Non-invasive diagnostic biomarker-based test of endometriosis

Abstract

The present invention relates to an in vitro method for assessing whether a subject has endometriosis comprising a step of measuring, in a biological sample of the subject, the expression level of at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP1.

Description

NON-INVASIVE DIAGNOSTIC BIOMARKER-BASED TEST OFENDOMETRIOSISFIELD OF INVENTION

[0001] The present invention relates to an in vitro non-invasive diagnostic biomarkerbased test of endometriosis and its application.BACKGROUND OF INVENTION

[0002] Endometriosis is defined as the presence of endometrial-like tissue outside the uterine cavity. This chronical disease is considered a debilitating gynecological pathology with a high incidence, between 6-10% among women of childbearing age. The prevalence of endometriosis has been estimated at 180 million women worldwide. Endometriosis lesions can occur at different locations in the body, including the pelvic peritoneum or ovary or can infiltrate pelvic structures below the peritoneal surface, thus defining three clinical phenotypes, superficial endometriosis (SUP), endometrioma (OMA) and deep endometriosis (DIE). The main symptoms consist of chronic pelvic pain, dysmenorrhea, dyspareunia, and infertility. Given the nonspecific symptoms of this disease, the gold standard for a definitive diagnosis is based on surgical procedure such as laparoscopy to visualize the lesions followed by a histological confirmation. The average delay between the onset of symptoms and diagnosis is 7 years. The lack of specific and heterogenous symptoms associated with a stigma or symptom normalization, the lack of awareness of the pathology by the general public, and by many practitioners and some gynecologists, are in part responsible for the diagnostic delay.

[0003] The management of the patient is evolving, and no systematic surgery and excision of the lesions are nowadays recommended. In fact, according to several gynecological scientific societies, there is a consensus to consider the patient’s predominant symptoms and / or the desire for a child before considering invasive treatment. Available medical therapies for endometriosis include non-hormonaltreatments such as painkillers and hormonal treatments such as combined oral contraceptives, progestins and gonadotropin-releasing hormone analogues.

[0004] Medical imaging has led to substantial improvements in the diagnosis of endometriosis. Transvaginal ultrasonography (TVUS) and / or magnetic resonance imaging (MRI) can be used to establish a diagnosis of OMA and DIE. However, SUP can usually not be visualized by imaging since the size and location of the lesions is below the threshold for detection. In addition, few radiologists are aware of the specific imaging modalities for the disease and patients who are suspected of endometriosis should be referred to a specialist of this disease.

[0005] In a clinical practice, a noninvasive diagnostic biomarker-based test could help shorten the time during which the patient does not know what she is suffering from and is not taken care of. The identification of validated biomarkers in blood is an unmet medical need that demands to be fulfilled. Many publications and patents are available describing the use of biomarkers alone (miRNA, protein, metabolites, RNA, circulating free DNA, antibodies, etc.) and / or in combination with various clinical performances.

[0006] W02021 / 168040 discloses compositions and methods that provide a high degree of sensitivity and a high degree of specificity for the pre-operative assessment of endometriosis in pre-menopausal women having a variety of endometriosis types. W02021 / 168040 discloses a large panel of more than 50 polypeptide biomarkers that are allegedly differentially present in subjects having endometriosis.

[0007] However, there is still a need for improved noninvasive diagnostic methods that not only have a high degree of sensitivity, but that also provide a high degree of specificity for diagnosing endometriosis and allowing a further adapted treatment, based on a small number of biomarkers to be measured, such as, for example, based on as few as five biomarkers.SUMMARY

[0008] This invention thus relates to an in vitro method for assessing whether a subject has endometriosis comprising a step of measuring, in a biological sample of the subject, the expression level of at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP 1.

[0009] In some embodiments, the method comprises measuring the expression level of at least three biomarkers, wherein the at least three biomarkers are:• MUC-16, and• ANXA1 or ARHGAP 1 , and• at least one biomarker selected from the group comprising or consisting of EREG, ANXA1, ARHGAP1, t-PA, PCSK9, ARSB, ESM-1 and TMPRSS15.

[0010] In some embodiments, the method comprises measuring the expression levels of three biomarkers, wherein the three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-PA.

[0011] In some embodiments, the method comprises measuring the expression levels of at least three biomarkers, wherein the at least three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-PA.

[0012] In some embodiments, the method comprises measuring the expression levels of five biomarkers, wherein the five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0013] In some embodiments, the method comprises measuring the expression levels of at least five biomarkers, wherein the at least five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0014] In some embodiments, the method comprises determining a score based on a linear combination of the measured expression levels (i.e., a combination with a linear model).

[0015] In some embodiments, the method comprises determining a score based on a nonlinear combination of the measured expression levels using a complex model (i.e.; a complex, non-linear model). In some embodiments, the score is determined using a machine learning algorithm selected from the group comprising or consisting of an artificial neural network (ANN), a perceptron algorithm, a deep neural network, a clustering algorithm, a k-nearest neighbors algorithm (k-NN), a decision tree algorithm, a random forest algorithm, a linear regression algorithm, a logistic regression algorithm, a linear discriminant analysis (LDA) algorithm, a quadratic discriminant analysis (QDA) algorithm, a support vector machine (SVM), a Bayes algorithm, a simple rule algorithm, a clustering algorithm, a meta-classifier algorithm, a Gaussian mixture model (GMM) algorithm, a nearest centroid algorithm, a gradient boosting algorithm (such as, e.g., an extreme gradient boosting [XG Boost] algorithm or an adaptative boosting [AdaBoost] algorithm), a linear mixed effects model algorithm, and a combination thereof.

[0016] In some embodiments, the biological sample is selected from the group comprising or consisting of any tissue, cell, fluid, saliva, tears, urine, sweat, sputum, liquid biopsy, blood, serum, plasma, ascites, cyst fluid, vaginal fluid and cervico-vaginal fluid. Preferably, the biological sample is selected from the group comprising or consisting of blood, serum and plasma.

[0017] In some embodiments, the level of the biomarkers is measured by immunoassays, such as, for example, ELISA, Western blot, immune-electrophoresis, immunostaining orProximity Extension Assay (PEA); enzymatic activity; flow cytometry; spectrometry, including mass spectrometry; RT-PCR; RT-qPCR; Northern Blot; or hybridization techniques.

[0018] This invention also relates to a diagnostic device, comprising means for measuring the expression level of biomarkers of a signature comprising at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP 1.

[0019] In some embodiments, the diagnostic device comprises means for measuring the expression level of biomarkers of the signature consisting of three biomarkers, wherein the three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0020] In some embodiments, the diagnostic device comprises means for measuring the expression level of biomarkers of the signature comprising at least three biomarkers, wherein the at least three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0021] In some embodiments, the diagnostic device comprises means for measuring the expression level of biomarkers of the signature consisting of five biomarkers, wherein the five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0022] In some embodiments, the diagnostic device comprises means for measuring the expression level of biomarkers of the signature comprising at least five biomarkers, wherein the at least five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0023] In some embodiments, the diagnostic device comprises i) capture reagents specific for the biomarkers of the signature, and ii) detection reagents for detecting the biomarkers of the signature.

[0024] In some embodiments, the capture reagents are bound to plates, chips, beads, membranes or arrays.

[0025] In some embodiments, the detection reagents are enzymes, radioactive proteins, or fluorescent proteins.

[0026] In some embodiments, the means for measuring the expression level of biomarkers of a signature comprise or consist of means that may be used in immunoassays, such as, for example, ELISA, Western blot, immune-electrophoresis, immunostaining or Proximity Extension Assay (PEA); enzymatic activity; flow cytometry; spectrometry, including mass spectrometry; RT-PCR; RT-qPCR; Northern Blot; or hybridization techniques, preferably immunoassay.DEFINITIONS

[0027] Unless defined otherwise, all technical and scientific terms used herein have the meaning commonly understood by a person skilled in the art to which this invention belongs. For sake of completeness, in the present invention, the following terms have the following meanings:

[0028] “A”, “an”, and “the” include plural forms unless the context clearly dictates otherwise. Thus, for example, reference to “a biomarker” includes reference to more than one biomarker.

[0029] “About” preceding a figure means plus or less 10% of the value of said figure.

[0030] “Biological sample” refers to any tissue such as, for example, a homogenized tissue sample (e.g., a tissue sample obtained by biopsy), cell, fluid, saliva, tears, urine, sweat, sputum, liquid biopsy, blood, serum, plasma, ascites, cyst fluid, vaginal fluid and cervico-vaginal fluid, or any other material derived from a subject or patient.

[0031] “Biomarker” refers to a substance within a biological system that is used as an indicator of endometriosis. According to the present invention, a biomarker may be a protein, a polypeptide, a nucleic acid molecule (e.g., a polynucleotide such as mRNA, cDNA), a metabolite, a hormone, or other analyte that is associated with endometriosis.

[0032] “Capture reagent” refers to a reagent that specifically binds a nucleic acid molecule or polypeptide to select or isolate the nucleic acid molecule or polypeptide.

[0033] “Comprises”, “comprising”, “containing”, “having” and the like mean “includes”, “including” and the like.

[0034] “Datasets” are collections of data used to build a machine learning mathematical model, so as to make data-driven predictions or decisions. In “supervised learning” (i.e., inferring functions from known input-output examples in the form of labelled training data), two types of machine learning datasets are typically dedicated to two respective kinds of tasks: “training”, i.e., fitting the parameters; and “testing”, i.e., checking independently of a training dataset exploited for building a mathematical model that the latter model provides satisfying results.

[0035] “Determining”, “assessing”, “assaying”, “measuring” and “detecting” refer to both quantitative and qualitative determinations, and as such, the term “determining” is used interchangeably herein with “assaying”, “measuring”, and the like.

[0036] “Difference of (the) level” refers to differences in the quantity of a particular biomarker in a sample as compared to a control or reference level. In some embodiments, a “difference of a level” may be a difference between the quantity of a particular biomarker present in a sample as compared to a control of at least about 1%, at least about 2%, at least about 3%, at least about 5%, at least about 10%, at least about 15%, at least about 20%, at least about 25%, at least about 30%, at least about 35%, at least about 40%,at least about 50%, at least about 60%, at least about 75%, at least about 80% or more. In some embodiments, a “difference of a level” may be a statistically significant difference between the quantity of a biomarker present in a sample as compared to a control. For example, a difference may be statistically significant if the measured level of the biomarker falls outside of about 1.0 standard deviations, about 1.5 standard deviations, about 2.0 standard deviations, or about 2.5 standard deviations of the mean of any control or reference group.

[0037] “Diagnosis” or “diagnostic” refer to detecting a disease or disorder or determining the stage or degree of a disease or disorder. Usually, a diagnosis of a disease or disorder is based on the evaluation of one or more factors and / or symptoms that are indicative of the disease. That is, a diagnosis can be made based on the presence, absence or amount of a factor which is indicative of presence or absence of the disease or condition. Each factor or symptom that is considered to be indicative for the diagnosis of a particular disease may not be exclusively related to the particular disease; i.e., there may be differential diagnosis that can be inferred from a diagnostic factor or symptom. Likewise, there may be instances where a factor or symptom that is indicative of a particular disease is present in an individual that does not have the particular disease. The diagnostic methods may be used independently, or in combination with other suitable diagnosing and / or staging methods.

[0038] “Endometriosis” refers to a chronic, hormone-dependent, inflammatory gynecological disorder characterized by the implantation of benign endometrial tissue in locations outside the uterine cavity, including the pelvic peritoneum, ovaries, and bowel. “Endometriosis” also refers to a variety of endometriosis types (e.g., endometriosis, superficial or peritoneal endometriosis, deep infiltrating endometriosis, endometriotic cysts, endometrioma, or another benign condition of the endometrium) and at a variety of disease states (e.g., early and late stage).

[0039] “Expression level” refers to expression of a biomarker, including the encoded polypeptide or protein, or the polynucleotide encoding the biomarker, in particular the mRNA level. Expression level of a biomarker may be determined by measurement of polypeptide or protein level, for example, by immunoassay using one or moreantibody(ies) that bind(s) with the polypeptide or protein. Alternatively, expression of a biomarker may be determined by measurement of a polynucleotide level, in particular of mRNA levels, for example, by RT-PCR, RT-qPCR (wherein qPCR stands for quantitative PCR), or using a microarray, or using sequencing methods.

[0040] “Including”, “include(s)” is used herein to mean, and is used interchangeably with, the phrase “including but not limited to.”

[0041] “Learning algorithm” or “machine learning algorithm”, as used herein, refers to computer-executed algorithms that automate analytical model building, e.g., for clustering, classification or profile recognition. Learning algorithms perform analyses on training datasets provided to the algorithm. Learning algorithms output a “model”, also referred to as a “classifier”, “classification algorithm” or “diagnostic algorithm”. Models receive, as input, test data and produce, as output, an inference or a classification of the input data as belonging to one or another class, cluster group or position on a scale, such as diagnosis, stage, prognosis, disease progression, responsiveness to a drug, etc.

[0042] “Polynucleotide” includes DNA, RNA and DNA / RNA hybrids, in particular cDNA and mRNA.

[0043] “Reference” refers to a standard of comparison. In one embodiment, the reference score is defined as a score allowing determination of the status of the subject or patient (e.g., healthy or affected with endometriosis). In one embodiment, the reference score is determined (or was previously determined) by comparing in a substantially identical manner the expression level(s) of at least one biomarker (or a combination of biomarkers) as described herein in samples of subjects of a substantially healthy population to the expression level(s) of said biomarker(s) in samples of subjects of a diseased population. A reference sample thereby provides a standard allowing for the evaluation of the information obtained from the sample of interest.

[0044] “Score” refers to any digit value obtained by the mathematical combination (univariate or multivariate, using linear or complex models) of at least two biomarkers as defined herein. In one embodiment, a score is an unbound digit value. In anotherembodiment, a score is a bound digit value, obtained by a mathematical function. In some embodiments, a score ranges from 0 to 1.

[0045] “Specifically binds” refers to a compound (e.g., antibody) that recognizes and binds a molecule (e.g., polypeptide), but which does not substantially recognize and bind other molecules in a sample, for example, a biological sample.

[0046] “Signature” of biomarkers refers to a set of biomarkers that may be used for diagnosing endometriosis. In one embodiment, the use of a signature instead of single biomarkers provides an enhancement in diagnosis of endometriosis, as measured by sensitivity, specificity and / or area under the Receiver Operating Characteristic (ROC) curve (AUC) than any single biomarker alone.

[0047] “Subject” refers to a mammal (such as a human, a non-human primate, murine, bovine, equine, canine, ovine, or feline), preferably a human. In one embodiment, a subject may be a “patient”, i.e., an animal, more preferably a human, who / which is awaiting the receipt of, or is receiving medical care or was / is / will be the object of a medical procedure, or is monitored for the development of a disease.

[0048] “Substantially healthy” refers to a subject or to a population of subjects that has / have not been previously diagnosed or identified as having or suffering from endometriosis.

[0049] Unless specifically stated or obvious from context, as used herein, the term “or” is understood to be inclusive.DETAILED DESCRIPTION

[0050] A biomarker refers to a broad subcategory of medical signs which can be measured accurately and reproducibly. Medical signs stand in contrast to medical symptoms, which are limited to those indications of health or illness perceived by subjects themselves. According to the National Institutes of Health Biomarkers Definitions Working Group, a biomarker is “a characteristic that is objectively measured andevaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention". It is also defined as “any substance, structure, or process that can be measured in the body or its products and influence or predict the incidence of outcome or disease" (Strimbu and Tavel, Curr Opin HIV AIDS. 2010 Nov; 5(6): 463-466).

[0051] The invention provides an in vitro method for assessing whether a subject has endometriosis comprising a step of measuring, in a biological sample of the subject, the expression level of at least two biomarkers, named signature, that are differentially expressed in subjects having endometriosis. The biomarkers of said signature are differentially expressed depending on endometriosis status, including, subjects having endometriosis vs. subjects that do not have endometriosis.

[0052] Within the meaning of the invention, a biomarker is differentially present if the mean or median expression level of the biomarker in samples from a group of control subjects (in particular in a group of substantially healthy subjects) is statistically different from the mean or median level of the biomarker in samples from a group of diseased subjects or patients.

[0053] A reference level may be, for example, the level that differentiates the most between levels in samples from a group of substantially healthy control subjects and levels in samples from a group of diseased subjects or patients.

[0054] A reference level may also be, for example, the level obtained from the subject at an earlier timepoint, i.e., prior to treatment.

[0055] A biomarker measured in a method of the invention may be detected in a biological sample of the subject (e.g., tissue, fluid), including, but not limited to any tissue (in particular a homogenized tissue sample (e.g., a tissue sample obtained by biopsy)), cell, fluid, saliva, tears, urine, sweat, sputum, liquid biopsy, blood, serum, plasma, ascites, cyst fluid, vaginal fluid, cervico-vaginal fluid, or other material derived from a subject or patient. Preferably the biological sample is blood, serum or plasma.

[0056] The present invention relates to an in vitro method for assessing whether a subj ect has endometriosis comprising a step of measuring, in a biological sample of the subject, the expression level of at least two biomarkers.

[0057] In one embodiment, the method of the present invention further comprises a step of comparing the measured expression levels with reference expression levels, thereby assessing whether the subject has endometriosis.

[0058] In one embodiment, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of at least two biomarkers selected from the group comprising or consisting of MUC-16, ANXA1, ARHGAP1, PECAM-1, GLB1, CES2, EREG, PRKCQ, ESM-1, TMPRSS15, IGLC2, CD84, ARSB, PON2, PCSK9, MAP2K6, t-PA, CTSH, SSC4D, MMP-3, IL17RB, VEGF-A and EN-RAGE.

[0059] The present invention thus relates to a signature comprising at least two biomarkers selected from the group comprising or consisting of MUC-16, ANXA1, ARHGAP1, PECAM-1, GLB1, CES2, EREG, PRKCQ, ESM-1, TMPRSS15, IGLC2, CD84, ARSB, PON2, PCSK9, MAP2K6, t-PA, CTSH, SSC4D, MMP-3, IL17RB, VEGF-A and EN-RAGE.

[0060] The present invention further relates to an in vitro method for assessing whether a subject has endometriosis comprising a step of measuring, in a biological sample of the subject, the expression level of at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP1. The present invention thus relates to a signature comprising at least two biomarkers, wherein said at least two biomarkers comprise MUC-16 and at least one of ANXA1 and ARHGAP1.

[0061] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is MUC-16 (Mucin 16, that may also be referred to as CA125, cancer antigen 125). MUC-16 / CA125 is most commonly known as a biomarker for ovarian cancer, though other cancers as well as a number of benign conditions also cause serum levels to be increased. MUC-16 / CA125 is a component of the ocular surface, respiratory tract, and epithelia of the female reproductive tract. MUC-16 / CA125 is a 14507 amino acid proteinwith UniProt Accession No. Q8WX17. In one embodiment of the present invention, the expression level of MUC-16 / CA125 is increased in subjects with endometriosis as compared to the expression level of MUC-16 / CA125 in substantially healthy subjects.

[0062] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is ANXA1 (Annexin Al). ANXA1 is a 346 amino acid protein with UniProt Accession No. P04083. ANXA1 is most commonly known as playing a role in the innate immune response as effector of glucocorticoid-mediated responses and regulator of the inflammatory process. In one embodiment of the present invention, the expression level of ANXA1 is increased in subjects with endometriosis as compared to the expression level of ANXA1 in substantially healthy subjects.

[0063] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is ARHGAP1 (Rho GTPase activating protein 1). ARHGAP1 is a 439 amino acid protein with UniProt Accession No. Q07960. ARHGAP1 is most commonly known as a GTPase activator for the Rho, Rac and Cdc42 proteins, converting them to the putatively inactive GDP -bound state. In one embodiment of the present invention, the expression level of ARHGAP1 is decreased in subjects with endometriosis as compared to the expression level of ARHGAP1 in substantially healthy subjects.

[0064] According to some embodiments, the method of the present invention comprises measuring the expression level of at least three biomarkers, wherein the at least three biomarkers are:• MUC-16, and• ANXA1 or ARHGAP 1 , and• at least one biomarker selected from the group comprising or consisting of EREG, ANXA1, ARHGAP1, t-PA, PCSK9, ARSB, ESM-1 and TMPRSS15.

[0065] The present invention thus relates to a signature comprising at least 3 biomarkers, wherein the at least 3 biomarkers comprise or consist of (i) MUC-16, (ii) ANXA1 or ARHGAP 1 and (iii) at least one biomarker selected from the group comprising orconsisting of EREG, ANXA1, ARHGAP1, t-PA, PCSK9, ARSB, ESM-1 and TMPRSS15.

[0066] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is EREG (Proepiregulin). EREG refers to a 169 amino acids protein with UniProt Accession No. 014944. Among its reported functions, EREG is a ligand of the EGF receptor / EGFR and ERBB4 stimulating EGFR and ERBB4 tyrosine phosphorylation. EREG contributes to inflammation, wound healing, tissue repair, and oocyte maturation by regulating angiogenesis and vascular remodeling and by stimulating cell proliferation. In one embodiment, the expression level of EREG is decreased in subjects with endometriosis compared to substantially healthy subjects.

[0067] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is t-PA (for Tissue type Plasminogen activator, that may also be referred to as PLAT). t-PA is a serine protease constituted of five functional domains through which it interacts with different substrates, binding proteins, and receptors. In the last years, great interest has been given to the clinical relevance of targeting t-PA in different diseases of the central nervous system, in particular, stroke. Among its reported functions in the central nervous system, t-PA displays both neurotrophic and neurotoxic effects. t-PA is a 562 amino acid protein with UniProt Accession No. P00750. In one embodiment of the invention, the expression level of t-PA is decreased in subjects with endometriosis compared to substantially healthy subjects.

[0068] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is PCSK9 (Proprotein convertase subtilisin / kexin type 9). PCSK9 refers to a 692 amino acids protein with UniProt Accession No. Q8NBP7. Among it reported functions, it is a crucial player in the regulation of plasma cholesterol homeostasis. In one embodiment of the invention, the expression level of PCSK9 is decreased in subjects with endometriosis compared to substantially healthy subjects.

[0069] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is ARSB (Aryl sulfatase B). ARSB refers to a 533 amino acids protein with UniProt Accession No. Pl 5848. Among it reported functions, it is a crucial player in theregulation of cell adhesion, cell migration and invasion in colonic epithelium. In one embodiment of the invention, the expression level of ARSB is decreased in subjects with endometriosis compared to substantially healthy subjects.

[0070] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is ESM-1 (Endothelial cell-specific molecule 1). ESM-1 refers to a 184 amino acids protein with UniProt Accession No. Q9NQ30. Among it reported functions, it is involved in angiogenesis; promotes angiogenic sprouting and may have potent implications in lung endothelial cell-leukocyte interactions. In one embodiment of the invention, the expression level of ESM-1 is increased in subjects with endometriosis compared to substantially healthy subjects.

[0071] In one embodiment, one of the biomarkers measured in the in vitro method of the invention is TMPRSS15 (Enteropeptidase). TMPRSS15 refers to a 1019 amino acids protein with UniProt Accession No. P98073. Among it reported functions, it is a crucial player in initiating activation of pancreatic proteolytic proenzymes (trypsin, chymotrypsin and carboxypeptidase A). It catalyzes the conversion of trypsinogen to trypsin which in turn activates other proenzymes including chymotrypsinogen, procarboxypeptidases, and proelastases. In one embodiment of the invention, the expression level of TMPRSS15 is decreased in subjects with endometriosis as compared to the expression level of TMPRSS15 in substantially healthy subjects.

[0072] Individual biomarkers are diagnostic biomarkers. But the inventors found that a specific combination of biomarkers, i.e., each signature as described herein, provides an enhanced performance in diagnostic of endometriosis, as measured by sensitivity, specificity and / or area under the Receiver Operating Characteristic (ROC) curve (AUC) as compared to any single biomarker alone, or any other combination of previously identified biomarkers. Specifically, the detection and measurement of the expression level of the plurality of biomarkers of the invention in a sample increases the sensitivity, accuracy, and specificity of the diagnostic test.

[0073] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least three, at least four, at least five, at least six, at least seven, at least eight, at least nine, at least ten, at least eleven or more biomarkers.

[0074] In one embodiment, the method comprises measuring the expression levels of (i) MUC-16, and (ii) of either ANXA1 or ARHGAP1 and (iii) of at least one, at least two, at least three, at least four, at least five, at least six, at least seven, at least eight, at least nine or more biomarkers.

[0075] In one embodiment, the method comprises measuring the expression levels of (i) MUC-16 and (ii) of either ANXA1 or ARHGAP1 and (iii) of at least one, at least two, at least three, at least four, at least five, at least six, or at least seven biomarkers selected from the group comprising or consisting of t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, ARHGAP1, EREG and ARSB.

[0076] In one embodiment, the method comprises measuring the expression levels of (i) MUC-16, (ii) of either ANXA1 or ARHGAP1, and (iii) of 1, 2, 3, 4, 5, 6, or 7 biomarkers selected from the group comprising or consisting of t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, ARHGAP1, EREG and ARSB.

[0077] In one embodiment, the method comprises measuring the expression levels of a combination of 2 biomarkers, a first biomarker being MUC-16 and a second biomarker selected from the group consisting of ANXA1 and ARHGAP1. In one embodiment, the method comprises measuring the expression levels of MUC-16 and ANXA1. In one embodiment, the method comprises measuring the expression levels of MUC-16 and ARHGAP1.

[0078] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least two biomarkers, wherein the at least two biomarkers are (i) MUC-16, and (ii) either ANXA1 or ARHGAP1.

[0079] In one embodiment, the method comprises measuring the expression levels of a combination of 3 biomarkers.

[0080] In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and t-PA. In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and TMPRSS15. In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and PCSK9. In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and ESM-1. In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and ARHGAP1. In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and EREG. In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and ARSB.

[0081] In one embodiment, the method comprises measuring the expression levels of MUC-16, ARHGAP1 and t-PA. In one embodiment, the method comprises measuring the expression levels of MUC-16, ARHGAP1 and TMPRSS15. In one embodiment, the method comprises measuring the expression levels of MUC-16, ARHGAP1 and PCSK9. In one embodiment, the method comprises measuring the expression levels of MUC-16, ARHGAP1 and ESM-1. In one embodiment, the method comprises measuring the expression levels of MUC-16, ANXA1 and ARHGAP1. In one embodiment, the method comprises measuring the expression levels of MUC-16, ARHGAP1 and EREG. In one embodiment, the method comprises measuring the expression levels of MUC-16, ARHGAP1 and ARSB.

[0082] In one embodiment, the signature comprises or consists of MUC-16, ANXA1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1 and PCSK9. In one embodiment, the signature comprises or consists of MUC- 16, ANXA1 and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1 and ARHGAP1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1 and EREG. In one embodiment, the signature comprises or consists of MUC-16, ANXA1 and ARSB.

[0083] In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16,ARHGAP1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1 and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1 and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1 and ARHGAP1. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1 and EREG. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1 and ARSB.

[0084] In some embodiments, the method comprises measuring the expression levels of three biomarkers, wherein the three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0085] In some embodiments, the signature comprises or consists of:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0086] In one embodiment, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of MUC-16, ANXA1 and EREG.

[0087] In one embodiment, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of MUC-16, ANXA1 and ARSB.

[0088] In one embodiment, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of MUC-16, ANXA1 and TMPRSS15.

[0089] In one embodiment, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of MUC-16, ARHGAP1 and PCSK9.

[0090] In one embodiment, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of MUC-16, ARHGAP1 and t-PA.

[0091] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least three biomarkers, wherein the at least three biomarkers are (i) MUC-16, (ii) ANXA1, and (iii) at least one biomarker selected from the group comprising or consisting of t-PA, TMPRSS15, PCSK9, ESM-1, ARHGAP1, EREG and ARSB.

[0092] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least three biomarkers, wherein the at least three biomarkers are (i) MUC-16, (ii) ARHGAP1, and (iii) at least one biomarker selected from the group comprising or consisting of t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, EREG and ARSB.

[0093] In one embodiment, the method comprises measuring the expression levels of a signature of 4 biomarkers. In one embodiment, the 4 biomarkers are (i) MUC-16, (ii) ARHGAP1 and (iii) 2 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, EREG and ARSB. In one embodiment, the 4 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 2 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0094] In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB and EREG. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB and ESM-1. Inone embodiment, the in vitro method comprises measuring the expression levels of MUC- 16, ANXA1, ARSB and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ESM-1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ESM-1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ESM-1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, PCSK9 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, t- PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB and EREG. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ESM-1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ESM-1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ESM-1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, PCSK9 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1 and EREG. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1 and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1 and ARSB.

[0095] In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB and EREG. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ESM-1 and PCSK9. In one embodiment, the signature comprisesor consists of MUC-16, ANXA1, ESM-1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ESM-1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB and EREG. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ESM-1 and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ESM-1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ESM-1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1 and EREG. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1 and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1 and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1 and ARSB.

[0096] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least 4 biomarkers. In one embodiment, the at least 4 biomarkers are (i) MUC-16, (ii) ARHGAP1, and (iii) 2 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, EREG and ARSB. In one embodiment, the at least 4 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 2 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0097] In some embodiments, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of a signature of five biomarkers. In one embodiment, the 5 biomarkers are (i) MUC-16, (ii) ARHGAP1 and (iii) 3 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, EREG and ARSB. In one embodiment, the 5 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 3 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0098] In one embodiment, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of a signature of five biomarkers, consisting of MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0099] In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, EREG and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, EREG and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, EREG and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, EREG and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, ESM-1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, ESM-1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, ESM-1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, PCSK9 and TMPRSS15.In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARSB, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG, ESM- 1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG, ESM-1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG, ESM-1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG, PCSK9 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, EREG, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ESM-1, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ESM-1, PCSK9 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ESM-1, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, PCSK9, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, EREG and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, EREG and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, EREG and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, EREG and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, ESM-1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, ESM-1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, ESM-1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, PCSK9 and TMPRSS15. In oneembodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ARSB, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG, ESM-1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG, ESM-1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG, ESM-1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG, PCSK9 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, EREG, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ESM-1, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ESM- 1, PCSK9 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, ESM-1, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ARHGAP1, PCSK9, t-PA and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, ARSB and EREG. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, ARSB and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, ARSB and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, ARSB and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC- 16, ANXA1, ARHGAP1, ARSB and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, EREG and ESM-1. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, EREG and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, EREG and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, EREG andTMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, ESM-1 and PCSK9. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, ESM-1 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, ESM-1 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, PCSK9 and t-PA. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, PCSK9 and TMPRSS15. In one embodiment, the in vitro method comprises measuring the expression levels of MUC-16, ANXA1, ARHGAP1, t-PA and TMPRSS15.

[0100] In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, EREG and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, EREG and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, EREG and t-PA. In one embodiment, the signature comprises or consists ofMUC-16, ANXA1, ARSB, EREG and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, ESM-1 and PCSK9. In one embodiment, the signature comprises or consists ofMUC-16, ANXA1, ARSB, ESM-1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, ESM-1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARSB, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG, ESM-1 and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG, ESM-1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG, ESM-1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, EREG, t-PA and TMPRSS15. Inone embodiment, the signature comprises or consists of MUC-16, ANXA1, ESM-1, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ESM-1, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ESM-1, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, PCSK9, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, EREG and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, EREG and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, EREG and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, EREG and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, ESM-1 and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, ESM-1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, ESM-1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ARSB, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG, ESM-1 and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG, ESM- 1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG, ESM-1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, EREG, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ESM-1, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ESM-1, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, ESM-1, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ARHGAP1, PCSK9, t-PA and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1,ARSB and EREG. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, ARSB and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, ARSB and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, ARSB and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, ARSB and TMPRSS15. In one embodiment, the signature comprises or consists of MUC- 16, ANXA1, ARHGAP1, EREG and ESM-1. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, EREG and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, EREG and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, EREG and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, ESM-1 and PCSK9. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, ESM-1 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, ESM-1 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, PCSK9 and t-PA. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, PCSK9 and TMPRSS15. In one embodiment, the signature comprises or consists of MUC-16, ANXA1, ARHGAP1, t-PA and TMPRSS15.

[0101] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least 5 biomarkers. In one embodiment, the at least 5 biomarkers are (i) MUC-16, (ii) ARHGAP1 and (iii) 3 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, EREG and ARSB. In one embodiment, the at least 5 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 3 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0102] In one embodiment, the in vitro method comprises measuring the expression levels of a signature of 6 biomarkers. In one embodiment, the 6 biomarkers are (i) MUC- 16, (ii) ARHGAP1 and (iii) 4 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM- 1, ANXA1, EREG and ARSB. In one embodiment, the 6 biomarkers are (i) MUC-16, (ii)ANXA1 and (iii) 4 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0103] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least 6 biomarkers. In one embodiment, the at least 6 biomarkers are (i) MUC-16, (ii) ARHGAP1 and (iii) 4 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, EREG and ARSB. In one embodiment, the at least 6 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 4 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0104] In one embodiment, the in vitro method comprises measuring the expression levels of a signature of 7 biomarkers. In one embodiment, the 7 biomarkers are (i) MUC- 16, (ii) ARHGAP1 and (iii) 5 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM- 1, ANXA1, EREG and ARSB. In one embodiment, the 7 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 5 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0105] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least 7 biomarkers. In one embodiment, the at least 7 biomarkers are (i) MUC-16, (ii) ARHGAP1 and (iii) 5 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM-1, ANXA1, EREG and ARSB. In one embodiment, the at least 7 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 5 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0106] In one embodiment, the in vitro method comprises measuring the expression levels of a signature of 8 biomarkers. In one embodiment, the 8 biomarkers are (i) MUC- 16, (ii) ARHGAP1 and (iii) 6 biomarkers selected from t-PA, TMPRSS15, PCSK9, ESM- 1, ANXA1, EREG and ARSB. In one embodiment, the 8 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 6 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0107] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least 8 biomarkers. In one embodiment, the at least 8 biomarkers are (i) MUC-16, (ii) ARHGAP1 and (iii) 6 biomarkers selected from t-PA, TMPRSS15,PCSK9, ESM-1, ANXA1, EREG and ARSB. In one embodiment, the at least 8 biomarkers are (i) MUC-16, (ii) ANXA1 and (iii) 6 biomarkers selected from ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0108] In one embodiment, the in vitro method comprises measuring the expression levels of a signature of 9 biomarkers. In one embodiment, the 9 biomarkers are MUC-16, ANXA1, ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0109] In one embodiment, the method comprises measuring the expression levels of a signature comprising at least 9 biomarkers. In one embodiment, the at least 9 biomarkers are MUC-16, ANXA1, ARHGAP1, t-PA, TMPRSS15, PCSK9, ESM-1, EREG and ARSB.

[0110] In some other embodiments, the in vitro method of the invention for assessing whether a subject has endometriosis comprises measuring the expression levels of any other biomarkers that are differentially present or expressed in subjects having endometriosis as compared to the expression level of this other biomarker in substantially healthy subjects.

[0111] Examples of other biomarkers differentially expressed in subjects having endometriosis as compared to substantially healthy subjects include, but are not limited to, PECAM-1, GLB1, CES2, PRKCQ, IGLC2, CD84, PON2, MAP2K6, CTSH, SSC4D, MMP-3, IL17RB, VEGF-A and EN-RAGE.

[0112] In some embodiments, the in vitro method of the invention comprises measuring the expression levels of a signature as described herein, and further comprises measuring the expression levels of at least one other biomarker. Said at least one other biomarker may selected from the group comprising or consisting of PECAM-1, GLB1, CES2, PRKCQ, IGLC2, CD84, PON2, MAP2K6, CTSH, SSC4D, MMP-3, IL17RB, VEGF-A and EN-RAGE.

[0113] In one embodiment, the biological sample is selected from the group consisting of any tissue, notably a homogenized tissue sample derived from the subject or a tissue sample obtained by biopsy from the subject, cell, fluid, saliva, tears, urine, sweat, sputum,liquid biopsy, blood, serum, plasma, ascites, cyst fluid, vaginal fluid and cervico-vaginal fluid, preferably the biological sample is blood, serum or plasma.

[0114] In some embodiments, the biomarkers measured in a method of the invention can be isolated, detected and their expression levels measured by any method known in the art.

[0115] In certain embodiments, this isolation is accomplished using the mass and / or binding characteristics of the biomarkers. For example, a sample comprising the biomarkers can be subject to chromatographic fractionation and subject to further separation by, e.g., acrylamide gel electrophoresis. Knowledge of the identity of the biomarker also allows their isolation by immunoaffmity chromatography. By “isolated biomarker” is meant at least 60%, by weight, free from proteins and naturally-occurring organic molecules with which the biomarker is naturally associated. Preferably, the preparation is at least 75%, more preferably 80, 85, 90 or 95% pure or at least 99%, by weight, a purified biomarker.

[0116] The expression level of biomarkers measured in an in vitro method according to the invention can be detected by any suitable method. The methods described herein can be used individually or in combination for a more accurate detection of the biomarkers (e.g., biochip in combination with mass spectrometry, immunoassay in combination with mass spectrometry, and the like).

[0117] In one embodiment, the level of the biomarkers is measured by immunoassays, including ELISA, Western blot, immune-electrophoresis, immunostaining and / or Proximity Extension Assay (PEA); enzymatic activity; flow cytometry; spectrometry, including mass spectrometry; RT-PCR; RT-qPCR; Northern Blot; or hybridization techniques, or any combination thereof.

[0118] In some embodiments, the expression level of the biomarkers is measured at the polynucleotide level using any well-known technique to the skilled man in the art (such as, for example, RT-qPCR).

[0119] In some embodiments, the expression level of biomarkers of the invention is measured by immunoassay. Immunoassay typically utilizes an antibody (or another compound that specifically binds the biomarker) to detect the presence or level of a biomarker in a sample.

[0120] This invention contemplates traditional immunoassays and protein-binding assay including, for example, Western blot, antibody-based assay, antigen-binding protein based assay, a protein-based array, an enzyme-linked immunosorbent assay (ELISA), flow cytometry, a protein array, a blot, a Western blot, nephelometry, turbidimetry, chromatography, mass spectrometry, enzymatic activity, proximity extension assay (PEA), and an immunoassay selected from RIA, immunofluorescence, immunochemiluminescence, immunoelectrochemiluminescence, immunoelectrophoresis, a competitive immunoassay, immunoprecipitation, nephelometry. Other forms of immunoassay include magnetic immunoassay, radioimmunoassay, and real-time immunoquantitative PCR (qPCR).

[0121] In some embodiments, the expression level of the biomarkers of the invention is measured with Proximity Extension Assay (PEA).

[0122] In some embodiments, the sample is analyzed by means of a biochip (also known as a microarray). The biomarkers of the invention are useful as hybridizable array elements in a biochip. Biochips generally comprise solid substrates and have a generally planar surface, to which a capture reagent (also called an adsorbent or affinity reagent) is attached. Frequently, the surface of a biochip comprises a plurality of addressable locations, each of which has the capture reagent bound there.

[0123] In some embodiments, the expression level of biomarkers of this invention are detected by mass spectrometry (MS). Mass spectrometry is a well-known tool for analyzing chemical compounds that employs a mass spectrometer to detect gas phase ions.

[0124] In some embodiments, the mass spectrometer is a laser desorption / ionization mass spectrometer. The analysis of proteins by LDI can take the form of Matrix-assisted Laser Desorption / ionization (MALDI) or of Electrospray Ionization (SELDI).

[0125] In some embodiments, the in vitro method of the invention for assessing whether a subject has endometriosis comprises combining the measured expression levels in a score.

[0126] In some embodiments, the in vitro method of the invention for assessing whether a subject has endometriosis comprises comparing said subject’s sample score to a reference score and determining whether the subject has endometriosis, based on this comparison.

[0127] Thus, in some embodiments, the measured expression levels of biomarkers can be transformed into a diagnostic result (e.g., a score). This may involve an algorithm which produce a diagnostic result as a function of the measured levels. Each individual biomarker of the signature of the invention makes a different contribution to the overall diagnostic result and each biomarker may be weighted differently.

[0128] In one embodiment, the comparison may be carried out manually or computer- assisted. Thus, the comparison may be carried out by a computing device. The values of the measured expression level of the biomarkers in the subject’s sample or the subject’s sample score and the reference score can be, e.g., compared to each other and the said comparison can be automatically carried out by a computer program executing an algorithm for the comparison. The computer program carrying out the said evaluation will provide the desired assessment in a suitable output format. For a computer-assisted comparison, the value of the measured or detected level may be compared to values corresponding to suitable references or to a reference score which are stored in a database by a computer program. The computer program may further evaluate the result of the comparison, i.e., automatically provide the desired assessment in a suitable output format.

[0129] In some embodiments, the expression levels of biomarkers in a subject’s sample can be computed in a score (i.e., the subject’ s sample score) and compared to the reference score known to be associated with the presence or absence of endometriosis.

[0130] In some embodiments, the reference score is determined (or was previously determined) by measuring and combining in a substantially identical manner the expression level(s) of at least two biomarkers (or a combination of biomarkers) asdescribed herein in samples of (i) substantially healthy subjects and of (ii) diseased subjects, and comparing these combined expression level(s).

[0131] In some embodiments, the reference score can be relative to a score derived from a reference population (comprising either substantially healthy or diseased subjects), including without limitation, such subjects having similar age range, subjects in the same or similar ethnic group, similar endometriosis history, no endometriosis history and the like.

[0132] In some embodiments, the reference population (comprising either substantially healthy or diseased subjects) comprises preferably at least 50, 60, 70, 80, 90, 92, 100, 120, at least 150, at least 180, at least 200, at least 500, at least 1000 well-characterized individuals. Said individuals from said appropriate reference population are used to establish a "reference score".

[0133] In some embodiments, the subject is identified as suffering from endometriosis when said subject’s sample score is higher than the reference score.

[0134] In some embodiments, the subject is identified as suffering from endometriosis when the subject’s sample score is lower than the reference score.

[0135] In some embodiments, a difference between the reference score and the subj ect’ s sample score is indicative of the presence (existence) of endometriosis.

[0136] In some embodiments, a difference between the reference score and said subject’s sample score is indicative of a risk of endometriosis development or the presence or further progression of endometriosis.

[0137] In some embodiments, a difference between the reference score and said subject’s sample score is indicative of a lowered risk of endometriosis development or the absence or bettering of endometriosis.

[0138] In some embodiments, the in vitro method of the invention for assessing whether a subject has endometriosis comprises determining a score based on the combination of the measured expression levels using a linear model.

[0139] In one embodiment, the score is determined by mathematically combining the expression levels measured in the subject’s sample, such as, for example, in a logistic regression or in a linear regression.

[0140] In some embodiments, the score, named subject’s sample score, is determined based on a linear combination of the measured expression levels.

[0141] In some embodiments, the in vitro method of the invention for assessing whether a subject has endometriosis comprises determining a score based on the combination of the measured expression levels using a complex model.

[0142] In some embodiments, the expression “using a complex model” refers to the use of a machine learning algorithm.

[0143] A variety of learning algorithms can be used to infer a condition or state of a subject. Machine learning algorithms may be supervised or unsupervised. Learning algorithms include, but are not limited to, artificial neural networks (e.g., back propagation networks), discriminant analyses (e.g., Bayesian classifier, Fischer analysis), support vector machines (SVM), decision trees (e.g., recursive partitioning processes, such as classification and regression trees [CART]), random forests, genetic algorithms, linear classifiers (e.g., multiple linear regression [MLR], partial least squares [PLS] regression, principal components regression [PCR], logistic regression, linear discriminant analysis [LDA], quadratic discriminant analysis [QDA]), perceptron algorithms, deep neural networks (DNN), clustering algorithms, k-nearest neighbors algorithms (k-NN), Gaussian mixture model (GMM) algorithms, nearest centroid algorithms, gradient boosting algorithms (such as, e.g., extreme gradient boosting [XG Boost] algorithms or adaptative boosting [AdaBoost] algorithms), linear mixed effects model algorithms, kernel matching pursuit algorithms, kernel fisher discriminate analysis algorithms, kernel principal components analysis algorithms, Bayesian probability function algorithms, Markov Blanket algorithms, hierarchical clustering and cluster analysis. The learning algorithm generates a model or classifier that can be used to make an inference, e.g., an inference about a disease state of a subject.

[0144] In one embodiment, the at least one machine learning algorithm was previously trained with at least one training dataset.

[0145] In one embodiment, the at least one training dataset comprises information relating to the expression levels and scores from samples previously obtained from diseased and substantially healthy subjects.

[0146] In one embodiment, the at least one training dataset comprises information relating to the expression level of biomarkers of the signatures of the invention from samples previously obtained from subjects of a reference population not having endometriosis.

[0147] In one embodiment, the at least one training dataset comprises information relating to the expression level of biomarkers of the signatures of the invention from samples previously obtained from subjects of a population having endometriosis.

[0148] In one embodiment, the at least one machine learning algorithm is selected from the group comprising an artificial neural network (ANN), a perceptron algorithm, a deep neural network, a clustering algorithm, a k-nearest neighbors algorithm (k-NN), a decision tree algorithm, a random forest algorithm, a linear regression algorithm, a logistic regression algorithm, a linear discriminant analysis (LDA) algorithm, a quadratic discriminant analysis (QDA) algorithm, a support vector machine (SVM), a Bayes algorithm, a simple rule algorithm, a clustering algorithm, a meta-classifier algorithm, a Gaussian mixture model (GMM) algorithm, a nearest centroid algorithm, a gradient boosting algorithm (such as, e.g., an extreme gradient boosting [XG Boost] algorithm or an adaptative boosting [AdaBoost] algorithm), a linear mixed effects model algorithm, and a combination thereof.

[0149] In one embodiment, the combination of the measured expression levels may be carried out manually or computer-assisted.

[0150] The creation of algorithms for converting measured expression levels into scores is well known in the art. For example, linear or non-linear (or complex) classifier algorithms can be used. These algorithms can be trained using data from any particulartechnique for measuring the biomarker(s). Suitable training data will have been obtained by measuring the biomarkers in "disease" and "control" samples, i.e., samples from subjects known to suffer from endometriosis and from subjects known not to suffer from endometriosis. In some embodiments, the control samples may include samples from subjects with a related disease which is to be distinguished from endometriosis. The classifier algorithm is modified until it can distinguish between the subject or patient and control samples, e.g., by adding or removing biomarkers from the analysis, by changes in weighting, etc. Thus, the method of the invention may include a step of analyzing biomarker expression levels in a subject's sample by using a classifier algorithm which distinguishes between endometriosis subjects and non- endometriosis subjects based on measured biomarker expression levels in samples taken from such subjects.

[0151] In one embodiment, the method comprises using a quantitative algorithm to determine if the expression levels of a combination at least two biomarkers or of a signature as described herein in the subject’s sample is higher or lower than a previously defined reference. The algorithm may be a trained algorithm. In various embodiments, the algorithm is drawn from the group consisting essentially of: linear or nonlinear regression algorithms; linear or nonlinear classification algorithms; ANOVA; neural network algorithms; genetic algorithms; support vector machines algorithms; hierarchical analysis or clustering algorithms; hierarchical algorithms using decision trees; kernel based machine algorithms such as kernel partial least squares algorithms, kernel matching pursuit algorithms, kernel fisher discriminate analysis algorithms, or kernel principal components analysis algorithms; Bayesian probability function algorithms; Markov Blanket algorithms; a plurality of algorithms arranged in a committee network; and forward floating search or backward floating search algorithms. Such algorithms may be used in supervised or unsupervised learning modes.

[0152] In a specific embodiment, quantitative algorithms can be used to determine the presence, extent, severity, or stage of disease, to determine the right treatment approach (e.g., oral contraceptives, disease-specific therapy, surgical intervention), to select the appropriate dose for a medical treatment, to determine whether a subject or patient islikely to respond to a particular medical or surgical treatment, to monitor response to treatment, or to monitor disease progression.

[0153] In one embodiment, the method according to the invention includes deriving a score from the quantitative algorithm or mathematical formula. In one embodiment, the mathematic formula is a logistic regression or a linear regression.

[0154] In another embodiment, the score can serve as a cut off value for distinguishing between two or more potential outcomes (e.g., high or low risk of endometriosis presence, progression or recurrence or stage of endometriosis).

[0155] In one embodiment, a score serves as a cutoff value in order to determine the presence or absence of endometriosis.

[0156] In one embodiment, a score serves as a cutoff value in order to determine the right treatment approach (e.g., oral contraceptives, disease-specific therapy, surgical intervention).

[0157] In one embodiment, a score serves as a cutoff value in order to select the appropriate dose for a medical treatment.

[0158] In one embodiment, a score serves as a cutoff value in order to determine whether a subject or patient is likely to respond to a particular medical or surgical treatment.

[0159] In one embodiment, a score serves as a cutoff value in order to monitor response to treatment.

[0160] In one embodiment, a score serves as a cutoff value in order to monitor disease progression.

[0161] The accuracy of a diagnostic test can be characterized using any method well known in the art, including, but not limited to, a Receiver Operating Characteristic curve (“ROC curve”).

[0162] Common tests for statistical significance include, among others, t-test, ANOVA, Kruskal-Wallis, Wilcoxon, Mann-Whitney and odds ratio. Biomarkers, alone or incombination, provide measures of relative likelihood that a subject belongs to a phenotypic status of interest. In one embodiment, the differential expression of the biomarkers of a signature of the invention in a subject sample can be useful in characterizing the subject as having endometriosis, or for selecting a treatment regimen (e.g., selecting that the subject be evaluated and / or treated by a surgeon that specializes in endometriosis).

[0163] In some embodiments, biomarkers and combinations thereof used in the in vitro methods of the present invention have an AUC greater than 0.54, greater than 0.60, greater than 0.65, greater than 0.70, greater than 0.75, greater than 0.78, greater than 0.80, greater than 0.85, greater than 0.88, greater than 0.89, or greater than 0.9.

[0164] In one embodiment, the method of the present invention is a replacement diagnostic test. As used herein, a replacement diagnostic test refers to a diagnostic test which performances enable the replacement of existing gold-standard diagnostic method (higher sensibility / specificity values or similar performances with other advantages).

[0165] In another embodiment, the method of the present invention is a rule-out test. As used herein, a rule-out test refers to a test used as an initial step in a diagnostic pathway to identify the group of women who need further testing with the current gold-standard diagnostic method. Although ideally a rule-out test has a high sensitivity and specificity, it may have a lower specificity but higher sensitivity than the current gold-standard diagnostic method. The rule-out test does not aim to improve the diagnostic accuracy of the current method but rather to reduce the number of individuals having an unnecessary diagnostic test. A negative result from a test with high sensitivity will exclude the disease with high certainty independent of the specificity. However, a positive result has less diagnostic value particularly when the specificity is low.

[0166] In any of the methods described herein, the step of correlating the measurement of the expression level of biomarker(s) with endometriosis can be performed on general- purpose or specially programmed hardware or software.

[0167] In an embodiment, the results of the measure of the expression levels of biomarkers of the invention are subjected to data processing. Data processing can beperformed by a software classification algorithm. Such software classification algorithms are well known in the art and one of ordinary skill can readily select and use the appropriate software to analyze the results obtained from a specific detection method.

[0168] In one embodiment, the resulting data are transformed into various formats for display.

[0169] In one embodiment, software used to analyze the data can include code that applies an algorithm to the analysis of the expression levels measured in the in vitro method of the invention.

[0170] In some embodiments, data (including, for example, expression levels) derived from the assays (e.g., PEA assays) that are generated using samples obtained from subject for which the status (e.g., affected or not with endometriosis) is known (herein after “known samples”) can then be used to “train” a classification model. The data that are derived from the expression levels and are used to form the classification model can be referred to as a “training data set”. Once trained, the classification model can recognize patterns in data derived from expression levels generated using unknown samples. The classification model can then be used to classify the unknown samples into classes (e.g., affected or not with endometriosis). This can be useful, for example, in predicting whether or not a particular biological sample is associated with a certain biological condition (e.g., diseased versus non-diseased).

[0171] In one embodiment, the training data set that is used to form the classification model may comprise raw data or pre-processed data.

[0172] Classification models can be formed using any suitable statistical classification (or “learning”) method that attempts to segregate bodies of data into classes based on objective parameters present in the data. Classification methods may be either supervised or unsupervised.

[0173] In supervised classification, training data containing examples of known categories are presented to a learning mechanism, which learns one or more sets of relationships that define each of the known classes. New data may then be applied to thelearning mechanism, which then classifies the new data using the learned relationships. Examples of supervised classification processes include linear regression processes (e.g., multiple linear regression (MLR), partial least squares (PLS) regression and principal components regression (PCR)), binary decision trees (e.g., recursive partitioning processes such as CART - classification and regression trees), random forest, artificial neural networks such as back propagation networks, discriminant analyses (e.g., Bayesian classifier or Fischer analysis), logistic classifiers, support vector classifiers (support vector machines), and a combination thereof.

[0174] In some embodiments, a supervised classification method is a logistic regression process. In other embodiments, a supervised classification method is a mixed model of neural networks and machine learning algorithms. Further details about mixing of deep learning algorithms are provided in “LeDell and Poirier, 2020, "H2o automl: Scalable automatic machine learning." Proceedings of the AutoML Workshop at ICML. Vol. 2020”.

[0175] In other embodiments, the classification models that are created can be formed using unsupervised learning methods. Unsupervised classification attempts to learn classifications based on similarities in the training data set, without pre-classifying the expression levels from which the training data set was derived. Unsupervised learning methods include cluster analyses. A cluster analysis attempts to divide the data into “clusters” or groups that ideally should have members that are very similar to each other, and very dissimilar to members of other clusters. Similarity is then measured using some distance metric, which measures the distance between data items, and clusters together data items that are closer to each other. Clustering techniques include the MacQueen’s K- means algorithm and the Kohonen’s Self-Organizing Map algorithm.

[0176] Learning algorithms asserted for use in classifying biological information are described, for example, in PCT International Publication No. WO 01 / 31580 (Barnhill et al., “Methods and devices for identifying patterns in biological systems and methods of use thereof’), U.S. Patent Application No.20020193950 Al (Gavin et al., “Method or analyzing mass spectra”), U.S. Patent Application No.20030004402 Al (Hitt et al., “Process for discriminating between biological states based on hidden patterns frombiological data”), and U.S. Patent Application No.20030055615 Al (Zhang and Zhang, “Systems and methods for processing biological expression data”).

[0177] The classification models can be formed on and used on any suitable digital computer. Suitable digital computers include micro, mini, or large computers using any standard or specialized operating system, such as a Unix, Windows™ or Linux™ based operating system. The digital computer that is used may be physically separated from the device that enables measurement of the biomarkers, or it may be coupled to the device.

[0178] The training data set and the classification models according to embodiments of the invention can be embodied by computer code that is executed or used by a digital computer. The computer code can be stored on any suitable computer readable media including optical or magnetic disks, sticks, tapes, etc., and can be written in any suitable computer programming language including C, C++, visual basic, Python, R, etc, preferably Python and R.

[0179] The learning algorithms described above are useful both for developing classification algorithms for the biomarkers already discovered, or for finding new biomarkers for endometriosis. The classification algorithms, in turn, form the base for diagnostic tests by providing diagnostic values (e.g., cut-off points) for biomarkers used in combination, or for the scores measured in the present invention.

[0180] According to the invention, the subject is a female subject, in particular a human female subject.

[0181] In some specific embodiments, the subject is a woman that is not in menopausal or postmenopausal status.

[0182] In some embodiments, the subject may be asymptomatic or pre- symptomatic for endometriosis or may already be displaying clinical symptoms, e.g., pelvic pain, dysmenorrhea, dyspareunia, dyschesia, fatigue and / or infertility. For pre-symptomatic subjects the invention is useful for predicting that symptoms may develop in the future if no preventative action is taken. For subjects already displaying clinical symptoms, theinvention may be used to confirm or resolve another diagnosis. The subject may already have begun treatment for endometriosis.

[0183] In some embodiments the subject may already be known to be predisposed to development of endometriosis e.g., due to family or genetic links. In other embodiments, the subject may have no such predisposition.

[0184] The invention also relates to a diagnostic device, comprising means for measuring the expression level of biomarkers of a signature as defined herein.

[0185] In one embodiment, the diagnostic device comprises means for measuring the expression levels of at least two biomarkers, wherein the at least two biomarkers are (i) MUC-16 and (ii) either ANXA1 or ARHGAP1.

[0186] In some embodiments, the signature consists of three biomarkers, wherein the three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0187] Therefore, according to these embodiments, the diagnostic device comprises means for measuring the expression levels of three biomarkers selected from:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0188] In some embodiments, the signature consists of five biomarkers, wherein the five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0189] Therefore, according to this embodiment, the diagnostic device comprises means for measuring the expression levels of MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0190] In some embodiments, the diagnostic device comprises i) capture reagents specific for the biomarkers of the signature, and / or ii) detection reagents for detecting the biomarkers of the signature.

[0191] In some embodiments, the capture reagents are bound to plates, chips, beads, membranes or arrays.

[0192] In some embodiments, the detection reagents are enzymes, radioactive proteins, or fluorescent proteins.

[0193] In some embodiments, the expression level of biomarkers of a signature is determined by immunoassays, including ELISA, Western blot, immune-electrophoresis, immunostaining and Proximity Extension Assay (PEA); enzymatic activity; flow cytometry; spectrometry, including mass spectrometry; RT-PCR; RT-qPCR; Northern Blot; or hybridization techniques, and the diagnostic device comprises means appropriate for carrying out these techniques.

[0194] The present invention also relates to a kit for detecting endometriosis or for determining if a subject is affected with endometriosis, the kit comprising reagents for measuring, in a biological sample, the expression levels of the biomarkers of the signatures as defined herein and instructions for use.

[0195] In one embodiment, the kit comprises reagents for measuring the expression levels of at least two biomarkers, wherein the at least two biomarkers are (i) MUC-16 and (ii) either ANXA1 or ARHGAP1.

[0196] In some embodiments, the signature consists of three biomarkers, wherein the three biomarkers are:MUC-16, ANXA1 and EREG, orMUC-16, ANXA1 and ARSB, orMUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0197] Therefore, according to these embodiments, the kit comprises reagents for measuring the expression levels of three biomarkers selected from:• MUC- 16, ANXA1 and EREG, or• MUC- 16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

[0198] In some embodiments, the signature consists of five biomarkers, wherein the five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0199] Therefore, according to this embodiment, the kit comprises reagents for measuring the expression levels of MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0200] In some embodiments, the kit includes instructions for use of a set of reagents. For example, a kit can include instructions for measuring the expression level of biomarkers using methods such as an immunoassay, a protein-binding assay, an antibodybased assay, an antigen-binding protein based assay, a Proximity Extension Assay (PEA), a protein-based array, an enzyme-linked immunosorbent assay (ELISA), flow cytometry, a protein array, a blot, a Western blot, nephelometry, turbidimetry, chromatography, mass spectrometry, enzymatic activity, proximity extension assay (PEA), and an immunoassay selected from RIA, immunofluorescence, immunochemiluminescence, immunoelectrochemiluminescence, immunoelectrophoretic, a competitive immunoassay, immunoprecipitation, or any other well-known technique to the skilled man in the art for measuring the level of the polynucleotides encoding polypeptide biomarkers.

[0201] In some embodiments, the kit includes instructions for practicing the methods disclosed herein (e.g., methods for training or deploying a predictive model to predict anassessment of disease activity). These instructions can be present in the subject kits in a variety of forms.

[0202] In some embodiments, the kit as described hereinabove is for use in assaying whether a subject or patient has endometriosis.

[0203] The in vitro methods according to the invention, for assessing whether a subject has endometriosis may be used to diagnose endometriosis in a subject as a first step of a method to determine the adapted treatment and therapeutic method to be apply to cure and / or alleviate endometriosis.

[0204] In these embodiments, the combination of expression levels of biomarkers or signature can therefore be used to select a course of treatment for a subject or patient.

[0205] Therefore, in some embodiments, the present invention relates to methods of treating endometriosis based on the results of the in vitro method for assessing whether a subject has endometriosis. In some embodiments, the present invention relates to a method for treating a patient identified as having endometriosis using the in vitro method of the present invention. In some embodiments, the present invention relates to a method for treating a patient identified as being at risk of developing endometriosis using the in vitro method of the present invention. In some embodiments, the present invention relates to a method for providing an adapted care to a patient identified as having endometriosis using the in vitro method of the present invention.

[0206] The invention thus relates to a method for treating a subject affected with endometriosis or for providing an adapted care to a subject affected with endometriosis, wherein said method comprises the steps of: a) assessing whether a subject has endometriosis or whether a subject is at risk of developing endometriosis, using the in vitro method as disclosed herein, and c) treating, or providing an adapted care to, the subject identified as having endometriosis or at risk of developing endometriosis.

[0207] In one embodiment, a subject identified by the method as described hereinabove, as having endometriosis or as being at risk of developing endometriosis, is treated withan adapted care. In one embodiment, the adapted care is an endometriosis treatment adapted to the subject. In one embodiment, a subject that is identified with endometriosis is treated with non-hormonal treatments, including painkillers; hormonal treatments, including combined oral contraceptive (for example with non-steroidal anti-inflammatory drugs (NSAIDs)), or progestins combined with NSAIDs, or gonadotropin-releasing- hormone (GnRH) analogues combined with progestin, or aromatase inhibitor combined with progestin, or androgen analogue, or menopausal hormonal therapy (conjugated equine estrogen or estradiol with norethisterone); and / or with surgery.

[0208] The invention also relates to a method for treating a subject affected with endometriosis, wherein said method comprises the steps of a) measuring, in a biological sample of the subject, an expression level of biomarkers of a signature comprising at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP1, b) determining that the subject has endometriosis, and c) treating said subject for endometriosis.

[0209] In some embodiments, the method for treating a subject affected with endometriosis comprises measuring the expression level of at least three biomarkers, wherein the at least three biomarkers are:• MUC-16, and• ANXA1 or ARHGAP 1 , and• at least one biomarker selected from the group comprising or consisting of EREG, ANXA1, ARGHAP1, t-PA, PCSK9, ARSB, ESM-1 and TMPRSS15.

[0210] In some embodiments, the method for treating a subject affected with endometriosis comprises measuring the expression level of three biomarkers, wherein the three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, orMUC-16, ARHGAP1 and t-PA.

[0211] In some embodiments, the method for treating a subject affected with endometriosis comprises measuring the expression level of five biomarkers, wherein the five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

[0212] In some embodiments, the method for treating a subject affected with endometriosis comprises measuring the expression level of biomarkers of a signature as described herein.

[0213] The invention also relates to a computer-implemented method for determining if a subject is affected with endometriosis or for determining a personalized course of treatment in a subject affected with endometriosis, comprising the steps of: a) receiving an input level, being the quantification of the expression level of the biomarkers of a signature as described herein, preferably comprising at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP1, determined in a sample previously obtained from the subject, b) analyzing and transforming the input level by organizing and / or modifying each input level to derive a score via at least one machine learning algorithm, c) generating an output, wherein the output is the subject’s sample score, and d) providing either:- an endometriosis diagnosis of the subject based on the output; or- a personalized course or information to determine a personalized course of treatment for the subject based on the output.

[0214] In some embodiments, the course of treatment of the computer-implemented method comprises non-hormonal treatments, including combined oral contraceptive (for example with non-steroidal anti-inflammatory drugs (NSAIDs)), or progestins combined with NSAIDs, or gonadotropin-releasing-hormone (GnRH) analogues combined with progestin, or aromatase inhibitor combined with progestin, or androgen analogue, ormenopausal hormonal therapy (conjugated equine estrogen or estradiol with nor ethisterone); and / or with surgery.

[0215] In some embodiments, the at least one machine learning algorithm used in the computer-implemented method, was previously trained with at least one training dataset.

[0216] In some embodiments, the at least one machine learning algorithm used in the computer-implemented method, is a linear algorithm, or a complex algorithm.

[0217] In some embodiments, the at least one machine learning algorithm used in the computer-implemented method is a complex algorithm, and the score is determined using a machine learning algorithm selected from the group comprising or consisting of an artificial neural network (ANN), a perceptron algorithm, a deep neural network, a clustering algorithm, a k-nearest neighbors algorithm (k-NN), a decision tree algorithm, a random forest algorithm, a linear regression algorithm, a logistic regression algorithm, a linear discriminant analysis (LDA) algorithm, a quadratic discriminant analysis (QDA) algorithm, a support vector machine (SVM), a Bayes algorithm, a simple rule algorithm, a clustering algorithm, a meta-classifier algorithm, a Gaussian mixture model (GMM) algorithm, a nearest centroid algorithm, a gradient boosting algorithm (such as, e.g., an extreme gradient boosting [XG Boost] algorithm or an adaptative boosting [AdaBoost] algorithm), a linear mixed effects model algorithm, and any combination thereof.

[0218] The present invention also relates to a computer system for assaying endometriosis in a subject.

[0219] As used herein, the term “computer system” refers to any and all devices capable of storing and processing information and / or capable of using the stored information to control the behavior or execution of the device itself, regardless of whether such devices are electronic, mechanical, logical, or virtual in nature. The term “computer system” can refer to a single computer, but also to a plurality of computers working together to perform the function described as being performed on or by a computer system. A method implemented using a computer system is referred to as a “computer-implemented method”.

[0220] In one embodiment, the computer system according to the present invention comprises:- at least one processor, and- at least one computer-readable storage medium that stores code readable by the processor.

[0221] As used herein, the term “processor” is meant to include any integrated circuit or other electronic device capable of performing an operation on at least one instruction word, such as, e.g., executing instructions, codes, computer programs, and scripts which it accesses from a storage medium. However, the term “processor” should not be construed to be restricted to hardware capable of executing software and refers in a general way to a processing device, which can for example include a computer, a microprocessor, an integrated circuit, or a programmable logic device (PLD). The processor may also encompass one or more graphics processing units (GPU), whether exploited for computer graphics and image processing or other functions. Additionally, the instructions and / or data enabling to perform associated and / or resulting functionalities may be stored on any processor-readable medium, including, but not limited to, an integrated circuit, a hard disk, a magnetic tape (including floppy disk and zip diskette), an optical disc (including Blu-ray, compact disc and digital versatile disc), a flash memory (including memory card and USB flash drive) a random-access memory (RAM) (including dynamic and static RAM), a read-only memory (ROM) or a cache. Instructions may be in particular stored in hardware, software, firmware or in any combination thereof.

[0222] Examples of processors include, but are not limited to, central processing units (CPU), microprocessors, digital signal processors (DSPs), general purpose microprocessors, application specific integrated circuits (ASICs), field programmable logic arrays (FPGAs), and other equivalent integrated or discrete logic circuitry.

[0223] The present invention also relates to a computer program comprising software code readable by the processor adapted to perform, when executed by said processor, the computer-implemented methods as described herein.

[0224] The present invention also relates to a computer-readable storage medium comprising code readable by the processor which, when executed by said processor, causes the processor to carry out the steps of the computer-implemented methods as described herein.

[0225] Examples of computer-readable storage medium include, but are not limited to, an integrated circuit, a hard disk, a magnetic tape (including floppy disk and zip diskette), an optical disc (including Blu-ray, compact disc and digital versatile disc), a flash memory (including memory card and USB flash drive) a random-access memory (RAM) (including dynamic and static RAM), a read-only memory (ROM) or a cache.

[0226] In one embodiment, the computer-readable storage medium is a non-transitory computer-readable storage medium.

[0227] In one embodiment, the code stored on the computer-readable storage medium, when executed by the processor of the computer system, causes the processor to:- receive an input level, i.e., the quantification of the expression levels of biomarkers of the signatures of the invention determined in a sample previously obtained from the subject,- analyze and transform the input level by organizing and / or modifying each input level to derive a score via at least one machine learning algorithm,- generate an output, wherein the output is the subject sample score, and- provide a diagnosis of the subject based on the output; or- provide a personalized course or information to determine a personalized course of treatment for the subject based on the output.

[0228] In one embodiment, the code stored on the computer-readable storage medium, when executed by the processor of the computer system, causes the processor to:- receive an input level, i.e., the quantification of the expression level of the biomarkers of the signatures of the invention determined in a sample previously obtained from the subject,- analyze and transform the input level by organizing and / or modifying each input level to derive a probability score and / or a classification label via at least one machine learning algorithm,- generate an output, wherein the output is the subject sample score, and- provide a diagnosis of the subject based on the output; or- provide a personalized course or information to determine a personalized course of treatment for the subject based on the output.BRIEF DESCRIPTION OF THE DRAWINGS

[0229] Figures 1A-1I are graphic representations of ROC (Receiver Operating Characteristic) curves showing the performances of signatures with unique biomarker proteins on discovery dataset of 95 control patients and 243 endometriosis patients. Fig. 1A relates to MUC-16. Fig. IB relates to t-PA. Fig. 1C relates to TMPRSS15. Fig. ID relates to ESM-1. Fig. IE relates to ANXA1. Fig. IF relates to ARHGAP1. Fig. 1G relates to ARSB. Fig. 1H relates to EREG and Fig. II relates to PCSK9. The corresponding AUC (Area Under Curve) numerical values and confidence intervals are given in Table 1.

[0230] Figures 2A-2E are graphic representations of ROC (Receiver Operating Characteristic) curves showing the performances of five 3-protein signatures computed with a linear model. Fig. 2A relates to the combination of biomarkers EREG, MUC-16 and ANXA1. Fig. 2B relates to the combination of biomarkers MUC-16, PCSK9 and ARHGAP1. Fig. 2C relates to the combination of biomarkers t-PA, MUC-16 and ARHGAP1. Fig. 2D relates to the combination of biomarkers ARSB, MUC-16 and ANXA1. Fig. 2E relates to the combination of biomarkers TMPRSS15, MUC-16 andANXA1. ROC curves and confidence intervals computed on test dataset are represented in lighter gray and ROC curves and confidence intervals computed on verification dataset are represented in darker gray. The corresponding AUC (Area Under Curve) computed numerical values are given in Table 2 (test dataset) and Table 3 (verification dataset).

[0231] Figures 3A to 3E are graphic representations of ROC (Receiver Operating Characteristic) curves showing the performances of five 3-protein signatures computed with a complex model. Fig. 3A relates to the combination of biomarkers EREG, MUC- 16 and ANXA1. Fig. 3B relates to the combination of biomarkers MUC-16, PCSK9 and ARHGAP1. Fig. 3C relates to the combination of biomarkers t-PA, MUC-16 and ARHGAP1. Fig. 3D relates to the combination of biomarkers ARSB, MUC-16 and ANXA1. Fig. 3E relates to the combination of biomarkers TMPRSS15, MUC-16 and ANXA1. ROC curves and confidence intervals computed on test dataset are represented in lighter gray and ROC curves and confidence intervals computed on verification dataset are represented in darker gray. The corresponding AUC (Area Under Curve) computed numerical values are given in Table 4 (test dataset) and Table 5 (verification dataset).

[0232] Figure 4 is a graphic representation of ROC (Receiver Operating Characteristic) curve showing the performances of a 5-protein signature computed with a complex model. Figure 4 relates to the combination of biomarkers MUC-16, ANXA1, EREG, t- PA and ESM-1. ROC curves and confidence intervals computed on test dataset are represented in lighter gray and ROC curves and confidence intervals computed on verification dataset are represented in darker gray. The corresponding AUC (Area Under Curve) computed numerical values for test and verification datasets are given in Table 6.EXAMPLESExample 1 :Performances of five 3-protein signatures as analyzed according to a linear model.In this example, we aim to identify the best signatures of 3 proteins for endometriosis diagnosis using linear regression techniques on Endodiag’s cohort and protein measurements obtained by PEA.Materials and MethodsCohort

[0233] Samples from the collection set of specimens from Endodiag’s cohort were used. All patients followed inclusion criteria of Endodiag’s study, including an absence of endometriosis lesion during visual inspection of a surgeon for control patients and a dual confirmation of endometriosis by visual inspection of a surgeon and characterization of lesion by a pathologist for disease patients. No patients were in menopausal or postmenopausal status. Two distinct cohorts were used for Example 1 : One “discovery” cohort composed of 95 control patients and 243 patients with endometriosis (disease patients) and one “verification” cohort composed of 23 control patients and 60 patients with endometriosis (disease patients).Protein measurements

[0234] Sera were obtained from patients after centrifugation of blood 15min at 1120g within 2-4 hours after sampling and protein levels were measured with PEA (Proximity Extension Assay).Feature selection analysis

[0235] In order to find the best combination of proteins for discriminating between control and disease patients, a first selection of proteins of interest was necessary among the 900 measured proteins. This reduction of number of proteins is defined as feature selection, and it was processed following these steps:Compute individual statistic tests of control vs disease patient data for each protein (Pearson Correlation test and Chi-square 2 test) and register the p-values. These tests are direct descriptive methods of protein measurements,Compute multivariate methods based on the generation of models with all proteins and the selection of 20 proteins with the highest variance importance in the models. The models generated were of 4 types (Recurse Feature Elimination, Lasso, Random Forest and light Gradient Boosting Modeling). These multivariate methods provide additional information on complementarities between proteins but are heuristic and need cross- validation. Thus, for robustness purposes, each multivariate model was computed 100 times,Compute final feature selection score, defined as the number of times a protein was selected in each run and each method (with an additional weight of 100 for Pearson correlation and Chi-square 2 results to balance the difference of number of runs between methods).

[0236] The feature selection analysis was applied to the “discovery” cohort and the 23 proteins with the best feature selection final score were selected for signature generation. The individual diagnostic performances of the set of 23 selected proteins, shown in Table 1, vary from an area under the curve (AUC) of 0.43 up to an AUC of 0.72.

[0237] Generation of signatures was performed following these steps:Separating dataset from the “discovery” cohort into train (70%) and test (30%) datasets,Generating thousands of combinations of 3 proteins out of the 26 selected proteins,- For each 3-protein combination, performing logistic regression on train dataset to obtain the best linear model for diagnostic of endometriosis,Selecting the best signatures based on AUC value computed on test dataset.

[0238] Finally, the selected signatures were applied to dataset from the “verification” cohort in order to confirm diagnostic performances.Results

[0239] In Table 1, we show individual AUC values on discovery and verification datasets for each of the 23 proteins obtained from feature selection. In Table 2, we highlight the 5 best signatures of 3 proteins, i.e., the signatures that have an AUC of 0.75 or higher on the test dataset. We show in Table 3 that these performances are confirmed on an independent dataset, computed from protein measurements on the “verification” cohort, as for each of these 5 signatures, the AUC is better or similar on verification dataset than on the test dataset. For each signature, two thresholds were defined as to ensure a sensitivity of 80% and 95%, respectively, to consider positioning of signatures as both a replacement diagnostic test and a rule-out test for endometriosis. The sensitivity / specificity couple values computed on verification dataset with these thresholds are shown for each signature in Table 3. In addition to AUC values shown in Table 1, Figures 1A to II show the individual diagnostic performances obtained using individual biomarkers on the discovery dataset instead of signatures. We observe that with all the 5 highlighted signatures, the combination of three proteins improves the diagnostic power over using single biomarkers, as no protein from Table 1 has an individual AUC exceeding 0.72. Figures 2A to 2E shows ROC graphics of all 3-protein signatures computed on both test and verification datasets.Table 1: Individual diagnostic performances for each protein present in feature selection on discovery datasetTable 2: Diagnostic performances of selected signatures on test datasetTable 3 Diagnostic performances of selected signatures on verification datasetConclusion

[0240] We obtained 5 signatures, each consisting of a linear combination of 3 proteins, which were computed using a train dataset and tested for selection and performances on a test dataset. We obtained diagnostic performances significantly better than individual performances of each protein, as AUC values of the signatures were all above 0.75 while AUC values of each protein did not exceed 0.72. Moreover, all signature diagnostic performances observed on the test dataset were confirmed using an independent verification dataset.Example 2:Performances of five 3-proteins signatures computed with a complex model.

[0241] In this example, we aim to improve diagnostic performances of signatures from Example 1 by recomputing them using more complex mathematical models, such as deep neural networks, on Endodiag’s cohort and protein measurements obtained by PEA.Materials and MethodsCohort and protein measurements

[0242] The same discovery and verification cohorts and datasets from Example 1 were used for Example 2.Best signature computation using complex mathematical models

[0243] For each of the 5 signatures of Example 1, thousands of models were generated to differentiate control patients from disease patients on data from the “discovery” cohort using the H2O AutoML package in Python (https: / / www.h2o.ai / hybrid-cloud / ). AutoML trains a variety of algorithms (e.g., GBMs, Random Forests, Deep Neural Networks, GLMs), yielding a healthy amount of diversity across candidate models, which are exploited by stacked ensembles to produce a powerful final model (LeDell and Poirier, 2020, "H2o automl: Scalable automatic machine learning." Proceedings of the AutoML Workshop at ICML. Vol. 2020). The discovery dataset was divided into a 70% trainingset for generation of candidate models and a 30% testing set for computation of model performances. For each signature, the model with the best AUC on the test model, while having a superior AUC on the train model, was selected.

[0244] One model was selected for each signature of 3 proteins.

[0245] The models were then applied to the verification dataset to confirm the diagnostic performances estimated on the test dataset.Results

[0246] The diagnostic performances of the 3 -protein signatures with complex mathematical model are shown on Table 4 for test dataset. All signatures have better diagnostic performances with mathematical complex model than with a linear model, as AUC values in Table 3 do not exceed 0.8 and AUC values in Table 4 are all superior to 0.82.

[0247] However, it is well known that, although very powerful to obtain almost perfect performances on a specific dataset, complex mathematical model can be overfitted on the training dataset and have unrobust performances that drop significantly when applied on an independent verification dataset. Thus, we applied each complex signature to the verification dataset and obtained the performances shown in Table 5. For each signature, two thresholds were defined as to ensure a sensitivity of 80% and 95%, respectively, to consider positioning of signatures as both a replacement diagnostic test and a rule-out test for endometriosis. The sensitivity / specificity couple values computed on verification dataset with these thresholds are shown for each signature in Table 5. Figures 3A to 3E shows ROC graphics of all 3 -protein complex signatures computed on both test and verification datasets. The performances in Table 5, with AUC values ranging from 0.76 to 0.84, are lower than those on the test dataset on Table 4. However, for the last 3-protein combinations of Table 5, the verification performances of complex signature are similar or better than verification performances of linear signatures from Table 3 (0.84 vs 0.79 for combination of t-PA, MUC-16 and ANXA1, 0.77 vs 0.78 for combination of ARSB, MUC-16 and ANXA1, 0.76 vs 0.75 for combination of TMPRSS15, MUC-16 and ANXA1, respectively).

[0248] As for signatures with linear model, we observe that, with complex mathematical model, the combination of three proteins improves the diagnostic power over using single biomarkers, as no protein from Table 1 has an individual AUC exceeding 0.72.Table 4: Diagnostic performances of signatures with complex mathematical model on test datasetTable 5 Diagnostic performances of signatures with complex mathematical model on verification dataset Conclusion

[0249] Applying complex mathematical models to combinations of 3 proteins that were previously selected in Example 1 enabled us to obtain better diagnostic performances when computed on test dataset. When these complex models were applied to verification dataset, the diagnostic performances were confirmed as being similar or slightly betterthan results obtained on the linear models, which is a very good result considering the difficulty to maintain robust performances on a complex mathematical model. Consequently, the obtained diagnostic performances were significantly better than individual performances of each protein, as AUC values of the signatures were all above 0.75 while AUC values of each protein did not exceed 0.69.Example 3 :Performances of a 5-protein signature with a complex mathematical model.

[0250] In this example, we aim to identify the best signature of 5 proteins (among the 23 proteins selected during the feature selection mentioned in Example 1) for an endometriosis rule-out test using not only linear regression techniques but also more complex mathematical models, such as deep neural networks, on Endodiag’s cohort and protein measurements obtained by PEA.Materials and MethodsCohort and protein measurements

[0251] The same discovery and verification cohorts and datasets from Examples 1 and 2 were used for Example 3.Selection of best 5-protein combination using linear regression techniques

[0252] In order to select the best combination of 5 proteins to diagnose endometriosis, we used the same feature selection as in Example 1 and slightly modified the signature selection methodology, i.e., we followed these steps:Separating data into train (70%) and test (30%) datasets,Generating thousands of combinations of 5 proteins, from list of 23 proteins selected in Example 1,- For each combination, performing logistic regression on train dataset to obtain the best linear model for diagnostic of endometriosis using the proteins of the combination,Selecting the protein combination with the best AUC value computed on test dataset.

[0253] The combination that produced the best linear signature is composed of MUC- 16, ANXA1, EREG, t-PA and ESM-1 proteins. Using those proteins, we looked to improve the performances of this linear signature using complex mathematical models.Best signature computation using complex mathematical models

[0254] Thousands of models were generated to differentiate control patients from disease patients using data from the selected combination of 5 proteins using the H2O AutoML package in Python (https: / / www.h2o.ai / hybrid-cloud / ). AutoML trains a variety of algorithms (e.g., GBMs, Random Forests, Deep Neural Networks, GLMs), yielding a healthy amount of diversity across candidate models, which are exploited by stacked ensembles to produce a powerful final model [LeDell 2020], The data was divided into a 70% training set for generation of candidate models and a 30% testing set for computation of model performances. The model with the best AUC on the test model, while having a superior AUC on the train model, was selected. The model was then applied to the verification dataset to confirm the diagnostic performances estimated on the test dataset.Results

[0255] The best model obtained from the combination of MUC-16, ANXA1, EREG, t- PA and ESM-1 proteins has an AUC of 0.86 on test dataset as shown in Table 6 and Figure 4. A threshold was defined as to ensure a sensitivity of at least 95%, in order to estimate performances of a rule-out diagnostic test. The sensitivity / specificity values computed on discovery dataset with this threshold is of 0.95 / 0.53, respectively, as shown in Table 6. However, it is well known that, although very powerful to obtain almost perfect performances on a specific dataset, complex mathematical model can be overfitted on the training dataset and have unrobust performances that drop significantly when applied on an independent verification dataset. Thus, we applied the complex signature to the verification dataset and obtained au AUC of 0.81 and sensitivity / specificity values of 0.95 / 0.39, respectively. Consequently, despite a limited drop of diagnostic performances between test and verification dataset, we can assess that the performancesremain, even on a verification cohort, of great interest for a rule-out test of endometriosis. Indeed, a specificity of 0.39 for 0.95 sensitivity is significantly higher than any specificity value obtained for 0.95 sensitivity on any of the 3-protein signatures in Examples 1 and 2 (see Table 3 and Table 5). Moreover, as for signatures in Examples 1 and 2, we observe that the combination of five proteins improves the diagnostic power over using single biomarkers, as no protein from Table 1 has an individual AUC exceeding 0.72.Table 6: Performances of signature of MUC-16, ANXA1, EREG, t-PA and ESM-1Conclusion

[0256] We obtained the best 5-protein signature for a rule-out test of endometriosis by combining proteins MUC-16, ANXA1, EREG, t-PA and ESM-1 and computing a complex mathematical model. The good performances of using the signature as a rule- out test were confirmed on an independent verification dataset with AUC only dropping from 0.86 to 0.81 and specificity value for 0.95 sensitivity remaining above 0.39. Consequently, with this 5-protein signature, we confirmed diagnostic performances significantly better than individual performances of each protein and rule-out test performances better than all sensitivity / specificity values from 3-protein signatures shown in Examples 1 and 2.

Claims

CLAIMS1. An in vitro method for assessing whether a subject has endometriosis comprising a step of measuring, in a biological sample of the subject, the expression level of at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP1.

2. The in vitro method according to claim 1, wherein the method comprises measuring the expression level of at least three biomarkers, wherein the at least three biomarkers are:• MUC-16, and• ANXA1 or ARHGAP 1 , and• at least one biomarker selected from the group comprising or consisting of EREG, ANXA1, ARHGAP1, t-PA, PCSK9, ARSB, ESM-1 and TMPRSS15.

3. The in vitro method according to claim 1 or claim 2, wherein the method comprises measuring the expression levels of at least three biomarkers, wherein the at least three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-PA.

4. The in vitro method according to claim 1 or claim 2, wherein the method comprises measuring the expression levels of at least five biomarkers, wherein the at least five biomarkers are MUC-16, ANXA1, EREG, t-PA and ESM-1.

5. The in vitro method according to any one of claims 1 to 4, wherein the method comprises determining a score based on the combination of the measured expression levels using a linear model.

6. The in vitro method according to any one of claims 1 to 4, wherein the method comprises determining a score based on the combination of the measured expression levels using a complex model.

7. The in vitro method according to any one of claims 1 to 6, wherein the biological sample is selected from the group consisting of any tissue, cell, fluid, saliva, tears, urine, sweat, sputum, liquid biopsy, blood, serum, plasma, ascites, cyst fluid, vaginal fluid and cervico-vaginal fluid, preferably the biological sample is blood, serum or plasma.

8. The in vitro method according to any one of claims 1 to 7, wherein the level of the biomarkers is measured by immunoassays, including ELISA, Western blot, immune-electrophoresis, immunostaining and / or Proximity Extension Assay (PEA) ; enzymatic activity ; flow cytometry ; spectrometry, including mass spectrometry ; RT-PCR ; RT-qPCR ; Northern Blot ; or hybridization techniques.

9. A diagnostic device, comprising means for measuring the expression level of biomarkers of a signature comprising at least two biomarkers, wherein the at least two biomarkers are MUC-16 and either ANXA1 or ARHGAP1.

10. The diagnostic device according to claim 9, wherein the signatures comprises at least three biomarkers, wherein the at least three biomarkers are:• MUC- 16, ANXA1 and EREG, or• MUC-16, ANXA1 and ARSB, or• MUC-16, ANXA1 and TMPRSS15, or• MUC- 16, ARHGAP 1 and PC SK9, or• MUC- 16, ARHGAP 1 and t-P A.

11. The diagnostic device according to claim 9 or claim 10, wherein the signature comprises at least five biomarkers, wherein the at least five biomarkers are MUC- 16, ANXA1, EREG, t-PA and ESM-1.

12. The diagnostic device according to any one of claims 9 to 11, comprising i) capture reagents specific for the biomarkers of the signature, and ii) detection reagents for detecting the biomarkers of the signature.

13. The diagnostic device according to claim 12, wherein the capture reagents are bound to plates, chips, beads, membranes or arrays.

14. The diagnostic device according to claim 12 or claim 13, wherein the detection reagents are enzymes, radioactive proteins, or fluorescent proteins.

15. The diagnostic device according to any one of claims 12 to 14, wherein the means for measuring the expression level of biomarkers of a signature comprise or consist of means that may be used in immunoassays, including ELISA, Western blot, immune-electrophoresis, immunostaining or Proximity Extension Assay (PEA) ; enzymatic activity ; flow cytometry ; spectrometry, including mass spectrometry ; RT-PCR ; RT-qPCR ; Northern Blot ; or hybridization techniques.

Images