5). These proteins are obvious diagnostic candidates for developing a PTB classifier. The impact of inversion formation was also illustrated using the ratio of overexpressed proteins to the ratio of underexpressed proteins (Figures 8 and 21). It is clear that this leads to an increase in the separation of PTB cases and controls. Previous analyses have suggested that the levels of some analytes may be influenced 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, the impact of BMI on separation was investigated by displaying the inversion values ​​across pregnancies only for patients whose BMI is less than 35 (Figure 21). This leads to a further improvement in separation.

[0302] Reversal selection and classifier development on the combined discovery and corroboration 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 reversal candidates. To evaluate the performance of the panel, the reversal 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 reversal for that sample. This type of classifier has no coefficients, which helps to avoid overfitting problems. Anyone skilled in the art can derive a similar logistic classifier using the same analytes by well-established techniques. The multivariate performance of the panel of three top reversals formed from four proteins is shown in Figure 8 as a histogram of AUC values ​​obtained by cross-validation and in the ROC curve. Previous analysis has suggested that the levels of some analytes may be influenced 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 (Figure 10). TTB is defined as the difference between GABD and gestational age at birth (GAB). This allows for 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 In this example, validation of IBP4 / SHBG sPTB predictors identified in a large-scale maternal serum proteomics effort in asymptomatic women in early pregnancy is presented.

[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 / 7–28 6 / 7 weeks GA were enrolled. Best clinical estimates of gestational age were prepared using a predefined protocol of menstrual day determination confirmed by early ultrasound biometry or ultrasound alone to establish days. Body mass index (BMI) was derived from height and self-reported prepregnancy weight. Pregnancies with multiple gestation and pregnancies with known or suspected significant fetal anomalies were excluded. Relevant information on subjects' demographic characteristics, past medical and reproductive history, current pregnancy history, and concomitant medications was collected and entered into an electronic case report form. After delivery, data on maternal and infant outcomes and complications were collected. All deliveries were adjudicated as full term (≥37 0 / 7 weeks GA), spontaneous preterm (including preterm 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 validation studies.

[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 of predictor development

[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: 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] Validation nested case / control analyses were performed on prespecified sPTB case and control samples unrelated to discovery and corroboration. 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 inclusion and exclusion criteria, and confirmed medical / pregnancy complications and birth GA allocation for all subjects at the time of sample collection and delivery. Detailed analysis protocols were pre-established, including validation study design, analysis plan, and blinding protocols. Personnel other than the Chief Clinical Officer (DCO) and clinical data manager were blinded to the subject case, control, and birth GA data allocation. Data analysis plans included prespecified validation claims and protocols for double 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. Areas 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 the validation results.

[0312] Validation study design

[0313] In the primary analysis, sPTB cases were defined as subjects who delivered due to preterm rupture of membranes before term (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 the blood collection interval (0 / 7-25 6 / 7 weeks GA) (Supplementary Materials). Discovery and validation identified two proteins (IBP4 and SHBG) used in the optimal narrow GA of the blood collection interval (19 0 / 7-21 6 / 7 weeks) and the ratio (IBP4 / SHBG) as the best predictor of sPTB by AUROC (Supplementary Materials). Discovery and validation showed that subjects with less extreme BMI values ​​had improved classification performance by IBP4 / SHBG (Supplementary Results). After discovery and validation analyses, we moved to analytical and clinical validation.

[0314] Validation sPTB cases were a total of 18 subjects collected at blood draw time 19 0 / 7–21 6 / 7 weeks GA (GABD) from a total of 81 subjects who were available across 17 0 / 7–28 6 / 7 weeks GA. 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 , 2015;Matei A, Tille Y., The R "sampling" package., European Conference on Quality in Survey Statistics., 2006) and chi-squared tests were used to evaluate 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;63(09):1-86 The distribution of births at term was compared to that outlined in the 2010 study (p. 10). Randomly generated control sets (10 in each group) were examined for sets yielding 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; vol. 21(20): p. 7881). overall To control for the large multiple testing error rate (α = 0.05), the 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 of GABD was 0.5–21 weeks GA) with and without BMI stratification. 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, 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 particular 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 clinical subjects (24), pooled serum standards (3) from healthy non-pregnant donors (HGS), pooled serum standards (3) from healthy pregnant donors (pHGS), and phosphate-buffered saline (2) that served as processing controls. For all analyses, serum samples were first depleted of high abundance and 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 before they were desalted and analyzed by reversed-phase liquid chromatography (LC) / MRM-MS. SIS peptides were used for normalization by generating a response ratio (RR) of the peak area of ​​a peptide fragment ion (i.e., transition) measured in serum divided 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 study subjects in the PAPR. 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 had one or more prior PTBs and to have experienced bleeding after 12 weeks of gestation in the study pregnancy compared to 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 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, by 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 method. Optimal performance was identified for the GA interval of 19 1 / 7 to 20 6 / 7 weeks. When BMI was not taken into account, the validated performance was AUROC=0.67 (p=0.02) (Supplementary Results). However, as expected, the GA intervals of >22 and ≤37 kg / m 2Performance improved with BMI stratification. 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 vs. control boundaries (Table 2). For sPTB vs. 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] Prevalence-adjusted positive predictive value (PPV), a measure of clinical risk, is shown as a function of predictor score in Figure 25. PPV was 0.01% higher than the background rate (population sPTB rate for singleton births in the United States 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 birth is shown in the box plots in FIG. 25. The earliest sPTB cases (<35 0 / 7 weeks GA) had higher predictor scores than the late term controls (≧39 0 / 7 weeks GA), whereas the late sPTB cases (≧35 0 / 7 to <37 0 / 7 weeks GA) had higher predictor scores than the early term controls (≧37 0 / 7 to <39 0 / 7 weeks GA) (Figure 25). Subjects were identified as high or low risk according to a predictor score cutoff corresponding to 2× relative risk (PPV of 14.6%). Live birth rates in the high 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 corroborated controls 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, which identified 75% of women destined for sPTB. 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) are currently considered the best indicators of clinical risk, but either individually or in combination they fail to predict the majority of sPTB.

[0328] An ideal sPTB predictive tool would be minimally invasive, performed early in pregnancy to coincide with the timing of routine obstetric visits, and accurately identify those at highest risk. This omics study suggests that perturbations in the physiological state of pregnancy can be detected with maternal serum analytes measured in sPTB subjects. Omics discovery studies of PTB include proteomic approaches (Gravett et al., Proteomic analysis of cervical-vaginal fluid: identification of novel biomarkers for detection of intra-amniotic infection. , J Proteome Res. 2007;6(1):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;88(2):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), Genomics (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 candidates genes that predispose to spontaneous preterm labor with intact membranes., Am J Obstet Gynecol., 2010; vol. 202 (issue 5): 431.e1~34 ; Swaggart et al., Genomics of preterm birth., Cold Spring Harb Perspect Med. , 2015;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 yielded a validated test to reliably predict the risk of sPTB in asymptomatic women.

[0329] The present 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. Also, the study size and the relatively wide blood collection window (17 The use of a predictor model with low complexity (i.e., ratio of two proteins) allowed for the identification of GA intervals where there was a significant change in protein concentration between sPTB cases and term controls. The use of a predictor model with low complexity (i.e., ratio of two proteins) suppressed 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 difficulty of building a classifier for conditions resulting from multiple etiologies, this predictor performed well with a cutoff of <37 0 / 7 vs. ≥37 0 / 7 weeks GA, with an AUROC of 0.75. Importantly, the accuracy of the predictor improved with earlier sPTB (e.g., <35 0 / 7 weeks GA), allowing detection of sPTB with the highest potential for morbidity. Subjects determined to be 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 in the majority of our study centers and was available in less than one-third of study subjects. It would be interesting to assess in future studies whether CL measurements improve on proteomic predictors or, instead, whether risk stratification with an IBP4 / SHBG classifier identifies women who would benefit most from serial CL measurements. Finally, it would be of great interest to investigate the performance of molecular predictors together with BMI variables or possibly in combination with other medical / pregnancy history and sociodemographic characteristics.

[0332] In conclusion, a predefined predictive test for sPTB based on serum measurements of IBP4 and SHBG in asymptomatic parous and nulliparous women was validated in a completely independent set of subjects. Further functional studies on these proteins, their genetic regulation, and associated pathways can help elucidate the molecular and physiological basis of sPTB. Application of this predictor should allow 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 the prevention of PTB.

[0333] Supplementary Materials and Methods

[0334] Discoveries and Supports

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

[0336] Discovery and supporting principles

[0337] sPTB was defined as described above in this example. Discovery and validation of predictors 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: 1-355). Fully independent sample sets were used for nested case / control analysis. 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 sPTB cases and controls for discovery and corroboration were individually adjudicated by the lead investigator, and discrepancies were resolved by the clinical site PI. A detailed analysis protocol was pre-established, including study design, analysis plan, and corroboration blinding protocol. Laboratory and data analysis personnel were blinded to the corroboration case, control, and GA data allocation. Predictor scores, calculated as described below, were assigned to all subject samples by an internal blinded statistician. 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 (Microsoft, Microsoft Excel 2013) function in Excel 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 confirmation study design

[0339] Discovery and corroboration sPTB cases totaled 86 and 50 subjects, respectively, collected over 17 0 / 7 to 28 6 / 7 weeks GA at blood draw (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 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., 2006, a set of 187 subjects was randomly selected from the 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 p-value (close to 1.0, with a minimum acceptable value of 0.950) was then selected and compared with the BMI distribution from the PAPR study as a whole. Univariate statistical analysis (chi-square test) on BMI data from the PAPR research database was used to select the set of controls that most closely approximated the distribution of BMI (close to 1.0, minimum acceptable value 0.950) and timing of delivery 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. The confidence interval for the PPV was calculated by a normal approximation of the error of the binomial proportion. Brown et al., Interval estimation for a binomial proportion 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 samples of interest. A mature MRM-MS assay measuring 147 proteins was applied to discovery and supportive 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] We developed a strategy 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 a "reversal" is similar to the 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 amplification and self-normalization of diagnostic signals because both proteins in the "inversion" underwent the same preanalytical and analytical processing steps. Inversion as a strategy to normalize peptide intensity measurements in complex proteomics workflows is also similar to a recently introduced approach 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. Predictor scores were defined as the natural logarithm of a single inversion value, which was itself a response ratio (defined above in this example). Finally, predictive performance was investigated in narrow overlapping 3-week gestation 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 of predictor AUROCs in the discovery and validation combination sets were calculated using bootstrap sampling, which was 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] Discovery, corroboration, and validation subject characteristics are summarized in Table 3. The percentage of subjects with one or more prior sPTB histories in discovery sPTB cases was higher than in corroboration or validation. Other characteristics were generally consistent across studies.

[0350] Findings and supporting analysis

[0351] Forty-four proteins were either up- or downregulated in the overlapping 3-week GA intervals and passed the analytical filters (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 of the overlapping 3-week GA intervals. The performance of a subset of inversions showing representative patterns is shown in Figure 29. Waves in performance were evident: IBP4 / SHBG and APOH / SHBG inversions had better AUROC values ​​in the early window, whereas ITIH4 / BGH3 and PSG2 / BGH3 peaked in later 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 AUROC=0.74 in the interval from 19 0 / 7 to 21 6 / 7 (Figure 29). Subjects were classified as having a pre-pregnancy BMI < 35 (kg / m 2When stratified by age, 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). The 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 yielded an AUROC of 0.75 (Table 5).

[0355] Example 3 Correlation of mass spectrometry and immunoassay data This example shows (1) results from Myriad RBM screening to identify IBP4 and other individual biomarkers of sPTB in early, mid, and late pregnancy collection windows; (2) 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 of PAPR were assayed for RBM (20 / 20 from early window, 10 / 10 from mid window, 10 / 10 from late window). For RBM, Human Discovery MAP 250+ v2.0 (Myriad RBM, Austin, TX) was used. The goal of these analyses is to develop multivariate models for predicting PTB using multiple analytes. We used four modeling methods: random forest (rf), boosting, lasso, and logistic (logit). We perform a first round of variable selection where each method independently selects the 15 best variables of that method. From these 15, the best analyte was independently selected by each of the four modeling methods using variable backward selection and estimation of the area under the ROC curve (AUC) using out-of-bag bootstrap samples. Table 6 shows the top hits from several multivariate models. Table 7 shows the early window (GABD 17-22 weeks) analyte rankings by various multivariate models. Table 8 shows the mid-window (GABD 23-25 ​​weeks) analyte rankings by various multivariate models. Table 9 shows the late window (GABD 26-28 weeks) analyte rankings by various multivariate models.

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

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

[0360] One hundred and twenty previously frozen serum samples from PAPR studies with known outcomes were selected to compare the ELISA and MS assays. The samples had a gestational age at blood collection (GABD) of 119-180 days. Samples were not excluded for maternal BMI. ELISAs were performed with commercial 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 protocols. 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, where 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 outcome were selected to compare the ELISA and MS assays. The samples had a gestational age at blood collection (GABD) of 133-148 days. ELISAs were performed with commercial 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 protocols. An internal standard was used to normalize samples run on different plates. Scores were calculated as ELISA concentration values ​​by LN ([IBP4] / [SHBG]) and LN (IBP4 RR / SHBG RR ) from MS, 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 to 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 a predicted gestational age at blood collection (GABD) of 133-146 days. Correlation analyses were performed on all BMI samples (Figure 34, right panel) or on a subset of samples with BMI >22 or ≤37 (Figure 34, left panel). ELISAs were performed with commercial 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, where 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 separations (p-values ​​were derived with unpaired t-test assuming equal standard deviations). Table 12 shows the IBP4 and SHBG ELISA kits showing sPTB vs. control separations (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 at 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 began at all three laboratories on the same day. ARUP used a Roche cobas e602 analyzer and Intermountain used an Abbott Architect CMIA. Both were semi-automated immunoassay instruments. Sera Prognostics used a proprietary proteomic analysis method that included immunodepletion of samples, enzymatic digestion, 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 r value is 0.937, which is different from Intermountain. 2 The value was 0.934.

[0365] Example 4 SNPs, insertions and deletions, and structural variants within the 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 in PreTRM IBP4 and SHBG peptides. This information is derived from the database of single nucleotide polymorphisms (dbSNP) build 146. A single missense mutation (G>C) A179P (dbSNP id:rs115336700) in SHBG has the highest overall allele frequency of 0.0048. Although this allele frequency is low, some 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 count), heterozygosity-average heterozygosity, validation-validation method (or blank for no validation), MAF-minor allele frequency, function-functional feature of the polymorphism, dbSNP allele-allele 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 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 by MS separately over the indicated gestational age ranges for sPTB cases and full-term controls are shown (Figure 44 and Figure 45). Curves were generated by mean smoothing the peptide relative ratios (endogenous peptide peak areas to corresponding SIS peak areas). Case vs. control signals correspond to approximately a maximum 10% difference for IBP4 and SHBG. When the scores calculated as ln(IBP4RR / SHBGRR) are plotted, the signal amplification is evident (approximately a maximum 20% difference) (Figure 46). These data demonstrate the diagnostic signal amplification obtained using the IBP4 / SHBG reversal strategy.

[0369] Since each protein is subjected to the same analytical and pre-analytical processing steps, forming a ratio of the levels of the two proteins can reduce variability. To study the impact on variability, the CVs of the individual proteins (RR of IBP4 and SHBG) and the CV of the IBP4 RR / SHBG RR ratio in pooled control serum samples (pHGS) from pregnant donors were determined. Pooled control samples with no biological variability were analyzed in multiple batches and 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 is common to amplify 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 a dataset derived from samples collected at 19 / 0-21 / 6 weeks gestation is shown in the top panel of Figure 47. 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 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 the 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 without biological variability were analyzed in multiple batches and over several days. The inversion variability is less than the variability associated with the individual proteins (Figure 49).

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

[0373] In Figure 48, the calculated CVs for pHGS samples (pooled pregnancy samples) analyzed in a laboratory on several instruments, over several days, in several batches are reported. The CVs were calculated using pHGS samples that lack biological variability and therefore represent a measure of the analytical variability introduced in the 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 PTB that are medically indicated based on conditions such as pre-eclampsia or gestational diabetes.

[0375] 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). 41 medically indicated PTB samples from the PAPR Biobank were analyzed in the laboratory and the PreTRM score was calculated. PreTRM™ scores were compared between subjects annotated as medically indicated for preeclampsia and subjects annotated for other procedures. Subjects who were medically indicated for preterm delivery due to preeclampsia had significantly higher scores than others (Figure 50).

[0376] In Figure 52, a heatmap of inversion intensity for diabetes annotation is shown. The 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 shows that inversions that stratify gestational diabetes can be identified, and therefore it is possible to build diagnostic tests 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) illustrate 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 illustrated, 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 proteins were analyzed by MRM-MS to identify surrogate peptide candidates 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 the SHBG peptide MS data and transition MS data. SHBG tryptic peptides derived from recombinant proteins and pooled pregnancy serum were analyzed by MRM-MS to identify surrogate peptide candidates 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. Serum samples from 312 women (104 sPTB cases, 208 term controls) at gestational age (GA) 17-25 weeks were analyzed by LC-MS (MRM) assays of 148 proteins from multiple pathways. MRM peak area data were analyzed by hierarchical clustering, t-tests, and association with GA. After analytical filtering, 25 proteins showed significant differences (p<0.05) in sPTB versus 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 in subintervals of the GA period. For example, at GA 19-21 weeks, seven additional proteins were elevated in sPTB and one was lower.

[0381] Table 20 lists 44 proteins that met the analytical filters 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 shows 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] Pathways such as inflammation, infection, and hemorrhage have been implicated in the pathogenesis of preterm birth. However, little is known about which proteins are measurable in blood and at what time during pregnancy they are disrupted. To answer these questions, we generated LC-MS (MRM) assays of 148 proteins from multiple pathways and analyzed serum samples from 312 women (104 sPTB cases, 208 full-term controls) at 17-25 weeks 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 for nearly all proteins. SIS peptides were used for normalization by generating response ratios of the peak areas of peptide fragment ions (i.e., transitions) measured in serum divided by the peak areas of the corresponding SIS transitions. Response ratios of 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 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-b activity. NCAM1 (CD56) is highly expressed on neuronal cells 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., vol. 114:744-754. Reversal of 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 (TGFBI) is a cell adhesion molecule expressed on vascular endothelial cells that inhibits angiogenesis by 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 in 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 and releases active IGF2. Low serum PAPP1 levels in 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 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 nodes is consistent with insufficient IGF2 activity in sPTB cases, which may result in abnormal placentation. Figure 56A shows a schematic 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 phosphorylation of IRS1 / 2 and Shc proteins. Shc activation leads to recruitment of the Grb2 / Sos complex, with subsequent activation of 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) as well as protein synthesis and cell proliferation (mTOR).

[0391] Figure 57 shows kinetic plots of differentially expressed proteins with functions related to metabolic hormone balance. Sex hormone binding globulin (SHBG), a placental protein, increases during pregnancy and determines the bioavailability and metabolism of sex steroid hormones. Decreased SHBG levels lead to higher free androgen and estrogen levels. Free androgens may be converted to estrogen by placental aromatase activity. The anti-progesterone 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 increased risk of miscarriage and premature birth. Stagnaro-Green A. and Pearce E., 2012, Nat. Rev. Endocrinol. , Vol. 8(No. 11): pp. 650-8. 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 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 antiangiogenic factor expressed in the ovarian gamma-secretase that stimulates the cleavage and inactivation of VEGFR-1. 10 Cathepsin D (CATD) cleaves prolactin to generate vasoinhibin, which inhibits angiogenesis. Elevated serum CATD and vasoinhibin 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( 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, 151(1):369-379.

[0393] Figure 59 shows the kinetic plots of differentially expressed proteins with functions related to innate immunity. LBP is bacterial LPS to Toll-like receptor-4 that induces inflammatory responses of 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 Fetuin-A-FA complex can bind to and activate the TLR4 receptor. Pal, D. et al., 2012, Nature Med., 18(8):1279-85.

[0394] FIG. 60 shows kinetic plots of differentially expressed proteins with functions in coagulation.

[0395] In FIG. 61, kinetic plots of differentially expressed serum / secreted proteins are shown.

[0396] In FIG. 62, kinetic plots of differentially expressed PSG / IBP are shown.

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

[0398] In FIG. 64, kinetic plots of differentially expressed complement / acute phase protein-1 are shown.

[0399] In FIG. 65, kinetic plots of differentially expressed complement / acute phase protein-2 are shown.

[0400] In FIG. 66, kinetic plots of differentially expressed complement / acute phase protein-3 are shown.

[0401] In FIG. 67, kinetic plots of differentially expressed complement / acute phase protein-4 are shown.

[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 Week 17, Day 0 to Week 28, Day 6.

[0403] In Figures 68-85, the mean relative ratio of each peptide transition was plotted against GABD using the R ggplot2 package, using a mean smoothing function (window = + / - 10 days). Graphs feature separate plots of cases vs. 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, underline, and peptide sequence. Analyte sequences may be abbreviated to allow titles to fit the size of the plot.

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

[0405] Analytes found to show significant case vs. control separation in the early window, e.g., with samples collected between 18-20 weeks of gestation, include, e.g., 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 vs. control separation in the later window, e.g., with samples collected between 26-28 weeks of gestation, include, e.g., ITIH4, HEP2, IBP3, IGF2, KNG1, PSG11, PZP, VASN, and VTDB. Separation of cases versus controls was improved using cutoffs of <35 0 / 7 weeks versus >35 0 / 7 weeks versus <37 0 / 7 weeks versus >37 0 / 7 weeks, 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 birth cutoffs. Those skilled in the art will appreciate that any analyte showing significant separation between cases and controls as shown in the accompanying figures for a given time window 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 vs. 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 date pregnancy without or with other date determination methods (e.g., date of last menstrual period, ultrasound date determination). 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 (2 per case), and two new controls. This analysis described in this example extends the commercial blood collection window beyond 19 and 20 weeks, generates additional data on prediction of sPTB <35 weeks based on multiple samples derived from all previous cases, led to the discovery of new analytes and reversals, defined molecular clock proteins, elucidated risk thresholds, and formed 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 Risk (PAPR) clinical study. The 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, as well as relevant information on patient demographic characteristics, past medical and reproductive history, current pregnancy history, and concomitant medications. After delivery, data on maternal and infant status and complications were collected. Serum and plasma samples were processed according to a protocol calling for standardized refrigerated centrifugation, aliquoting of samples into 0.5 ml 2D barcoded cryovials, and subsequent freezing at -80°C.

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

[0411] Serum samples were subsequently depleted of high abundance 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. Filtered samples were depleted using a MARS-14 column (4.6 x 100 mm, Agilent Technologies). Samples were cooled to 4 °C in an autosampler, the depletion column was run at room temperature, and 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 approximating those of the surrogate peptide analytes. The SIS fortified samples were mixed and divided into two equal volumes. Each aliquot was placed in a -80°C storage container until the workup was ready to continue. 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). The eluted peptides were dried by lyophilization. The lyophilized samples were reconstituted in a reconstitution solution containing only the internal standard that monitors the quality of the LC-MS step (IS Recon).

[0413] The fully processed samples were analyzed using dynamic multiple reaction monitoring (dMRM). Peptides were separated on a 2.1×100 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 that served both diagnostic and qualitative purposes. Chromatographic peaks were integrated using MassHunter quantitative analysis software (Agilent Technologies). The ratios of the chromatographic peak areas of the surrogate peptide analytes to the corresponding SIS chromatographic peak areas were reported.

[0414] A summary of serum analyte proteins, peptides, and transitions, SIS transitions, and IS Recon standards measured with the dMRM method are shown in Table 21. MARS-14 depleted proteins identify 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 since this peptide is conserved in several HLA type I isotypes. LYAM3 is referred to here as LYAM1 since 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 the 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 (e.g., 119-132 refers to gestational days 119-132, which is equivalent to 17 and 18 weeks of gestation) overlapping by one week. Performance in each two week window is reported for 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. 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 equates to 17, 18, and 19 weeks of gestation) overlapping by 2 weeks. Performance in each 3-week window was evaluated using two different case vs. control cutoffs (<37 0 / 7 vs. >=37 0 / 7, <35 0 / 7 vs. >=37 0 / 7, <35 0 / 7 vs. >=37 0 / 7, <= ... 0 / 7 vs. >=35 0 / 7) and with (rBMI) and without (aBMI) BMI stratification are reported.

[0417] Figures 86-95 show kinetic plots of various peptide transitions for cases vs. 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 vs. 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 an average smoothing function (window = + / - 10 days). The graph features separate plots of cases vs. 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). The plot title shows the protein abbreviation, underline, and peptide sequence. Analyte sequences may be abbreviated to allow the title to fit the size of 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-regulated or down-regulated for each of the combinations (i.e., overlapping 2-week or 3-week windows, BMI limits, and SPTB definitions), and if the majority of combinations were marked as up-regulated, the analyte was called 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 of the up-regulated analyte relative ratio values ​​by the down-regulated analyte relative ratio values ​​and taking the natural logarithm of the result. This results in 1650 inversions (55 x 30 = 1650). For each inversion, the area under the ROC curve (AUCROC) indicating SPTB and TERM separation, and a p-value indicating whether the AUCROC value was significantly different from AUCROC = 0.5 (i.e., SPTB and TERM separation is not significant) were calculated. For inversions with AUCROC > 0.6 and p-value < 0.05, the performance of each inversion in different conditions (e.g., with and without pregnancy window, BMI restriction, and two sPTB cutoffs) are tabulated. Tables 27-42 show the inversion classification performance at 17 and 18 weeks of gestation. Tables 47-58 show reversal classification performance for 17, 18, and 19 weeks of gestation. Tables 43-46 show reversal classification performance for 17-21 weeks of gestation. Additional potentially significant reversals are shown in Table 59.

[0420] Also, improved performance of predictors formed from more than one reversal (17-21 weeks) was demonstrated. Briefly, reversals 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 reversals (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 reversals (SumLog), but one of skill in the art could select other models (e.g., logistic regression). It is also contemplated that this multiple reversal approach will be applied to combinations of reversals specific to preterm rupture of membranes (PPROM) versus preterm labor without PPROM (PTL), fetal sex, and gestational age. It is further contemplated that the predictor can include indicator variables that select the subset of reversals to be used if blood draw period, fetal sex, or gestational age is known.

[0421] Figure 110 shows the relationship between predictor score (ln IBP4 / SHBG) and prevalence-adjusted relative risk (positive predictive value) of sPTB using cutoffs of <37 0 / 7 gestational weeks vs. >=37 0 / 7 gestational weeks. Samples were taken from 19 1 / 7 weeks to 20 6 / 7 weeks with BMI >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 bars 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) of sPTB using cutoffs of <35 0 / 7 gestational weeks vs. >=35 0 / 7 gestational weeks. Samples were taken from 19 1 / 7 weeks to 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 bar graphs 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 different phenotypes: sPTB, PPROM, and PTL. PPROM occurs more often early and is associated with infection or inflammation. PTL may occur later and is generally considered to be a less severe phenotype. There were more significant reversals, and the performance was higher for PPROM, which are dominated by proteins known to be involved in inflammation and infection. It is contemplated to select reversals to build independent testing methods for PPROM and PTL, or to use a combination of more than one reversal in a single predictor to maximize performance overall. The analysis shown in Tables 61-64 required AUC>0.65 and p<0.05 for either PPROM or PTL.

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

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

[0427] Clinical Observations: Primiparas and multiparas

[0428] Inversion performance (17-21 weeks) was further evaluated independently for two different phenotypes of sPTB, primiparas and multiparous women. In Tables 65-68, the top performing inversions (17-21 weeks) are shown separately for primiparas (first-time mothers) and multiparous subjects. First-time mothers are the ones most in need of testing to predict the probability of PTB, as doctors do not have a pregnancy history to determine / estimate the risk. These results allow the test to be independent for these two groups, or to combine the high performing inversions in a single classifier to predict both risks. The analyses shown in Tables 65-68 required AUC>0.65 and p<0.05 for either primiparas 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 case-control cutoffs 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 case-control cutoffs 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 case-control cutoffs 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 case-to-control cutoffs 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. Some inversions were found to have fetal sex-specific predictive performance. Figure 106 shows the fetal sex-specific IBP4 and SHBG analytes and scores (IBP4 / SHBG). IBP4 is significantly higher in subjects carrying male fetuses. Inversion performance remains comparable for gestational ages 19-21 weeks without BMI stratification (Figure 106). Additionally, PAPR found that investigational male fetuses were at increased risk for sPTB, with a p-value of 0.0002 and an odds ratio of 1.6. That is, fetal sex can be incorporated into the predictors (e.g., inversion value + fetal sex). The analyses shown in Tables 69-72 required AUC>0.65 and p<0.05 for either male or female fetuses.

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

[0433] Example 11 Correlation of mass spectrometry and 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 (only available 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 multiple coating solutions

[0439] Each U-PLEX coupled antibody (600 μL) solution was combined in a single tube and vortexed to mix. When less than 10 antibodies were combined, the solution volume was increased to 6 mL with stop solution to a final 1× concentration. Note that in these experiments, there was only a single antibody 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 for 1 h at room temperature or overnight at 2-8 °C 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 for 1 hour at room temperature with shaking at approximately 700 rpm. The plate was then washed with at least 150 μL of 1× MSD wash buffer. *The plates were washed three times with 1× MSD wash buffer. 50 μL of detection antibody solution was added to each well. After sealing, the plates were incubated for 1 hour at room temperature 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 pair combinations in both capture detector orientations. Capture antibodies were prepared at 10 μg / mL, coupled to U-PLEX linkers, and coated onto U-PLEX plates. SHBG R&D Systems calibrators were diluted in diluent 43 to generate a 7-point standard curve along with assay diluent blanks. Samples were diluted in diluent 43 and tested in the assay as follows: serum SHBG "high" and "low" samples: 100-fold and 500-fold dilutions, and serum pregnancy pool: 100-fold, 200-fold, 400-fold, 800-fold dilutions. Detection antibodies were tested at 1 μg / mL in diluent 3. Binding to the standard curve and native analytes in serum was evaluated. The top analyte pairs were then tested with NIBSC and Biospacific calibrators diluted as 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 7-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. Detection antibodies were tested at 1 μg / mL in diluent 12. Standard curves and binding to native analytes in serum were evaluated.

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

[0449] Antibody pair 12 was 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 detection) 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 a calibrator that would work with all pairs. However, all antibody pairs recognized native analytes in serum high samples, serum low samples, and pregnancy pool samples. 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 difference 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 varies 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 3-fold higher signals with the NIBSC standard.

[0454] IGFBP-4 antibody screening

[0455] Antibody pair 2 standard curves 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 dilutions of 32 and 64. Pair 2 showed approximately 12 fold higher signal in the samples, resulting in a 2-4 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 1000-fold and samples fell between standard calibrators 1-3. The median measured concentration was 58.4 μg / mL. The CV for duplicate measurements was low with a median CV of 2.4%. The median measured concentration for IGFBP-4 was 234 ng / mL. The median CV between duplicate samples 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, such embodiments also falling 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 and 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 16-5

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 18-24

Table 18-25

Table 18-26

Table 18-27

Table 18-28

Table 18-29

Table 18-30

Table 18-31

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 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 25-4

Table 26-1

Table 26-2

Table 27-1

Table 27-2

Table 27-3

Table 27-4

Table 28-1

Table 28-2

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 32-1

Table 32-2

Table 32-3

Table 32-4

Table 32-5

Table 33-1

Table 33-2

Table 34-1

Table 34-2

Table 35-1

Table 35-2

Table 35-3

Table 35-4

Table 36-1

Table 36-2

Table 36-3

Table 37-1

Table 37-2

Table 37-3

Table 37-4

Table 38-1

Table 38-2

Table 38-3

Table 39-1

Table 39-2

Table 39-3

Table 39-4

Table 39-5

Table 39-6

Table 39-7

Table 40-1

Table 40-2

Table 40-3

Table 40-4

Table 40-5

Table 40-6

Table 41-1

Table 41-2

Table 41-3

Table 42-1

Table 42-2

Table 42-3

Table 42-4

Table 43-1

Table 43-2

Table 43-3

Table 43-4

Table 44-1

Table 44-2

Table 44-3

Table 44-4

Table 45-1

Table 45-2

Table 45-3

Table 45-4

Table 46-1

Table 46-2

Table 46-3

Table 46-4

Table 47-1

Table 47-2

Table 47-3

Table 48-1

Table 48-2

Table 48-3

Table 49-1

Table 49-2

Table 49-3

Table 50-1

Table 50-2

Table 51-1

Table 51-2

Table 51-3

Table 51-4

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 53-1

Table 53-2

Table 53-3

Table 53-4

Table 54-1

Table 54-2

Table 54-3

Table 55-1

Table 55-2

Table 55-3

Table 55-4

Table 55-5

Table 56-1

Table 56-2

Table 56-3

Table 56-4

Table 56-5

Table 56-6

Table 57-1

Table 57-2

Table 57-3

Table 57-4

Table 58-1

Table 58-2

Table 58-3

Table 59-1

Table 59-2

Table 59-3

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 66-5

Table 67-1

Table 67-2

Table 67-3

Table 67-4

Table 67-5

Table 67-6

Table 68-1

Table 68-2

Table 68-3

Table 68-4

Table 68-5

Table 69-1

Table 69-2

Table 69-3

Table 70-1

Table 70-2

Table 70-3

Table 70-4

Table 71-1

Table 71-2

Table 71-3

Table 71-4

Table 71-5

Table 72-1

Table 72-2

Table 72-3

Table 72-4

Table 73

Table 74

Table 75

Table 76

Table 77-1

Table 77-2

Claims

[Claim 1] The method described in the specification.

Images