JP2022530636A - Classifier for detection of endometriosis - Google Patents

Abstract

As used herein, improved methods for the detection of endometriosis are described. In general, methods include, but are not limited to, applying machine learning algorithms to the miRNA level to detect, predict, diagnose, or monitor the presence or absence of endometriosis. The disclosure addresses, among other things, the need in the art for minimally invasive, accurate and more efficient methods for detecting, diagnosing and monitoring endometriosis.

Description

Mutual reference This application is filed on April 29, 2019, entitled "Classifier for the Detection of Endometriosis", US Provisional Applications Nos. 62 / 840, 300 and October 31, 2019. Claims the benefits of PCT / US2019 / 059006 entitled "Quantitative Algorithms for Endometriosis", both of which are incorporated by reference in their entirety.

Background Endometriosis is a common condition that affects women of adolescent and reproductive age. The disease is thought to be caused by the migration of endometrial tissue from its normal position, which lines the uterus, to other parts of the body, primarily into the abdominal cavity. It is common to affect the ovaries and intestinal walls. The migrated endometrial tissue, as well as its normal location, proliferates and diminishes with the menstrual cycle as a result of the action of ovarian hormones. Endometriosis can cause a number of symptoms including, but not limited to, abdominal pain, gastrointestinal upset, excessive bleeding, infertility, and menstrual disorder.

Women experiencing recurrent pelvic pain or infertility may be suspected of having endometriosis. Endometriosis, an inflammatory disorder in which endometrial cells proliferate outside the uterus, affects nearly 10% of women of reproductive age. It is found in 50-60% of women of reproductive age with chronic pelvic pain, and up to 50% of women with infertility. Despite its prevalence, endometriosis often remains undiagnosed for years. The average time from onset of symptoms to correct diagnosis can range from 5 to 10 years. The disease can be difficult to recognize, especially in the early stages, based on a description of the patient's symptoms, and a definitive diagnosis of the condition currently requires laparoscopy and surgical procedures. Laparoscopy is the current "gold standard" approach for visual confirmation of the pathology of endometriosis and collection of lesioned tissue for histological analysis.

Summary This disclosure addresses the need in the art for, among other things, a minimally invasive, accurate and more efficient method for detecting, diagnosing and monitoring endometriosis.

In one aspect, it is a method of detecting the presence or absence of endometriosis in a female subject, wherein (a) an expression profile of a panel of miRNAs associated with endometriosis in a fluid sample from a female subject. In the step of detecting, the panel of miRNA associated with endometriosis comprises miR-342 or miR451a, and (b) the expression of the panel of miRNA associated with endometriosis using a machine learning algorithm. A step applied to the profile, in which the machine learning algorithm has a materiality measure assigned to the properties of the miRNA, i. The importance scale is assigned to miR-342 and the importance scale assigned to miR-342 is selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. Greater than the importance measure assigned to at least one miRNA, or ii. The importance scale was assigned to miR-451a and the importance scale assigned to miR-451a was assigned to at least one miRNA selected from the group consisting of miR-3613, miR-125b, and let-7b. Methods are provided that include steps that are greater than the importance scale and (c) detect the presence or absence of endometriosis in a female subject using a machine learning algorithm. In some cases, the female subject has symptoms of endometriosis. In some cases, female subjects have symptoms of endometriosis selected from the group consisting of abdominal pain, gastrointestinal upset, excessive bleeding, infertility, and irregular menstruation. In some cases, the method comprises diagnosing endometriosis, monitoring endometriosis, predicting endometriosis, or predicting the risk of endometriosis. In some cases, the subject has not previously been diagnosed with endometriosis. In some cases, the subject was previously diagnosed with endometriosis and the method confirms the presence of endometriosis in a female subject. In some cases, the method further comprises the step of diagnosing endometriosis in a female subject if the presence of endometriosis is detected. In some cases, the method further comprises the step of predicting or monitoring endometriosis in a female subject if the presence of endometriosis is detected. In some cases, the method further comprises the step of treating the female subject with endometriosis if the presence of endometriosis is detected. In some cases, the method further comprises the step of determining that the female subject has a non-endometriosis condition if the absence of endometriosis is detected. In some cases, the method further comprises treating a female subject with a non-endometriotic condition if the absence of endometriosis is detected. In some cases, the importance measure assigned to miR-342 is on at least one miRNA selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. Greater than the assigned importance scale. In some cases, the importance scale assigned to miR-342 is smaller than the importance scale assigned to at least one other miRNA. In some cases, the importance measure assigned to miR-342 is on at least one miRNA selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. Greater than the assigned importance scale and greater than the importance scale assigned to at least one miRNA selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. small. In some cases, the importance scale assigned to miR-342 is smaller than the importance scale assigned to at least two other miRNAs. In some cases, the importance measure assigned to miR-342 is on at least two miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. Greater than the assigned importance scale. In some cases, the importance measure assigned to miR-342 is on at least three miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. Greater than the assigned importance scale. In some cases, the importance measure assigned to miR-342 is on at least 4 miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. Greater than the assigned importance scale. In some cases, the importance scale assigned to miR-342 is greater than the importance scale assigned to miR-150, miR-3613, miR-451a, let-7b, and miR-125b. In some cases, the importance scale assigned to miR-451a is greater than the importance scale assigned to at least one miRNA selected from the group consisting of miR-3613, miR-125b and let-7b. .. In some cases, the importance scale assigned to miR-451a is greater than the importance scale assigned to at least two miRNAs selected from the group consisting of miR-3613, miR-125b and let-7b. .. In some cases, the importance scale assigned to miR-451a is greater than the importance scale assigned to miR-3613, miR-125b, and let-7b. In some cases, the importance scales are such that the highest to lowest rankings of the importance scales are miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. Assigned. In some cases, fluid samples contain cells. In some cases, the body fluid sample is a cell-free sample. In some cases, the body fluid sample is a blood sample, plasma sample, saliva sample, or serum sample. In some cases, the miRNA panel is an acellular miRNA. In some cases, the miRNA panel is a cell-associated miRNA or an exosome-associated miRNA. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale ranking assigned to a miRNA property, with the highest to lowest ranking. , MiR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. In some cases, the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. In some cases, the machine learning algorithm is a random forest algorithm. In some cases, the method detects endometriosis in a female population with a specificity of greater than 80%. In some cases, the group of women is premenopausal women. In some cases, the female population includes women with smooth myomas, cystadenomas, chronic pelvic infections, teratomas, endometrial tumors, or paraovarian cysts. In some cases, the female population includes women with stage I / II endometriosis. In some cases, the female population includes women with stage III / IV endometriosis, or women with all four stages of endometriosis (stages I / II / III / IV). .. In some cases, the female population includes women who received hormone therapy within 3 months from the date the sample was obtained, or women in any phase of the menstrual cycle. In some cases, the female population comprises a cohort containing at least 100 women. In some cases, the female population includes a cohort containing at least 1000 women. In some cases, machine learning algorithms are trained with expression data from at least 100 samples. In some cases, machine learning algorithms are trained with expression data from at least 500 samples. In some cases, machine learning algorithms are trained with expression data from at least 1000 samples. In some cases, machine learning algorithms are trained for a population of women, including women with stage I-IV endometriosis. In some cases, the method has an AUC for detecting endometriosis greater than 0.85 in a female population. In some cases, the method has an AUC for detecting endometriosis greater than 0.90 in a female population. In some cases, the method has an AUC for detecting endometriosis greater than 0.92 in a female population. In some cases, the method detects endometriosis in a female population with a specificity of greater than 80%. In some cases, the method detects endometriosis in a female population with a specificity greater than 85%. In some cases, the method detects endometriosis in a female population with a specificity greater than 90%. In some cases, the method detects endometriosis in a female population with a specificity greater than 92%. In some cases, the method detects endometriosis in a female population with a specificity greater than 95%. In some cases, the method detects endometriosis in a female population with a sensitivity greater than 80%. In some cases, the method detects endometriosis in a female population with a sensitivity greater than 85%. In some cases, the method detects endometriosis in a female population with a sensitivity greater than 90%. In some cases, the method detects endometriosis in a female population with a specificity of greater than 90% and a sensitivity of less than 85%. In some cases, the method detects endometriosis in a female population with a specificity greater than 95% and a sensitivity of less than 85%. In some cases, the method detects endometriosis in a female population with less than 90% sensitivity. In some cases, the method detects endometriosis in a female population with less than 85% sensitivity. In some cases, the method detects endometriosis in a female population with less than 80% sensitivity. In some cases, the method further comprises treating the female subject with a procedure without surgery if the absence of endometriosis is detected. In some cases, the method further comprises the step of treating the female subject if the presence of endometriosis is detected, the treatment being hormonal treatment, surgery, laparoscopic surgery, statins, non-steroidal. Sexual anti-inflammatory drugs (NSAIDs), oral contraceptives, progestin, gonadotrophin release (GnRH) agonists, GnRH antagonists, androgen, antiprogesterone, selective estrogen receptor modulator (SERM), selective progesterone receptor modulator (SPRM), atorvastatin , Serivastatin, fluvastatin, robastatin, mevastatin, pitabastatin, pravastatin, rosvasstatin, simvastatin, parasetamol, COX-2 inhibitor, or aspirin. In some cases, miRNA expression levels are detected by quantitative real-time polymerase chain reaction (RT-PCR), microarrays, sequencing, or next-generation sequencing.

In another aspect, a method of classifying endometriosis in a female subject, (a) a step of obtaining a body fluid sample containing miRNA, wherein the body fluid sample is from a female subject. , (B) A step of performing a quantitative real-time polymerase chain reaction, microarray or sequencing assay on a set of miRNAs in a body fluid sample, wherein the set of miRNAs is associated with two or more endometriosis. Steps containing more different miRNAs and (c) the amount of two or more different miRNAs associated with endometriosis in the biological sample are compared to the amount of control RNA in the body fluid sample. A female subject is placed in utero by inputting the steps of determining normalized miRNA levels of two or more different miRNAs in, and (d) the normalized miRNA levels into a trained algorithm. In the step of classifying endometriosis as positive or negative, the trained algorithm has two or more different miRNA importance rankings, and the trained algorithm is the recipient behavioral characteristic (ROC). Optimized for higher specificity than sensitivity by selecting the optimal cutoff point in the voting percentage distribution of two or more different miRNAs associated with curves or endometriosis, steps And computerized reports identifying female subjects as either positive or negative for endometriosis, based on the steps of classifying female subjects as positive or negative for endometriosis in (e) and (d). Methods are provided, including steps to output on the screen. In some cases, the trained algorithm is optimized to detect endometriosis with a specificity greater than 80% in a female population. In some cases, the trained algorithm is optimized to detect endometriosis in a female population with a specificity of greater than 90% and a sensitivity of less than 85%. In some cases, the trained algorithm detects stage I / II endometriosis with a specificity greater than 90% or greater than 95% in a female population. In some cases, the method has an AUC for detecting endometriosis greater than 0.85 in a female population. In some cases, the method has an AUC for detecting endometriosis greater than 0.9 in a female population. In some cases, the method has an AUC for detecting endometriosis greater than 0.92 in a female population. In some cases, the method detects endometriosis in a female population with a specificity greater than 85%. In some cases, the method detects endometriosis in a female population with a specificity greater than 90%. In some cases, the method detects endometriosis in a female population with a specificity greater than 95%. In some cases, the method detects endometriosis in a female population with a sensitivity greater than 80%. In some cases, the method detects endometriosis in a female population with a sensitivity greater than 90%. In some cases, the method detects endometriosis in a female population with a sensitivity of less than 90%, less than 85%, or less than 80%. In some cases, a group of women includes at least 100 women. In some cases, a population of at least 100 women includes women with leiomyoma. In some cases, the method has an area under the curve (AUC) value greater than 0.85, regardless of the stage of endometriosis or hormonal treatment. In some cases, the method further comprises the step of treating the female subject with endometriosis after the report identifies the female subject as negative for endometriosis. In some cases, the method further comprises the step of treating the female subject with a non-endometriotic condition after the report identifies the female subject as negative for endometriosis. In some cases, the method repeated steps (a)-(e) for additional fluid samples obtained at least 3 months later, when the report identified the female subject as negative for endometriosis. Further included. In some cases, the trained algorithm is a importance measure assigned to at least one miRNA selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. Assign a greater importance scale to miR-342. In some cases, the importance scale assigned to miR-342 is smaller than the importance scale assigned to at least one other miRNA. In some cases, the importance measure assigned to miR-342 is on at least two miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. Greater than the assigned importance scale. In some cases, the importance measure assigned to miR-342 is on at least three miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. Greater than the assigned importance scale. In some cases, the importance measure assigned to miR-342 is on at least 4 miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. Greater than the assigned importance scale. In some cases, the importance scale assigned to miR-342 is greater than the importance scale assigned to miR-150, miR-3613, miR-451a, let-7b, and miR-125b. In some cases, the importance scale assigned to miR-451a is greater than the importance scale assigned to at least one miRNA selected from the group consisting of miR-3613, miR-125b and let-7b. .. In some cases, the importance scale assigned to miR-451a is greater than the importance scale assigned to at least two miRNAs selected from the group consisting of miR-3613, miR-125b and let-7b. .. In some cases, the importance scale assigned to miR-451a is greater than the importance scale assigned to miR-3613, miR-125b, and let-7b. In some cases, the importance scales are such that the highest to lowest rankings of the importance scales are miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. Assigned to. In some cases, fluid samples contain cells. In some cases, the body fluid sample is a cell-free sample. In some cases, the body fluid sample is a blood sample, plasma sample, saliva sample, or serum sample. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale ranking assigned to a miRNA property, with the highest to lowest ranking. , MiR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. In some cases, the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. In some cases, the machine learning algorithm is a random forest algorithm. In some cases, the group of women is premenopausal women. In some cases, the female population includes, in any combination, women with leiomyomas, cystadenomas, chronic pelvic infections, teratomas, endometrial tumors, or paraovarian cysts. In some cases, the female population includes women with stage I / II endometriosis. In some cases, the female population includes women with stage III / IV endometriosis, or women with all four stages of endometriosis (stages I / II / III / IV). .. In some cases, the female population includes women who received hormone therapy within 3 months from the date the sample was obtained, or women in any phase of the menstrual cycle. In some cases, the female population comprises a cohort containing at least 100 women. In some cases, the female population includes a cohort containing at least 500 women. In some cases, the female population includes a cohort containing at least 1000 women. In some cases, machine learning algorithms are trained with expression data from at least 100 samples. In some cases, machine learning algorithms are trained with expression data from at least 1000 samples. In some cases, machine learning algorithms are trained for a population of women, including women with stage I-IV endometriosis. In some cases, the method further comprises treating the female subject with a procedure without surgery if the absence of endometriosis is detected. In some cases, the method further comprises the step of treating the female subject if the presence of endometriosis is detected, the treatment being hormonal treatment, surgery, laparoscopic surgery, statins, non-steroidal. Sexual anti-inflammatory drugs (NSAIDs), oral contraceptives, progestin, gonadotrophin release (GnRH) agonists, GnRH antagonists, androgen, antiprogesterone, selective estrogen receptor modulator (SERM), selective progesterone receptor modulator (SPRM), atorvastatin , Serivastatin, fluvastatin, robastatin, mevastatin, pitabastatin, pravastatin, rosvasstatin, simvastatin, parasetamol, COX-2 inhibitor, or aspirin. In some cases, the method comprises diagnosing endometriosis, monitoring endometriosis, predicting endometriosis, or predicting the risk of endometriosis. In some cases, the subject has not previously been diagnosed with endometriosis.

Yet another embodiment is a method of diagnosing and treating endometriosis in a female subject, wherein (a) an expression profile of a panel of miRNAs associated with endometriosis in a saliva sample from a female subject. The steps to detect, the panel of miRNAs associated with endometriosis, include miR-125b and at least one other miRNA, and (b) the machine learning algorithm miRNAs associated with endometriosis. In the steps applied to the expression profile of the panel, the machine learning algorithm has a materiality scale assigned to the properties of miRNA, and the importance scale of miR-125b is miR-150, miR-3613, miR. Steps that are greater than the importance scale of -451a, let-7b, or miR-342, and (c) the step of diagnosing endometriosis in a female subject using a machine learning algorithm, and (d). Methods are provided that include treating endometrial disease diagnosed in a female subject with treatment for endometrial disease. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale assigned to the properties of the miRNA, the importance scale of miR-125b. , MiR-150, let-7b, miR-451a, or miR-3613. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale assigned to the properties of the miRNA, the importance scale of miR-125b. , MiR-150, let-7b, miR-451a, or miR-3613. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale assigned to the properties of the miRNA, the importance scale of miR-125b. , MiR-150, let-7b, miR-451a, or miR-3613. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale assigned to the properties of the miRNA, the importance scale of miR-125b. , MiR-150, let-7b, miR-451a, and miR-3613.

In yet another embodiment, a method of characterizing a female subject as having endometriosis, (a) a step of obtaining a body fluid sample containing miRNA, wherein the body fluid sample is from a female subject. , And (b) a step of performing a quantitative real-time polymerase chain reaction, a microarray assay, or sequencing of a set of miRNAs in a body fluid sample, the two sets of miRNAs associated with endometriosis. The amount of the step and (c) two or more different miRNAs associated with endometriosis in the biological sample, including or more different miRNAs, is compared to the amount in the control RNA, body fluid. By inputting the steps to determine the normalized miRNA levels of two or more different miRNAs in the sample and (d) the normalized miRNA levels into the trained algorithm, the female subject is uterine. A step in classifying endometriosis as positive or negative, where the trained algorithm cuts optimally in the recipient behavioral characteristics (ROC) curves of two or more different miRNAs associated with endometriosis. In steps and (e) (d), the off-point is optimized for sensitivity of at least 80% and the trained algorithm has two or more different miRNA importance rankings. Includes stepping on a computer screen to output a report identifying female subjects as either positive or negative for endometriosis, based on the steps to classify female subjects as positive or negative for endometriosis. , The method is provided.

In yet another embodiment, a method of characterizing a female subject as having or not having endometriosis, (a) a step of obtaining a body fluid sample containing miRNA, which is a body fluid. The sample is from a female subject with abdominal pain, gastrointestinal upset, excessive bleeding, infertility or irregular menstruation, and the subject has not previously been diagnosed with endometriosis, step and ( b) A step in performing a quantitative real-time polymerase chain reaction or sequencing of a set of miRNAs in a body fluid sample, wherein the set of miRNAs comprises two or more different miRNAs associated with endometriosis. , Step and (c) the amount of two or more different miRNAs associated with endometriosis in the biological sample compared to the amount in the control RNA, two or more in the body fluid sample. Classify female subjects as positive or negative for endometriosis by entering the steps of determining normalized miRNA levels for many different miRNAs and (d) the normalized miRNA levels into a trained algorithm. At least by having the trained algorithm select the optimal cutoff point in the recipient behavioral characteristics (ROC) curves of two or more different miRNAs associated with endometriosis. Optimized for 80% specificity and trained algorithms have two or more different miRNA importance rankings, steps and positive or negative endometriosis in (e) (d) Based on the steps to classify female subjects as, methods are provided that include printing a report on a computer screen that identifies the female subject as either positive or negative for endometriosis. In some cases, the trained algorithm calculates a voting score indicating the likelihood of having endometriosis. In some cases, the trained algorithm classifies a female subject as having endometriosis if the voting score is greater than 36%. In some cases, the method has an area under the curve (AUC) value greater than 0.85 for stage I / II endometriosis. In some cases, the method has a subcurve area (AUC) value greater than 0.85 for distinguishing between endometriosis and leiomyoma. In some cases, the method has an area under the curve (AUC) value greater than 0.85, regardless of the stage of endometriosis or hormonal treatment. In some cases, the method is optimized for specificity greater than 90% and sensitivity less than 85%.

Yet another embodiment is a method of detecting endometriosis in a female subject, wherein (a) the expression profile of a panel of miRNAs associated with endometriosis is detected in a sample containing miRNAs from the subject. And (b) applying the machine learning algorithm to the expression profile to detect endometriosis in a sample from a subject, wherein the machine learning algorithm is (i) miR-342, miR-. 451a, and miR-3613, (ii) miR-342, miR-451a, miR-3613, and miR-125b, (iii) miR-342, miR-451a, miR-3613, miR-125b, and let-7b. , (Iv) miR-342, miR-451a, let-7b, and miR-125b, (v) miR-342, miR-451a, let-7b, and miR-3613, (vi) miR-342, miR- 451a, and let-7b, (vii) miR-125b, miR-150, miR-342, miR-451a, and let-7b, and (viii) miR-125b, miR-150, miR-342, miR-3613. , MiR-451a, and let-7b are trained for a group of miRNA properties selected from the group, including steps. In some cases, the method further comprises the step of obtaining a sample containing miRNA from the subject prior to (a). In some cases, the sample is a saliva sample. In some cases, the sample is a serum sample. In some cases, the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. In some cases, the method has an AUC for detecting endometriosis greater than 0.85 in a female population. In some cases, the group of women is premenopausal women. In some cases, the female population is premenopausal and older than 18 years of age. In some cases, the female population is premenopausal and under 49 years of age. In some cases, the female population is negative for severe anemia, hyperplasia, polyps, and malignant tumors. In some cases, the female population includes women who received hormone therapy within 3 months from the date the sample was obtained. In some cases, hormone therapy comprises oral contraceptives or GnRH agonists. In some cases, machine learning algorithms are trained with expression data from at least 100 samples. In some cases, machine learning algorithms are trained on a population of women with surgically confirmed endometriosis. In some cases, machine learning algorithms are trained for a population of women, including women with stage I or II endometriosis. In some cases, machine learning algorithms are trained for a population of women, including women with stage I-IV endometriosis. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale assigned to the properties of the miRNA, which is the importance scale of miR-342. , MiR-150, let-7b, or at least one of miR-125. In some cases, the step of applying a machine learning algorithm to an expression profile involves applying a machine learning algorithm with a specific importance scale assigned to the properties of the miRNA, the importance scale of miR-125b. , MiR-150, let-7b, miR-451a, or miR-3613. In some cases, the step of applying a machine learning algorithm to an expression profile is to train the machine learning algorithm for a group of (viii) miRNA properties and a specific importance measure assigned to the miRNA properties. The highest to lowest rankings of miRNA properties, including the step of applying the machine learning algorithm having, are miR-125b, let-7b, miR-3613, miR-150, miR-342, and miR-451a. .. In some cases, the steps of applying a machine learning algorithm to an expression profile are the step of training the machine learning algorithm for a group of (viii) miRNA properties and the specific importance scale ranking assigned to the miRNA properties. The ranking from highest to lowest is miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150, including the step of applying a machine learning algorithm having. In some cases, the step of applying a machine learning algorithm to an expression profile is to train the machine learning algorithm for a group of (viii) miRNA properties and a specific importance measure assigned to the miRNA properties. The importance scale ranking of the properties of miRNA, including the step of applying the machine learning algorithm having, follows any column in Table 9.

In yet another embodiment, (a) the step of storing information related to the condition of a female patient in a standardized format on multiple network-based non-transient storage devices, and (b) graphically by at least one user. At least one user has been updated in a step that provides remote access to users over the network so that information related to the female patient's condition can be updated in real time via the user interface. Using a step that provides information in the form of a miRNA expression profile from a female patient, and (c) an application of a machine learning algorithm, a miRNA expression profile from a female patient can be presented by a content server to the female patient's uterus. The steps to convert to the possibility of having endometriosis, (d) the step of remembering the possibility that the female patient has endometriosis, and (e) the updated information is stored. Each time, the content server automatically generates a message containing the possibility that the female patient has endometriosis, and (f) each user has a female patient with endometriosis. Methods are provided, including the step of sending a message in real time to all of the users over the computer network for immediate access to the possibilities. In some cases, the machine learning algorithms are (i) miR-342, miR-451a, and miR-3613, (ii) miR-342, miR-451a, miR-3613, and miR-125b, (iii). miR-342, miR-451a, miR-3613, miR-125b, and let-7b, (iv) miR-342, miR-451a, let-7b, and miR-125b, (v) miR-342, miR- 451a, let-7b, and miR-3613, (vi) miR-342, miR-451a, and let-7b, (vii) miR-125b, miR-150, miR-342, miR-451a, and let-7b. , And (viii) trained on a group of miRNA properties selected from the group consisting of miR-125b, miR-150, miR-342, miR-3613, miR-451a, and let-7b. In some cases, the method further comprises the step of obtaining a sample containing miRNA from the subject. In some cases, the sample is a blood sample, plasma sample, or serum sample. In some cases, the sample is a saliva sample. In some cases, the sample is a serum sample. In some cases, the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. In some cases, the method has an AUC for detecting endometriosis greater than 0.85 in a female population. In some cases, the female population includes women who received hormone therapy within 3 months from the date the sample was obtained. In some cases, hormone therapy comprises oral contraceptives or GnRH agonists. In some cases, the sample is a cell-free serum sample. In some cases, the sample is a cell-free saliva sample. In some cases, the method further comprises treating the subject to treat endometriosis based on the likelihood reported in (e). In some cases, treatments include hormonal treatments, statins, or non-steroidal anti-inflammatory drugs (NSAIDs). In some cases, hormonal treatments include oral contraceptives, progestins, GnRH agonists, GnRH antagonists, androgens, antiprogesterones, SERMs, or SPRMs. In some cases, statins include atorvastatin, cerivastatin, fluvastatin, lovastatin, mevastatin, pitavastatin, pravastatin, rosuvastatin, or simvastatin. In some cases, NSAIDs include paracetamol, COX-2 inhibitors, or aspirin.

In some embodiments, the present disclosure is a method of detecting endometriosis in a female subject, wherein (a) an expression profile of a panel of miRNAs associated with endometriosis in a sample from the subject. The detection step and (b) the step of applying the machine learning algorithm to the expression profile to detect endometriosis in a sample from a subject, wherein the machine learning algorithm is (i) miR-342, miR. -451a, and miR-3613, (ii) miR-342, miR-451a, miR-3613, miR-125b, (iii) miR-342, miR-451a, miR-3613, miR-125b, let-7b, (Iv) miR-342, miR-451a, let-7b, miR-125b, (v) miR-342, miR-451a, let-7b, miR-3613, (vi) miR-342, miR-451a, let -7b, (vii) miR-125b, miR-150, miR-342, miR-451a, let-7b, and (viii) miR-125b, miR-150, miR-342, miR-3613, miR-451a, And a method comprising a step, which is trained on a group of miRNA properties selected from the group consisting of let-7b. In some embodiments, the method comprises obtaining a sample containing miRNA from a subject. In some embodiments, the sample is a blood sample. In some embodiments, the sample is a serum sample. In some embodiments, the sample is a plasma sample. In some embodiments, the sample is a cell-free sample or a cell-free blood, plasma, or serum sample. In some embodiments, the sample is a blood sample from which cells have been removed by centrifugation after collection in a collection tube by venous puncture without additional additives. In some embodiments, the blood sample is centrifuged or filtered to remove the cells. In some embodiments, the sample is a blood, plasma, or serum sample. In some embodiments, the sample is a urine sample. In some embodiments, the sample is a body fluid sample. In some embodiments, the body fluids are sweat, saliva, tears, urine, blood, plasma, serum, vaginal fluid, cervico-vaginal fluid, whole blood, menstrual effluent. , Menstrual blood, spinal fluid, plumonary fluid, or sputum. In some embodiments, the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. In some embodiments, the algorithm is a random forest algorithm. In some embodiments, the method has an AUC for detecting endometriosis greater than 0.85 in a female population. In some embodiments, the female population is premenopausal women. In some embodiments, the female population is premenopausal and above 18 years of age. In some embodiments, the female population is premenopausal and is less than 49 years old. In some embodiments, the female population is negative for severe anemia, hyperplasia, polyps, and malignant tumors. In some embodiments, the female population comprises women who received hormone therapy within 3 months from the date the sample was obtained. In some embodiments, the hormone therapy comprises an oral contraceptive and / or a GnRH agonist. In some embodiments, the algorithm is trained with expression data from at least 100 samples. In some embodiments, the algorithm is trained with expression data from at least 50 samples. In some embodiments, the algorithm is trained with expression data from at least 200 samples. In some embodiments, the algorithm is trained with expression data from at least 500 samples or at least 1000 samples. In some embodiments, the algorithm is trained on a population of women with surgically confirmed endometriosis. In some embodiments, the algorithm is trained for a population of women, including women with stage I or II endometriosis. In some embodiments, the algorithm is trained for a population of women, including women with stage I-IV endometriosis. In some embodiments, the step of applying a machine learning algorithm (eg, a random forest algorithm) to an expression profile comprises assigning a importance scale to the properties of miRNA, and the importance scale of miR-342 is miR-. Larger than that of 150, let-7b, or miR-125. In some embodiments, the step of applying a machine learning algorithm to an expression profile comprises assigning a importance scale to the properties of miRNA, and the importance scale of miR-125b is miR-150, let-7b, or. It is larger than that of miR-125. In some embodiments, applying the machine learning algorithm to an expression profile comprises training the machine learning algorithm for the characteristics of (viii) and assigning an importance scale to the characteristics of the miRNA, the characteristics of the miRNA. The highest to lowest importance scale rankings are miR-125b, let-7b, miR-2613, miR-150, miR-342, and miR-451a. In some embodiments, the step of applying the machine learning algorithm to the expression profile includes the step of training the machine learning algorithm for the characteristics of (viii) and the step of assigning a measure of importance to the characteristics of the miRNA of the miRNA. The highest-to-lowest importance measure (eg, characteristic importance) ranking of a characteristic is miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. In some embodiments, the step of applying the machine learning algorithm to the expression profile comprises training the machine learning algorithm for the characteristics of (viii) and assigning an importance scale to the characteristics of the miRNA, the characteristics of the miRNA. The importance scale ranking of is according to Table 9. In some embodiments, the method uses a supervised learning algorithm with the characteristic importance assigned in Table 9 to detect, diagnose, or diagnose the risk of endometriosis in a female subject. Further includes steps to evaluate. In some embodiments, the method further comprises performing endometriosis treatment to treat endometriosis detected or diagnosed in a female subject.

In some embodiments, the disclosure comprises (a) storing information related to the condition of a female patient in a standardized format on multiple network-based non-transient storage devices, and (b) at least one person. A step that provides remote access to a user over a network so that the user can update information related to the female patient's condition in real time via a graphical user interface, with at least one user. Using a step and application of (c) a machine learning algorithm to provide updated information in the form of a miRNA expression profile from a female patient, a miRNA expression profile from a female patient can be obtained by a content server. Steps to convert to the possibility that a female patient has endometriosis, (d) a step to remember the possibility that a female patient has endometriosis, and (e) updated information. Each time the content server remembers, the content server automatically generates a message containing the possibility that the female patient has endometriosis, and (f) the user tells the female patient that the female patient has endometrial membrane. It provides methods, including the step of sending a message in real time to at least some of the users over a computer network for immediate access to the possibility of having a illness. In some embodiments, the machine learning algorithms are (i) miR-342, miR-451a, and miR-3613, (ii) miR-342, miR-451a, miR-3613, miR-125b, (iii). miR-342, miR-451a, miR-3613, miR-125b, let-7b, (iv) miR-342, miR-451a, let-7b, miR-125b, (v) miR-342, miR-451a, let-7b, miR-3613, (vi) miR-342, miR-451a, let-7b, (vii) miR-125b, miR-150, miR-342, miR-451a, let-7b, and (viii) Trained on a group of miRNA properties selected from the group consisting of miR-125b, miR-150, miR-342, miR-3613, miR-451a, and let-7b. In some embodiments, the method comprises obtaining a sample containing miRNA from a subject. In some embodiments, the sample is a saliva sample. In some embodiments, the sample is a serum sample. In some embodiments, the sample is a blood, plasma, or serum sample. In some embodiments, the sample is a urine sample. In some embodiments, the sample is a body fluid sample. In some embodiments, the body fluids are sweat, saliva, tears, urine, blood, plasma, serum, vaginal fluid, cervical-vaginal fluid, whole blood, menstrual effluent, menstrual blood, spinal fluid, lung fluid, Or sputum. In some embodiments, the machine learning algorithms are a random forest algorithm, a k-nearest neighbor algorithm (KNN), a support vector machine (SVM), and a naive bays. In some embodiments, the machine learning algorithm is a random forest algorithm. In some embodiments, the method has an AUC for detecting endometriosis greater than 0.85 in a female population. In some embodiments, the female population comprises women who received hormone therapy within 3 months from the date the sample was obtained. In some embodiments, the hormone therapy comprises an oral contraceptive or a GnRH agonist. In some embodiments, the method further comprises a step of detecting, diagnosing, or assessing the risk of endometriosis in a female subject based on a measure of importance. In some embodiments, the method further comprises performing endometriosis treatment to treat endometriosis detected or diagnosed in a female subject.

Incorporation by Reference All publications, patents, and patent applications referred to herein are the same as each individual publication, patent, or patent application indicated to be specifically and individually incorporated by reference. To some extent, the whole is incorporated herein by reference.

The novel properties of the invention are described in detail in the appended claims. A better understanding of the properties and advantages of the invention can be obtained by referring to the following detailed description, as well as the accompanying drawings, which describe exemplary embodiments and utilize the principles of the invention. There will be.

FIG. 1 shows spraying and boxplots of miRNA expression in the sera of control vs. endometriosis patients, showing that all markers are effective in identifying endometriosis to some extent. Data are from 6 miRNAs (miR-125b, miR-451a, miR-3613, miR-150, miR-342, and let-7b) normalized to the level of the small nuclear RNA gene U6. Indicates the expression level. The data are plotted with the median indicated by a line and the interquartile range (IQR) indicated by a box. Whiskers and outliers are displayed according to Tukey's method, where the whiskers are less than or equal to 1.5 times the IQR (25th percentile minus IQR or 75th percentile plus IQR). It is plotted and points outside this range are plotted individually. ^* P <0.05, ^*** p <0.001, Mann-Whitney U test followed by Bonferroni correction for multiple comparisons.

FIG. 2 shows a spray / boxplot of miRNA expression during the proliferative or secretory phase, showing that the expression of 6 miRNAs does not change significantly during the proliferative vs. secretory phase of the menstrual cycle. The data are divided by phase of the menstrual cycle at the time of serum sampling and show the level of miRNA expression in the control subject normalized to the level of the small nuclear RNA gene U6. The data are plotted with the median indicated by a line and the interquartile range (IQR) indicated by a box. No significant difference was found (p> 0.05, Mann-Whitney U test).

FIG. 3 shows a spray / boxplot of miRNA expression with or without hormone treatment, showing that the level of miRNA expression in the study does not significantly differ depending on the hormone administration status to the subject. The data show miRNA expression levels in subjects with endometriosis analyzed by the presence or absence of hormone treatment (HT). Levels were normalized to the level of the small nuclear RNA gene U6. The data are plotted with the median indicated by a line and the interquartile range (IQR) indicated by a box. No significant difference was found (p> 0.05, Mann-Whitney U test).

FIG. 4 shows a spray / boxplot of miRNA expression by rASRM staging, where all markers distinguish between control and severe (III / IV) patients, while the ability to distinguish between subsets (eg, control). Vs. I / II or I / II vs. III / IV) show different. The data show miRNA expression levels in endometriosis subjects divided by the stage of endometriosis (I / II according to rASRM guidelines, minimal / mild; III / IV, moderate / severe). Levels were normalized to the level of the small nuclear RNA gene U6. The data are plotted with the median indicated by a line and the interquartile range (IQR) indicated by a box. Groups were compared using the Kruskal-Wallis test (nonparametric one-way ANOVA) and pairwise means of each subgroup were compared using Dunn's multiple comparison test. ^* P <0.05, ^** p <0.01, ^*** p <0.001.

FIG. 5 shows a receiver operating characteristic (ROC) curve showing the performance of the classifier algorithm in an independent data set. The graph is an analysis of a random forest model using 6 miRNA biomarkers (miR-125b-5p, miR-150-5p, miR-342-3p, miR-451a, miR-3613-5p, let-7b). Is shown. The model was derived in this (n = 100) dataset (“training”) and is retroactive Cosar et al. Fertil Steril. 2016 Aug; 106 (2): 402-9. Doi: 10.1016 / j.fertnstert .2016.04.013 (“Cosar”) Tested against a dataset (“test”) of study (n = 48).

FIG. 6 shows an example of a computer system for performing the methods described herein.

Figure 7 shows the distribution of voting percentages from the RF model of the retrospective dataset (Cosar); black bars for subjects with surgically defined endometriosis, white bars with endometriosis. Shows a histogram showing subjects who do not. The 43% diagnostic threshold (cutoff) shown by the vertical dashed line provides 96% specificity and 83% sensitivity for the RF model of this dataset.

Detailed Description Overview Given that delays in the diagnosis of endometriosis can exacerbate women's costs and negative experiences, earlier identification and treatment of the disease can be of great benefit to women. It can prevent complications of potentially advanced diseases, including infertility, and reduce the financial burden of untreated endometriosis. Laparoscopy is the current clinical recommendation for a definitive diagnosis of endometriosis, but it poses a cost and risk to the patient and is accurate when performed without histological confirmation. There is a limit to the identification. Surgical interventions are rarely given to patients with early-stage illness or unclear symptoms. Imaging methods such as ultrasound may be appropriate for detecting advanced endometriosis or endometrial species, but are insufficient to detect most diseases such as common peritoneal lesions or adhesions. be.

Analyzing a combination of biomarkers rather than a single biomarker can improve the detection of endometriosis. Although the serum cancer antigen CA-125 has been used as a circular marker of disease, elevated levels of CA-125 primarily reflect the stage of progression of endometriosis and other diseases (eg, uterine fibroids, ovaries). It does not have sufficient diagnostic sensitivity or specificity because it also increases in cancer and inflammatory disorders in the uterus. A systematic review of 141 studies of 122 blood biomarkers, including endometrial antibodies, reported that none of the biomarkers studied in the review met the criteria for triage or replacement diagnostic trials. (See Nisenblat et al. Cochrane Database Syst Rev. 2016 (5): Cd012179).

The present disclosure characterizes and monitors samples from subjects who have symptoms of endometriosis, have endometriosis, are at risk of having endometriosis, or are suspected of having endometriosis. And provide a new way to analyze. The present disclosure also provides methods for detecting, diagnosing, monitoring, and / or predicting such subjects, as well as methods for treating such subjects. In general, the methods provided herein include the detection or quantification of biomarkers, in particular non-coding RNAs (eg, miRNAs) in a sample from a subject. In some cases, the methods provided herein involve the application of machine learning algorithms.

Definitions As used herein, the term "cell-free", when used with respect to nucleic acids, refers to nucleic acids that were not associated with cells when they were obtained from the body. For example, nucleic acids can be present in body fluids, such as blood or saliva, in a cellular state in that the nucleic acids are not associated with cells. However, cell-free nucleic acids may have originally associated with cells, such as endometrial cells before entering the bloodstream or other body fluids. In contrast, in the body, nucleic acids that are simply associated with cells are generally not considered "cell-free." For example, nucleic acids extracted directly from cells are generally not considered the term "cell-free" as used herein.

As used herein, a "cell-free sample" generally refers to a biological sample, in particular the level at which the miRNA level in which cells are absent or determined is not in the cell portion but in the liquid portion of the sample. Refers to a biological fluid sample that is present in a very small amount. In some examples, the cell-free portion of the sample is obtained by centrifugation, filtration, fractionation column, or other method. In some cases, body fluids can be cell-free in nature. Typically, a cell-free body fluid sample does not contain intact cells; however, it can contain cell fragments, exosomes, or cell debris. In some cases, the sample is processed or used immediately after sampling; in some cases, the sample is stored for later use. Body fluid samples can be stored using any suitable storage method known in the art, for example, the sample can be frozen at about −20 ° C. to about −70 ° C.

As used herein, a "cell-free serum" sample is generally a serum sample that has been processed approximately immediately after collection to avoid cell destruction by coagulation mechanisms and the like.

Traditional notation is used herein to describe the polynucleotide sequence, where the left-hand end of the single-stranded polynucleotide sequence is the 5'end and the left-hand side of the double-stranded polynucleotide sequence is. It is called the 5'direction.

Terms such as "subject," "patient," and "individual" are used interchangeably herein to refer to any animal suitable for the methods described herein. In certain non-limiting embodiments, the patient, subject, or individual is a human. In some cases, the methods provided herein may include cells from such a subject, patient, or individual. In some cases, the method may be performed in vitro or in situ, at least in part.

As used herein, "microRNA" or "miRNA" generally describes a small non-coding RNA molecule that is approximately 15 to about 50 nucleotides in length, preferably 17 to 23 nucleotides in length. , For example, can play a role in regulating gene expression through a process called RNA interference (RNAi). RNAi describes a phenomenon in which the presence of an RNA sequence that is complementary or antisense to a sequence in the target gene's messenger RNA (mRNA) results in inhibition of target gene expression. miRNAs are generally processed from hairpin precursors (premiRNAs) of about 70 or more nucleotides derived from primary transcripts (premiRNAs) through sequential cleavage by RNAse III enzymes. miRBase is available at www. mirbase. A comprehensive microRNA database at org. Generally, miRNA genes are transcribed into precursors or premiRNAs, which are processed into mature miRNAs. PremiRNAs generally occur in the form of hairpins, which contain a 5'arm (or side) connected to a loop and then the loop connected to a 3'arm (or side). Processing of the precursor miRNA can result in the formation of two mature forms of miRNA, including the 5p form derived from the 5'side or arm of the precursor miRNA loop, and the precursor miRNA hairpin 3 '3p forms derived from the side or arm are included.

As used herein, "or" may refer to "and", "or", or "and / or" and may be used both exclusively and inclusively. For example, the term "A or B" can refer to "A or B", "A not B", "B not A", and "A and B". In some cases, the context may define a particular meaning.

As used herein, the term "one (a)" may refer to the plural or singular. In other words, "one (a)" usually refers to "one or more". Similarly, the term "an" can refer to the singular or plural.

As used herein, RNA (s) is used interchangeably and may refer to a single RNA or multiple RNAs. Similarly, miRNAs (s) are used interchangeably and can refer to a single miRNA or multiple miRNAs.

As used herein, "non-coding RNA" (ncRNA) generally refers to an endogenous RNA molecule that is not translated into protein in the cell. Exemplary types of ncRNAs include transfer RNA (tRNA), ribosome RNA (rRNA), microRNA (miRNA), piRNA, snoRNA, snRNA, exRNA, scRNA, and long ncRNAs (eg, Xist and HOTAIR). .. In some embodiments, the endometriosis test described herein determines the level of one or more ncRNAs that are not miRNAs, in addition to the specific microRNAs described herein. May include doing.

As used herein, the term "random forest" builds a large number of decision trees during training and outputs classes that are the form of individual tree classes (classifications) or average predictions (regression). Refers to ensemble learning methods for classification, regression, and other tasks that work with. The implementation of random forests for data classification has been described in various contexts.

As used herein, the term "support vector machine" (SVM) refers to a supervised learning method that analyzes the data used for classification and regression analysis. Given a set of training examples, each marked as belonging to one or the other of the two categories, the SVM training algorithm builds a model that assigns new examples to one category or the other and assigns it to a non-stochastic binary. Use a linear classifier (although there are methods such as Platt scaling for using SVMs in stochastic classification situations). The SVM model is a representation of an example as a point in space, mapping individual categories of examples so that they are separated by as wide a clear gap as possible. New examples are then mapped within that same space and they are expected to belong to the category based on the side of the corresponding gap.

As used herein, the term "sequencing" generally refers to methods and techniques for sequencing nucleotide bases in one or more polynucleotides. A polynucleotide can be, for example, a nucleic acid molecule such as deoxyribonucleic acid (DNA) or ribonucleic acid (RNA), including a variant or derivative thereof (eg, single-stranded DNA). Sequencing is currently limited to, but not limited to, a sequencing system by Illumina®, Pacific Biosciences (PacBio®), Oxford Nanopore®, or Life Technologies (Ion Torrent®). It can be implemented in the various systems available. Alternatively, or in addition, sequencing can be performed using nucleic acid amplification, polymerase chain reaction (PCR) (eg, digital PCR, quantitative PCR, or real-time PCR), or isothermal amplification. Such a system can provide multiple raw genetic data corresponding to the genetic information of a subject (eg, human) produced by the system from a sample provided by the subject. In some examples, such systems provide sequencing reads (also referred to herein as "reads"). The read may contain a set of nucleobases corresponding to the sequence of the sequenced nucleobase.

The terms "adaptor" and "adapter" may be used interchangeably. The adapter or tag can be linked to the polynucleotide sequence by any approach, including ligation, hybridization, primer extension, or other approaches. Adapters or tags can be added to the nucleic acid to facilitate attachment to the sequencing flow cell, facilitate binding of the sequencing primer to the nucleic acid, or count individual copies of the nucleic acid sequence in the mixture (eg,). Used as a unique molecular identifier).
In various embodiments, the amount of target analyte can be normalized to a normalized control using a "delta CT method" or "ΔCT method" that involves calculating ΔCT. In certain embodiments, ΔCT is a quantitative nucleic acid used to detect a normalized control from the CT of a quantitative nucleic acid detection assay (eg, qPCR assay) used to detect a target analyte. Calculated by subtracting the CT (cycle threshold) of a detection assay (eg, qPCR assay). In certain embodiments, the multiple difference in the amount of the normalized control and the target analyte is calculated from ΔCT-in certain embodiments, the multiple difference in the amount of the normalized control and the target analyte is given in Equation 2. Calculated from ΔCT according to ^−ΔCT . In some embodiments, the normalized control is a housekeeping nucleic acid (eg, DNA or RNA encoding a housekeeping polypeptide). In some cases, the normalized control is the small nuclear RNA gene U6.

Subjects The methods and compositions described herein are applicable to human and non-human subjects, including veterinary subjects. Preferred subjects are "patients", that is, living humans receiving medical care for a disease or condition (eg, endometriosis), or suspected of having such a disease or condition, or such. Those at risk of having a disease or condition. This includes individuals who have been investigated for signs of pathology (eg, endometriosis) but whose disease has not been determined.

Preferred patients or subjects for the methods and compositions described herein are female patients of pubertal or postpubertal age, premenopausal, peri-pubertal, menopausal, or postmenopausal (intrauterine). Because endometriosis can persist after menopause). Therefore, in general, the methods and compositions provided herein may be useful to female subjects of a wide range of ages, generally above 10 years or above 18 years. Often, the subject is under 49 years of age. In some cases, the subject is premenopausal. In some cases, the subject is premenopausal and is older than 18 years of age. In some cases, the subject is premenopausal and is less than 49 years old. The subject can be in any phase of the menstrual cycle, eg, the luteal phase or the proliferative phase.

In some cases, the subject may be at risk of having endometriosis. Subjects at risk of having endometriosis may have, for example, a family history of endometriosis, symptoms of endometriosis, or a past history of endometriosis. Endometriosis can be any stage of endometriosis. In some cases, the subject has or is suspected of having stage I, II, III, or IV endometriosis. In some cases, the subject has or is suspected of having endometrial tumor. In some cases, the subject has endometriosis of any stage (eg, stages I-IV), but the methods provided herein do not specify the stage of the subject. Endometriosis is detected as a general condition. In some cases, the subject has an early stage endometriosis. In some cases, the subject has stage I / II endometriosis. In some cases, the subject has stage III / IV endometriosis.

In some cases, the subject may be suspected of having endometriosis. Such subjects may not show symptoms of endometriosis. However, in other cases, such subjects may have dysmenorrhea, pain associated with bowel movements or urination, deep dyspareunia, chronic lower abdominal pain, chronic low back pain, adnexal mass, infertility, or excessive bleeding. May show symptoms of endometriosis. In some cases, the results of previous or concurrent studies of endometriosis may suggest that the subject has endometriosis. In some cases, the subject is suspected of having endometriosis due to multiple factors. For example, a subject may be suspected of having endometriosis due to the presence of symptoms in an overall clinical setting consistent with endometriosis.

In some cases, the subject has or may be suspected of having a non-endometriotic condition. In general, as used herein, the term "non-endometriotic condition" refers to an abnormal reproductive condition that is not endometriosis. Non-limiting examples of non-endometrial conditions are uterine myoma, smooth myoma, cyst, cutaneous cyst, serous cystadenoma, cystadenoma, ovarian cyst, mucosal cystadenoma, pelvic infection, malformation, and /. Or including paraovarian cysts. In some cases, such subjects may include dysmenorrhea, pain associated with bowel movements or urination, deep dyspareunia, chronic lower abdominal pain, chronic low back pain, adnexal mass, infertility, or excessive bleeding. May show symptoms of endometriosis. In some cases, the non-endometriotic condition is benign. In some cases, the non-endometriotic condition is malignant.

In some cases, the subject may be receiving hormonal treatment. Hormone treatments include GnRH (gonadotrophin-releasing hormone) agonists (with or without estrogen / progesterone replacement therapy or tibolone treatment) and antagonists, levonolgestrel-releasing intrauterine devices (eg, Milena), danazole, antiprogesterone, guestlinone, aromatase. It may include an inhibitor, a selective estrogen receptor modulator (SERM), or a selective progesterone receptor modulator (SPRM).

Sample The sample is preferably a body fluid sample. Body fluids include sweat, saliva, tears, urine, blood, plasma, serum, vaginal fluid, cervical-vaginal fluid, whole blood, menstrual effluent (eg, menstrual blood), It can be spinal fluid, plumonary blood, sputum, or any other bodily fluid. In a preferred embodiment, the sample is a saliva or menstrual effluent (eg, menstrual blood) sample. In some cases, the sample contains white blood cells (WBC). In some embodiments, the sample is a plasma sample. In some embodiments, the sample is a cell-free, or cell-free blood, plasma, or serum sample. In some embodiments, the sample is taken into a collection tube by venous puncture without additives (eg, without anticoagulant or coagulant) and then centrifuged (eg, 2500 xg) to isolate the cells. It is a removed blood sample. In some cases, the sample (eg, blood, serum, plasma, etc.) is subjected to centrifugation, filtration, fractionation, or other methods. In some cases, the sample contains peripheral blood mononuclear cells (PBMC), and in some cases, the sample contains peripheral blood lymphocytes (PBL). As used herein, the term "saliva" does not include sputum. This is because sputum belongs to a mucous or mucous sample. In some embodiments, saliva, peripheral blood samples or menstrual effluents (eg, menstrual blood) may be separated into cellular and non-cellular fractions by suitable methods (eg, centrifugation, filtration). can. In some embodiments, nucleic acids (eg, miRNAs or ncRNAs) can be extracted from cellular (eg, cell-containing) or non-cell (or cell-free) fractions. In some embodiments, the analysis described herein for miRNA or ncRNA expression contains cells of any of the samples (eg, blood, plasma, serum, saliva, menstrual blood, menstrual effluent, etc.). This can be done for fractions or non-cellular fractions.

In some cases, the sample comprises tissue, eg, tissue from a biopsy. In some cases, the tissue is endometrial tissue.

In some embodiments, the sample comprises cell-free non-coding RNA (eg, cell-free miRNA). In some cases, the sample comprises purified or extracted non-coding RNA (eg, miRNA). In some embodiments, the sample comprises non-coding RNA encapsulated in exosomes (eg, miRNA). In some embodiments, the sample comprises non-coding RNA (eg, miRNA) encapsulated in cells (eg, by leukocytes).

Taking Samples As used herein, "obtaining a sample" includes having a sample obtained (eg, from a third party that obtained the sample directly from the subject). It involves obtaining a sample directly or indirectly. In some embodiments, the sample is subsequently taken from the subject by the same organization (eg, a testing laboratory) that obtains biomarker data from the sample. In some embodiments, the sample is received (eg, by a testing laboratory) from another entity from which the sample was taken from the subject (eg, a doctor, nurse, blood collector, or other healthcare provider). In some embodiments, the sample is taken from the subject by a healthcare professional under the direction of another entity (eg, a test laboratory) and subsequently into the entity (eg, a test laboratory). Provided. In some embodiments, the sample is taken at home by the subject or the subject's caregiver (eg, family, home caregiver), followed by an organization (eg, test study) that obtains biomarker data from the sample. It is provided to the room).

In some embodiments, a test sample of blood can be obtained from the subject. In some embodiments, the blood sample is a peripheral blood sample. In some embodiments, the blood sample is a whole blood sample. In some embodiments, the sample is a blood sample and comprises whole blood, peripheral blood, serum, plasma, PBL, PBMC, T cells, CD4 T cells, CD8 T cells, or macrophages. Blood samples can be obtained by minimally invasive methods, such as blood sampling. Blood samples can be obtained by venipuncture.

In some embodiments, a test sample of saliva can be obtained from the subject. Methods of obtaining a saliva sample may include, but are not limited to, expelling (eg, exhaling) the subject's mouth, aspiration, or removal by swab or other collection tool. Methods for extracting RNA molecules from saliva are described, for example, in Pandhita, Pet al. Clin Chem. It can be found at 2013 Jul; 59 (7): 1118-22. A wide range of saliva collection and recovery devices (collecting the sample in a clean manner and providing stabilization of nucleic acids in the sample) are available as kits and are available as kits from commercially available suppliers such as DNA Genotek (eg Oragene). -RNA, as well as the products described in US2011102002A1 and WO2008040126A1), and Norgen Biotek, which are suitable for use with the methods of the present disclosure. Such kits are suitable for use by the patient individually or with minimal assistance by a healthcare provider (eg, a physician).

After collection, the sample (eg, saliva) is organic by the addition of antimicrobial agents (eg, Normocin, sodium azide), RNase inhibitors (eg, polyvinylsulfonic acid, RNassin®, RNaseOUT ™). Stabilized by disruption in solution (eg, trizol, phenol-chloroform, phenol-chloroform-isoamyl alcohol) or by detergent in detergent in combination with broad-spectrum protease (eg, SDS combined with proteinase K). Can be made to.

RNA (eg, miRNA) Expression Profiling The methods, kits, and systems disclosed herein specifically detect RNA (eg, ncRNA, miRNA) in a biological sample to determine an expression profile. It may include profiling, or quantification. In some cases, RNA (eg, miRNA, ncRNA) can be isolated from a biological sample. In some cases, RNA (eg, miRNA, ncRNA) can be isolated from a cell-free source.

In some cases, expression levels are determined by hybridization-based methods such as Northern blots, Southern blots, molecular beacons, molecular inversion probes, or microarray hybridization. In some cases, hybridization-based methods include hybridization of probes to target RNAs (eg, ncRNAs, miRNAs), or hybridization of multiple different probes to different target RNAs (eg, ncRNAs, miRNAs). include.

In some cases, expression levels are determined by the amplification process or by the polymerase chain reaction (PCR). In some cases, expression levels are determined by quantitative PCR, real-time PCR, reverse transcriptase PCR, or other types of PCR. PCR may include the use of probes such as the TaqMan probe.

In some cases, the expression level is determined by sequencing. Examples of sequencing are Sanger sequencing, high throughput sequencing, pyrosequencing, ligation sequencing, synthetic sequencing, hybridization sequencing, RNA-Seq (Illumina), Digital Gene Expression (Helicos), next generation. Sequencing, Synthetic Single Molecular Sequencing (SMSS) (Helicos), Large Parallel Sequencing, Clonal Single Molecular Array (Solexa), Shotgun Sequencing, Maxim-Gilbert Sequencing, Primer Walking, or any of them. Can include combinations of. In some embodiments, sequencing may initially include reverse transcriptase and / or PCR amplification steps to increase the abundance of miRNAs analyzed or to add suitable sequencing adapters. Sequencing is not limited to, but is limited to, such as by Illumina®, Pacific Biosciences (PacBio®), Oxford Nanopore®, or Life Technologies (Ion Torrent®), etc. It can be implemented in the various systems available. Alternatively, or in addition, sequencing can be performed using nucleic acid amplification, polymerase chain reaction (PCR) (eg, digital PCR, quantitative PCR, or real-time PCR), or isothermal amplification. Examples of sequencing can include Sanger sequencing, Next Generation sequencing, and RNA sequencing.

Biomarker RNA (eg miRNA, ncRNA)
The methods and compositions herein are one or more ncRNAs associated with endometriosis (eg, miRNAs) for detecting, predicting, or monitoring the severity of endometriosis. ) (Eg, detection of the presence or absence of at least one ncRNA), or measurement of the level of one or more miRNAs or ncRNAs associated with endometriosis from a patient sample. In some cases, detection of more than one miRNA further comprises applying an algorithm trained to expression levels of more than one miRNA or ncRNA associated with endometriosis. Trained algorithms suitable for application include any of the classification algorithms described herein. In some cases, the trained algorithm is a machine learning algorithm. In some cases, the machine learning algorithms are random forest algorithms, k-nearest neighbor algorithms (KNN), support vector machines (SVMs), and naive bays.

In some cases, the miRNA classifier set used is any one of miR-125b, miR-150, miR-342, miR-3613, miR-451a, and let-7b. Include in numbers or combinations. In some cases, the miRNA classifier sets used are miR-342, miR-451, and miR-3613. In some cases, the miRNA classifier sets used are miR-342, miR-451, miR-3613, and miR-125. In some cases, the miRNA classifier sets used are miR-342, miR-451, miR-3613, miR-125, and let-7. In some cases, the miRNA classifier sets used are miR-342, miR-451, let-7, and miR-125. In some cases, the miRNA classifier sets used are miR-342, miR-451, let-7, and miR-3613. In some cases, the miRNA classifier sets used are miR-342, miR-451, and let-7. In some cases, the miRNA classifier sets used are miR-125, miR-150, miR-342, miR-451, and let-7. In some cases, the miRNA classifier sets used are miR-125, miR-150, miR-342, miR-3613, miR-451, and let-7. In some cases, the miRNA classifier sets used are miR-342, miR-451a, and miR-3613. In some cases, the miRNA classifier sets used are miR-342, miR-451a, miR-3613, and miR-125b. In some cases, the miRNA classifier sets used are miR-342, miR-451a, miR-3613, miR-125b, and let-7b. In some cases, the miRNA classifier sets used are miR-342, miR-451a, let-7b, and miR-125b. In some cases, the miRNA classifier sets used are miR-342, miR-451a, let-7b, and miR-3613. In some cases, the miRNA classifier sets used are miR-342, miR-451a, and let-7b. In some cases, the miRNA classifier sets used are miR-125b, miR-150, miR-342, miR-451a, and let-7b. In some cases, the miRNA classifier sets used are miR-125b, miR-150, miR-342, miR-3613, miR-451a, and let-7b. In some cases, the miRNA classifier sets used are miR-342-3p, miR-451a, and miR-3613-5p. In some cases, the miRNA classifier sets used are miR-342-3p, miR-451a, miR-3613-5p, and miR-125b-5p. In some cases, the miRNA classifier sets used are miR-342-3p, miR-451a, miR-3613-5p, miR-125b-5p, and let-7b-5p. In some cases, the miRNA classifier sets used are miR-342-3p, miR-451a, let-7b-5p, and miR-125b-5p. In some cases, the miRNA classifier sets used are miR-342-3p, miR-451a, let-7b-5p, and miR-3613-5p. In some cases, the miRNA classifier sets used are miR-342-3p, miR-451a, and let-7b-5p. In some cases, the miRNA classifier sets used are miR-125b-5p, miR-150-5p, miR-342-3p, miR-451a, and let-7b-5p. In some cases, the miRNA classifier sets used are miR-125b-5p, miR-150-5p, miR-342-3p, miR-3613-5p, miR-451a, and let-7b-5p. Is.

Treatment In some embodiments, the methods of the present disclosure include assigning or giving treatment to a patient who has, is at risk of developing, or is suspected of having endometriosis. .. Appropriate treatment can be assigned or administered to a patient suffering from endometriosis by detecting the clinical condition of the patient using the classifiers or miRNAs described herein. These treatments may include, but are not limited to, hormonal therapy, chemotherapy, immunotherapy, and surgical procedures. Similarly, the methods of the present disclosure can be used to assign or administer treatment to patients with reduced fertility due to endometriosis. In this way, by determining the extent to which a patient's fertility is reduced, appropriate treatment can be assigned or administered through the detection of the biomarkers found herein. Related procedures include, but are not limited to, hormonal therapy, chemotherapy, immunotherapy, and surgical procedures.

In some embodiments, a classifier that outputs the level of one or more miRNAs (eg, circulating miRNAs) in the patient's biological sample, or the clinical status determined from it, is the patient's disease (eg, intrauterine). Used to detect, diagnose, monitor, or predict endometriosis). In some embodiments, the level of one or more miRNAs (eg, circulating miRNAs) in the test sample obtained from the patient is compared to the level from the reference sample obtained from the patient at a previous time point. obtain. In some cases the patient is clinically monitored; in some cases the patient can act as her own baseline control. For example, changes in the level of one or more miRNAs (eg, circulating miRNAs) that increase or decrease may indicate the development of endometriosis in a patient. In some embodiments, the test sample is obtained at multiple time points. In these embodiments, measuring the level of one or more miRNAs (eg, circulating miRNAs) in a test sample is an indicator of whether the patient has or is at risk of having endometriosis. offer. In some cases, the level of one or more miRNAs (eg, circulating miRNAs) in a test sample obtained from a patient may be a level from a reference sample from a different patient, or a composite or mean value of a different patient. Will be compared.

In some embodiments, the level of one or more miRNAs (eg, circulating miRNAs) in a patient's biological sample (or a classifier that outputs the clinical status determined from it) is the efficacy or disease of the treatment. Used to monitor the prognosis of. In some embodiments, the level of one or more miRNAs (eg, circulating miRNAs) in a test sample obtained from a treated patient is from a reference sample obtained from that patient prior to the start of treatment. Can be compared to levels. Clinical monitoring of the procedure typically requires the patient to act as his or her own baseline control. In some embodiments, the test sample is obtained at multiple time points after application of the treatment. In these embodiments, measuring the level of one or more miRNAs (eg, circulating miRNAs) in a test sample provides an indicator of the extent and duration of the in vivo effect of the treatment. In some cases, the level of one or more miRNAs (eg, circulating miRNAs) in the test sample obtained from the treated patient may be from a reference sample from different patients, or a combination of different patients or Compared to the average value.

Measurement of biomarker levels (or classification status) may make it possible to monitor the course of treatment of the disease. The effectiveness of the treatment regimen for disease can be monitored by detecting an effective amount of one or more biomarkers in a sample obtained over time from a subject and comparing the amounts of the detected biomarkers. .. For example, the first sample can be obtained before the subject undergoes treatment, and one or more subsequent samples are taken after or during the treatment of the subject. Changes in biomarker levels throughout the sample may provide an indicator of the effectiveness of the therapy.

In some embodiments, the disclosure provides a method for monitoring the level of miRNA in response to a treatment (or clinical condition such as endometriosis or non-endometriosis). For example, in certain embodiments, the present disclosure provides a method of determining the effectiveness of treatment in a subject by measuring the level of one or more miRNAs described herein. In some embodiments, the level of one or more miRNAs can be measured over time, and the level at one time point after the start of treatment is compared to the level at another time point after the start of treatment. .. In some embodiments, the level of one or more miRNAs can be measured over time and the level at a point in time after the start of treatment is compared to the level before the start of treatment.

The present disclosure provides therapies for the treatment or prevention of endometriosis (eg, drugs, surgery). A non-limiting list of known methods and materials for the treatment of endometriosis includes, but is not limited to, analgesics, hormonal treatments, chemotherapeutic treatments, and surgical treatments. In some cases, a patient or subject in which endometriosis is detected by the methods provided herein may undergo additional tests or procedures to confirm endometriosis. For example, a patient or subject undergoes surgery to further diagnose or characterize endometriosis, eg, the stage of endometriosis, or to treat endometriosis (eg, laparoscopic surgery). May be good. Analgesics used in the treatment of endometriosis include simple analgesics such as paracetamol, COX-2 inhibitors, aspirin, and other non-steroidal anti-inflammatory drugs well known in the art, as well as morphine and codeine. , Oxycodone, and other numbing analgesics such as those well known in the art. Hormonal treatments include oral contraceptives, progestins (didrogesterone, medroxyprogesterone acetate, medroxyprogesterone depotacetate, norethisterone, levonorgestrel, etc.), progesterone and progesterone-like substances, GnRH agonists (leuprelin, busererin, goseleline, histreline, deslorerin). , Nafaleline, and tryptreline, synthetic androgens such as androgen and danazole, GnRH agonists (eg, eragolix) (with or without estrogen / progesterone replacement therapy or tiboron treatment), levonorgestrel-releasing intrauterine devices (eg, Milena). ), Danazole, antiprogesterone, guestlinone, selective estrogen receptor modulator (SERM), or selective progesterone receptor modulator (SPRM), as well as, but not limited to, aromatase inhibitors. Surgical procedures include, but are not limited to, laparoscopic surgery, hysterectomy, and oophorectomy. Other procedures particularly well suited for use in the present disclosure are well known in the art. In some embodiments, the patient can be treated with statins including, but not limited to, atorvastatin, cerivastatin, fluvastatin, lovastatin, mevastatin, pitavastatin, pravastatin, rosuvastatin, and simvastatin.

The present disclosure provides therapies for the treatment or prevention of non-endometriotic conditions (eg, drugs, surgery). A non-limiting list of known methods and materials for the treatment of non-endometriosis includes, but is not limited to, analgesics, antibiotics, chemotherapy, and surgical procedures. Analgesics include simple analgesics such as paracetamol, COX-2 inhibitors, aspirin, and other non-steroidal anti-inflammatory drugs well known in the art, as well as morphine, codeine, oxycodone, and well known in the art. It may include both narcotic analgesics such as others. Surgical procedures include, but are not limited to, laparoscopic surgery, hysterectomy, and oophorectomy. In some cases, the patient or subject may undergo additional testing or monitoring following a negative result from the endometriosis test or assay described herein. Such subjects or patients may be tested, for example, for diseases or conditions of non-endometriosis, such as cancer, benign cysts, uterine fibroids, and the like. In some cases, following a negative result, the subject or patient waits a reasonable amount of time (eg, longer than 1 week, longer than 1 month, longer than 6 months) for endometriosis. Can be tested again. In some cases, such tests may be by the methods provided herein. In some cases, depending on the outcome of negative endometriosis, the subject or patient may avoid undergoing surgical intervention (eg, laparoscopic surgery).

Sample Classification The presence and / or expression data of miRNAs according to the methods of the present disclosure (eg, data representing the presence or absence or level of a particular miRNA from a sample from a patient) can be used to classify the sample. .. For example, a sample may be classified or predicted to be from a) a patient with endometriosis, or b) from a patient without endometriosis. Many statistical classification techniques are suitable as an approach for performing such classifications. In a supervised learning approach, a group of samples from two or more groups (eg, endometriosis or otherwise) is analyzed or processed by a statistical classification method. The presence / absence or expression level of miRNA can be used as the basis for a classifier that distinguishes between two or more groups. The new sample can then be analyzed or processed so that the classifier can associate the new sample with one of two or more groups. Commonly used supervised classifiers include neural networks (multilayer perceptrons), support vector machines, k-nearby, Gaussian mixed models, Gaussian, naive Bayes, decision trees, and radial basis function (RBF) classifiers. Including, but not limited to. Linear classification methods include Fisher's linear discrimination, logistic regression, naive Bayes classifier, perceptron, and support vector machine (SVM). Other classifiers for use in the methods according to the present disclosure include quadratic classifiers, k-nearby, boosting, decision trees, random forests, neural networks, pattern recognition, Bayesian networks, and hidden Markov models. Other classifiers containing any of these improvements or combinations commonly used in supervised learning may also be suitable for use in the methods described herein.

Classifications that use supervised methods are generally performed by the following methodologies:

1. 1. Collect the training set. These are, for example, the expression levels of one or more miRNAs described herein from a sample from a patient with endometriosis, or a book from a sample from a patient without endometriosis. It may include the expression level of one or more miRNAs described herein. The training sample is used to "train" the classifier.

2. 2. Determines the input "characteristic" representation of the function to be learned. The accuracy of the trained function depends on how the input object is represented. Typically, the input object is transformed into a characteristic vector containing some characteristics that describe the object. Properties may include a set of miRNAs found in a sample from a patient or subject.

3. 3. Determine the structure of the function to be learned and the corresponding learning algorithm. Select a learning algorithm, such as an artificial neural network, decision tree, Bayes classifier, or support vector machine. Use the learning algorithm to build a classifier.

4. Build a classifier (eg, a classification model). Run the learning algorithm on the collected training set. The parameters of the learning algorithm can be adjusted by optimizing the performance of a subset of training sets (called validation sets) or through cross-validation. After adjusting and training the parameters, the performance of the algorithm can be measured in a test set of untreated samples separate from the training set. The constructed model may include characteristic coefficients or importance measures assigned to individual characteristics.

In some cases, the individual property is the level of miRNA or miRNA. In some cases, the level of miRNA is a normalized value, an average value, a median value, an average value, an adjusted average, or another adjusted level or value. Individual properties may include or consist of sets or panels of miRNAs, such as the sets provided herein.

In some cases, the machine learning algorithm has a importance scale or characteristic importance assigned to the characteristics of the miRNA. In some cases, the importance scale or characteristic importance is pre-determined; for example, the importance scale may have been reached by training in previous datasets. In some cases, the importance measure assigned to miR-342 is one or more additional miRNAs (eg, miR-150, miR-3613, miR-451a, let-7b, or miR-125b, Or greater than the importance scale or characteristic importance assigned to other miRNAs associated with endometriosis). In some cases, the importance scale assigned to miR-451a is greater than the importance scale assigned to one or more of miR-3613, miR-125b, or let-7b.

In some embodiments, applying a machine learning algorithm (eg, a random forest algorithm) to an expression profile applies a machine learning algorithm with a particular importance scale or characteristic importance assigned to a characteristic of the miRNA. Including that, the importance scale of miR-342 is greater than at least one of miR-150, let-7b, or miR-125. In some embodiments, applying a machine learning algorithm to an expression profile comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-125b. The importance scale of is greater than at least one of miR-150, let-7b, or miR-125. In some embodiments, the highest-to-lowest importance scale or characteristic importance ranking of miRNA properties is miR-125b, let-7b, miR-2613, miR-150, miR-342, and miR-451a. (Or any subset of them). In some embodiments, the highest-to-lowest importance scale (eg, characteristic importance) rankings of miRNA properties are miR-342, miR-451a, miR-3613, miR-125b, let-7b, and. It is miR-150.

In some embodiments, a machine learning algorithm (eg, a random forest algorithm) is applied to the expression profile of miRNAs derived from saliva samples. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-125b. Is greater than at least one of let-7b, miR-3613, miR-150, miR-342, or miR-451a. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, let-7b. The importance scale or characteristic importance of is greater than at least one of miR-3613, miR-150, miR-342, or miR-451a. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, miR-3613. The importance scale or characteristic importance of is greater than at least one of miR-150, miR-342, or miR-451a. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-150. The importance scale or characteristic importance of is miR-342, or miR. Greater than at least one of 451a. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, including miR-342. Importance scale or characteristic importance of is greater than miR-451a. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, miR-451a. Importance scale or characteristic Importance of is less than at least one of miR-341, miR-150, miR-3613, let-7b, or miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, including miR-342. The importance scale or characteristic importance of is less than at least one of miR-125b, let-7b, miR-3613, or miR-150. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-150. The importance scale or characteristic importance of is less than at least one of miR-3613, let-7b, or miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to the characteristics of the miRNA, which is important for miR3613. The sex scale or trait importance is less than miR-125b or let-7b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, let-7b. Importance scale or characteristic importance of is less than miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, which is important for the miRNA. The sex scale or characteristic importance is in the order of miR-125b, let-7b, miR-3613, miR-150, miR-342, and miR-451a.

In some embodiments, a machine learning algorithm (eg, a random forest algorithm) is applied to the expression profile of miRNAs derived from a sample (eg, a serum sample). In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a materiality scale or characteristic importance assigned to a particular miRNA characteristic, including miR-342. The importance scale or characteristic importance of is greater than at least one of miR-125b, miR-451a, miR-3613, miR-150, or let-7b. In some embodiments, the importance scale or characteristic importance of miR-125b is greater than at least one of miR-451a, miR-3613, miR-150, or let-7b. In some embodiments, the importance scale or characteristic importance of miR-451a is greater than at least one of miR3613, miR-150, or let-7b. In some embodiments, the importance scale or characteristic importance of miR-3613 is greater than miR-150 or let-7b. In some embodiments, the importance scale or characteristic importance of miR-150 is greater than let-7b. In some embodiments, the importance scale or characteristic importance of let-7b is less than at least one of miR-150, miR-3613, miR-451a, miR-125b, or miR-342. In some embodiments, the importance scale or characteristic importance of miR-150 is less than at least one of miR-3613, miR-451a, miR-125b, or miR-342. In some embodiments, the importance scale or characteristic importance of miR-3613 is less than at least one of miR-451a, miR-125b, or miR-342. In some embodiments, the importance scale or characteristic importance of miR-451a is less than miR-125b or miR-342. In some embodiments, the importance scale or characteristic importance of miR-125b is less than miR-342. In some embodiments, the miRNA importance scale or characteristic importance scale or characteristic importance is in the order miR-342, miR-125b, miR-451a, miR-3613, miR-150, and let-7b. be.

In some embodiments, a machine learning algorithm (eg, a random forest algorithm) is applied to the expression profile of miRNAs derived from a sample (eg, a serum sample). In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, including miR-342. The importance scale or characteristic importance of is greater than at least one of miR-451a, miR-3613, miR-125b, let-7b, or miR-150. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, miR-451a. The importance scale or characteristic importance of is greater than at least one of miR-3613, miR-125b, let-7b, or miR-150. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, including miR-3613. The importance scale or characteristic importance of is greater than at least one of miR-125b, let-7b, or miR-150. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, miR-125b. Importance scale or characteristic Importance of is greater than let-7b or miR-150. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, let-7b. Importance scale or characteristic importance of is greater than miR-150. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-150. The importance scale or characteristic importance of is less than at least one of let-7b, miR-125b, miR-3613, miR-451a, or miR-342. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, let-7b. Importance scale or characteristic Importance of is less than at least one of miR-125b, miR-3613, miR-451a, or miR-342. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, miR-125b. The importance scale or characteristic importance of is less than at least one of miR-3613, miR451a, or miR-342. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, including miR-3613. Importance scale or characteristic Importance of is less than miR-342 or miR-451a. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, miR-451a. Importance scale or characteristic importance of is less than miR-342. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, which is important for the miRNA. The sex scale or characteristic importance is in the order of miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150.

In some embodiments, a machine learning algorithm (eg, a random forest algorithm) is applied to the expression profile of miRNAs derived from a sample (eg, a serum sample). In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-125b. The importance scale or characteristic importance of is greater than at least one of miR-3613, miR-451a, miR-150, miR-342, or let-7b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, miR-3613. The importance scale or characteristic importance of is greater than at least one of miR-451a, miR-150, miR-342, or let-7b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, miR-451a. The importance scale or characteristic importance of is greater than at least one of miR-150, miR-342, or let-7b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-150. Importance scale or characteristic Importance of is greater than miR-342 or let-7b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, including miR-342. Importance scale or characteristic importance of is greater than let-7b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, let-7b. Importance scale or characteristic Importance of is less than at least one of miR-342, miR-150, miR-451a, miR-3613, or miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, including miR-342. The importance scale or characteristic importance of is less than at least one of miR-150, miR-451a, miR-3613, or miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA characteristic, miR-150. The importance scale or characteristic importance of is greater than at least one of miR-451a, miR-3613, or miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a miRNA property, miR-451a. Importance scale or characteristic Importance of is less than miR-3613 or miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a specific importance scale or characteristic importance assigned to a characteristic of the miRNA, miR-3613. Importance scale or characteristic importance of is less than miR-125b. In some embodiments, applying such a machine learning algorithm comprises applying a machine learning algorithm having a particular importance scale or characteristic importance assigned to a characteristic of the miRNA, wherein here. The importance scale or characteristic importance of miRNA is in the order of miR-125b, miR-3613, miR-451a, miR-150, miR-342, and let-7b.

Once the classifier (eg, classification model) has been determined (“trained”) as described above, it can be used to analyze or process samples, eg, miRNAs according to the methods described herein. Patient samples containing can be classified.

An unsupervised learning approach can also be used in conjunction with the methods described herein. Clustering is an unsupervised learning approach in which a clustering algorithm correlates a set of samples without the use of markers. The most similar samples are sorted into "clusters". New samples can be sorted into clusters, thereby classifying with other members most closely related.

The methods provided herein may include the use of trained classifiers or algorithms to analyze sample data, in particular to detect endometriosis. In some cases, levels of RNA derived from the sample (eg, miRNA, ncRNA) are used to develop or train the algorithms or classifiers provided herein. In some cases, RNA levels (eg, miRNA, ncRNA levels) are measured in samples from asymptomatic patients or patients with one or more symptoms of endometriosis to detect endometriosis. A classifier or algorithm (eg, a trained algorithm) is applied to the resulting data to predict, predict, or monitor.

Training of a multidimensional classifier (eg, an algorithm) can be done using a large number of samples. For example, multidimensional classifier training is at least about 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130. , 140, 150, 160, 170, 180, 190, 200, or more. In some cases, multidimensional classifier training is at least about 200, 210, 220, 230, 240, 250, 260, 270, 280, 290, 300, 350. It can be done using, 400, 450, 500, or more samples. In some cases, multidimensional classifier training is at least about 525, 550, 600, 650, 700, 750, 800, 850, 900, 950, 1000, 1100. This can be done using samples of 1200, 1300, 1400, 1500, 1600, 1700, 1800, 2000, or more.

Further disclosed herein are a classifier set and a method of generating one or more classifier sets. The classifier set includes one or more RNAs (eg, miRNA, ncRNA), such as let-7a, let-7b, let-7c, let-7d, let-7e, let-7f, miR-135a, miR. -135b, miR-18a, miR-125b, miR-143, miR-145, miR-150, miR-342, miR-451a, miR-500a, miR-3613, and miR-6755, individually or in any combination. Can be included in. In some cases, the miRNA classifier sets used are miR-342, miR-451a, and miR-3613. In some cases, the miRNA classifier sets used are miR-342, miR-451a, miR-3613, and miR-125b. In some cases, the miRNA classifier sets used are miR-342, miR-451a, miR-3613, miR-125b, and let-7b. In some cases, the miRNA classifier sets used are miR-342, miR-451a, let-7b, and miR-125b. In some cases, the miRNA classifier sets used are miR-342, miR-451a, let-7b, and miR-3613. In some cases, the miRNA classifier sets used are miR-342, miR-451a, and let-7b. In some cases, the miRNA classifier sets used are miR-125b, miR-150, miR-342, miR-451a, and let-7b. In some cases, the miRNA classifier sets used are miR-125b, miR-150, miR-342, miR-3613, miR-451a, and let-7b.

The classifier and / or the classifier probe set can be used to either include or exclude the sample as healthy (eg, as derived from a healthy subject). For example, a classifier can be used to classify a sample as being derived from a healthy subject. Alternatively, a classifier can be used to classify the sample as being derived from an unhealthy subject (eg, as being derived from an unhealthy subject). Alternatively, or in addition, the classifier can be used to either include or exclude the sample as endometriosis (eg, as derived from a subject with endometriosis). .. For example, a classifier can be used to classify a sample as being derived from a subject suffering from endometriosis. In another example, a classifier can be used to classify a sample as derived from a subject who does not have endometriosis.

The methods disclosed herein may include assigning a classification to one or more samples from one or more subjects. The step of assigning a classification to a sample may include applying the algorithm to the level of one or more RNAs (eg, miRNA, ncRNA) derived from the sample.

The algorithm may provide a record of its output, including sample classification and / or confidence level. In some cases, the output of the algorithm may be that the subject may have a condition, eg endometriosis.

The algorithm can be a trained algorithm. The algorithm may include a linear classifier. The linear classifier may include one or more linear discriminant analysis, Fisher's linear discriminant, naive Bayes classifier, logistic regression, perceptron, support vector machine, or a combination thereof. The linear classifier can be a support vector machine (SVM) algorithm.

Algorithms include one or more linear discriminant analysis (LDA), simple perceptron, elastic net logistic regression, logistic regression, (kernel) support vector machine (SVM), diagonal linear discriminant analysis (DLDA). , Gorab classifier, Palzenbase, (Kernel) Fisher discriminant classifier, k-neighborhood method, iterative RELIEF, classification tree, maximum likelihood classifier, random forest, neighborhood centroid, microarray predictive analysis (PAM), It may include k median clustering, Fuzzy C-mean clustering, Gaussian mixed model, or a combination thereof. The algorithm may include a diagonal linear discriminant analysis (DLDA) algorithm. The algorithm may include a neighborhood centroid algorithm. The algorithm may include a random forest algorithm.

The methods provided herein can help determine, with a high degree of accuracy, sensitivity, and / or specificity, whether a patient has endometriosis. In some cases, predictive accuracy (eg, regarding detecting endometriosis or distinguishing endometriosis from non-endometriosis) is 55%, 60%, 65%, 70. %, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 98.5%, 99.0%, 99. 1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, 99.95%, or 99 Greater than .99%. In some embodiments, the prediction accuracy is 100%. In some cases, sensitivity (eg, regarding detecting endometriosis or distinguishing endometriosis from non-endometriosis) is 55%, 60%, 65%, 70%. , 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 98.5%, 99.0%, 99.1 %, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, 99.95%, or 99. Greater than 99%. In some embodiments, the sensitivity is 100%. In some cases, the specificity (eg, regarding detecting endometriosis or distinguishing endometriosis from non-endometriosis) is 55%, 60%, 65%, 70. %, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 98.5%, 99.0%, 99. 1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, 99.95%, or 99 Greater than .99%. In some cases, the specificity is 100%. In some cases, the positive predictive value of the method (eg, regarding detecting endometriosis or distinguishing endometriosis from non-endometriosis) is 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 98.5%, 99.0 %, 99.1%, 99.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, 99.95% %, Or greater than 99.99%. In some cases, the positive predictive value is 100%. The thresholded AUC in any of the methods provided herein is 0.9, 0.91, 0.92, 0.93, 0.94, 0.95, 0.96, 0.97. , 0.98, 0.99, 0.995, or 0.999. Conversely, the method may predict or determine whether the subject has no endometriosis or is at low risk of endometriosis. Negative predictive value (eg, regarding detecting endometriosis or distinguishing endometriosis from non-endometriosis) is 55%, 60%, 65%, 70%, 75%. , 80%, 85%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 98.5%, 99.0%, 99.1%, 99 From 0.2%, 99.3%, 99.4%, 99.5%, 99.6%, 99.7%, 99.8%, 99.9%, 99.95%, or 99.99% May also be large. In some cases, the negative predictive value is substantially equal to 100%.

In some cases, the methods provided herein are optimized for specificity by selecting the optimal cutoff point. Above that, the cutoff point can be characterized that the sample is from a subject with endometriosis, below that, the sample is from a subject negative for endometriosis. It can be a threshold or cutoff identified as. In some situations, the sample is identified as coming from a subject who is positive for endometriosis below the cutoff and negative for endometriosis above the cutoff. .. In some cases, the cutoff point or threshold is on the receiver operating characteristic (ROC) curve or voting percentage distribution. In some cases, the method has a specific specificity, eg, greater than 80%, greater than 85%, greater than 90%, greater than 95%, 98%. Optimized to achieve greater specificity. In some cases, the specificity is optimized for values that exceed the sensitivity of the assay. In some cases, the specificity is optimized to exceed a certain value, eg, greater than 85%, resulting in less than a certain value, eg, less than 95%, less than 90%, less than 85%, Assay sensitivities of less than 80% can be achieved. In some cases, the specificity is optimized to exceed a certain value, eg, greater than 90%, resulting in less than a certain value, eg, less than 95%, less than 90%, less than 85%, Assay sensitivities of less than 80% can be achieved. In some cases, the specificity is optimized to exceed a certain value, eg, greater than 95%, resulting in less than a certain value, eg, less than 95%, less than 90%, less than 85%. Assay sensitivities of less than 80% can be achieved. Such assays are especially non-life-threatening because they can help subjects avoid unnecessary interventions such as surgical interventions that can occur if the assay is associated with a high percentage of false positives. Can be suitable. In some cases, the method achieves the specificity described herein (eg, greater than 80%, greater than 90%) for a diverse population of women. The population can be a group of more than 50 women, more than 100 women, more than 500 women, more than 1000 women, and so on. The population may include women with various benign conditions, such as cysts, uterine fibroids, leiomyomas, cystadenomas, chronic pelvic infections, teratomas, paraovarian cysts, and the like.

In some cases, the methods provided herein are optimized for sensitivity by selecting the optimal cutoff point. Above that, the cutoff point can be characterized that the sample is from a subject with endometriosis, below that, the sample is from a subject negative for endometriosis. It can be a threshold or cutoff identified as. In some situations, the sample is identified as coming from a subject who is positive for endometriosis below the cutoff and negative for endometriosis above the cutoff. In some cases, the cutoff point or threshold is on the receiver operating characteristic (ROC) curve or voting percentage distribution. In some cases, the method has specific sensitivities, such as greater than 80%, greater than 85%, greater than 90%, greater than 95%, greater than 98%. Optimized to achieve. In some cases, the sensitivity is optimized to a value that exceeds the specificity of the assay. In some cases, the sensitivity is optimized to exceed a certain value, eg, greater than 85%, resulting in less than a certain value, eg, less than 95%, less than 90%, less than 85%, 80. The specificity of the assay can be less than%. Such assays may be particularly suitable for screening tests for endometriosis. In some cases, the method achieves the sensitivities described herein (eg, greater than 80%, greater than 90%) for diverse female populations. The population can be a group of more than 50 women, more than 100 women, more than 500 women, more than 1000 women, and so on. The population may include women with various benign conditions, such as cysts, uterine fibroids, leiomyomas, cystadenomas, chronic pelvic infections, teratomas, paraovarian cysts, and the like.

The methods disclosed herein may include assigning a classification to one or more samples from one or more subjects. The step of assigning a classification to a sample may include applying the algorithm to the level of one or more RNAs (eg, miRNA, ncRNA) derived from the sample.

The algorithm may provide a record of its output, including sample classification and / or confidence level. In some cases, the output of the algorithm may be that the subject may have a condition, eg endometriosis.

In some embodiments, the present disclosure is a procedure for treating endometriosis detected herein based on a classification generated using the machine learning algorithms described herein. Provides application to the subject. Treatments for endometriosis include, but are not limited to, analgesics (eg, NSAIDs), hormonal treatments, chemotherapy, and surgical treatments. Analgesics used in the treatment of endometriosis include simple analgesics such as paracetamol, COX-2 inhibitors, aspirin, and other non-steroidal anti-inflammatory drugs well known in the art, as well as morphine and codeine. , And drug-induced analgesics such as oxycodone. Hormonal treatments include oral contraceptives; progestins such as dydrogesterone, medroxyprogesterone acetate, medroxyprogesterone depotacetate, norethisterone, levonorgestrel, and others well known in the art; progesterone and progesterone-like substances; leuprorelin, GnRH agonists such as, but not limited to, buserelin, gosereline, histrelin, deslorerin, nafarelin, triptorelin, and leuprorelin; synthetic androgens such as androgens and danazoles; GnRH antagonists; and aromatase inhibitors. Surgical procedures include, but are not limited to, laparoscopic surgery, hysterectomy, and oophorectomy.

In some embodiments, the GnRH antagonist is administered to treat endometriosis detected herein. Peptides (goserelin acetate, buserelin, histrelin, deslorerin, nafarelin, and triptorelin, leuprorelin) and non-peptides (Elagolix / ABT-620, NBI-56418, eg Taylor et al. N Engl J Med. 2017 Jul 6; 377 (1): 28-40), various antagonists of GnRH suitable for clinical administration are the second choice of endometriosis in individuals with refractory endometriosis. Available for treatment.

In some embodiments, the fluid sample collection and diagnosis described above is performed after a defined period of time (eg, 1 month, 6 months, or 1 year) and if endometriosis is not detected, intrauterine. The initial dose of the drug used to treat endometriosis (eg, either the hormone analogs or antagonists described herein) is adjusted downwards. In some embodiments, the fluid sample collection and diagnosis described above is performed after a defined period of time (eg, 1 month, 6 months, or 1 year) and if endometriosis is detected, the drug. The initial dose of is adjusted upwards. In some embodiments, the fluid sample collection and diagnosis described above is performed after a defined period of time (eg, 1 month, 6 months, or 1 year) and if endometriosis is not detected, of the drug. Administration is terminated.

Computer-implemented methods The expression levels of one or more RNAs (eg, miRNAs, ncRNAs) can be analyzed in a digital computer and associated with the subject's condition (eg, endometriosis). If desired, such a computer directly to a scanner or the like (eg, qPCR system, multiplex fluorescent plate reader, FACS instrument, or sequencer) that receives signals related to experimentally determined miRNA or ncRNA expression levels. Are linked together. Alternatively, the expression level may be entered by other means. As described above, the computer converts the raw signal into expression levels (absolute or relative) and measures the measured expression levels at one or more reference expression levels or stages of such values (. It can be programmed to compare with (scale). The computer also assigns a value or other specification to the expression level based on a comparison with one or more reference expression levels and aggregates such values or instructions for multiple genes in the expression profile. It can also be programmed as. The computer should also output values or other indications that provide an indicator of the presence of endometriosis, as well as either raw or intermediate data used in determining such values or indications. It can also be programmed.

Typical computers (US 6,785,613, see FIGS. 4 and 5) include a central processor, system memory, input / output controllers, external devices such as a printer via a parallel port, a display via a display adapter. It may include buses interconnecting major subsystems such as screens, serial ports, keyboards, fixed disk drives, and floppy (registered) disk drives that can operate to accept floppy (registered) disks. A number of other devices, such as a scanner via an I / O controller, a mouse connected to a serial port, or a network interface may be connected. The computer contains a computer-readable medium that holds the code for causing the computer to perform various functions. These functions include controlling the automation device, receiving inputs, and delivering outputs, as described above. The automated device may include a robot arm for delivering reagents to determine expression levels, as well as a small container for performing expression analysis, such as a microtiter well.

The methods, systems, kits, and compositions provided herein may also be capable of producing and transmitting results through a computer network. In some cases, the sample is first taken from the subject (eg, a patient with one or more symptoms of endometriosis, or a patient without symptoms). In some cases, the sample is assayed and the level of RNA (eg, miRNA, ncRNA) is measured. Computer systems can be used to analyze the data and perform sample classification. Results can be sent to different types of end users over a computer network. In some cases, the subject (eg, patient) may have access to the results by using stand-alone software and / or web-based applications on a local computer that has access to the Internet. Can be. In some cases, the results may be accessed via a mobile application provided on a mobile digital processing device (eg, mobile phone, tablet, etc.). In some cases, the results may be accessed by a healthcare provider (eg, a physician) to help the physician identify and track the patient's condition. In some cases, the results may be used for other purposes such as education and research.

Computer Programs The methods, kits, and systems disclosed herein may include at least one computer program or its use. A computer program may include a sequence of instructions that can be executed by the CPU of a digital processing device, written to perform a specified task. Computer-readable instructions can be implemented as program modules, such as functions, objects, application programming interfaces (APIs), data structures, etc., that perform specific tasks or implement specific abstract data types. In light of the disclosures provided herein, one of ordinary skill in the art will recognize that computer programs may be written in different versions and in different languages.

The functionality of computer-readable instructions can be combined or distributed as desired in a variety of environments. A computer program may usually provide a sequence of instructions from one position or multiple positions. In various embodiments, the computer program is one or more web applications, one or more mobile applications, one or more stand-alone applications, one or more web browser plug-ins, extensions, add-ins, etc. Alternatively, the add-in or a combination thereof is included in part or in whole.

A system for classifying one or more samples and their use thereof are further disclosed herein. The system is a digital processing device, including (a) an operating system and a memory device configured to make executable instructions, (b) (i) one or more RNAs (eg,) derived from a sample from a subject. , MiRNA, ncRNA) RNA (eg, miRNA, ncRNA) expression profile of a first software module configured to receive, (ii) subject-derived RNA (eg, miRNA, ncRNA). A second software module configured to analyze, and (iii) a third configured to classify samples from subjects based on a classification system containing two or more classes. It may include a computer program containing instructions that can be executed by a digital processing device to classify samples from subjects, including software modules. At least one of the classes can be selected from endometriosis. Analyzing a gene expression profile derived from a subject may include applying an algorithm. Analyzing a gene expression profile can include normalizing an RNA (eg, miRNA, ncRNA) expression profile derived from a subject (eg, a constitutive RNA, eg, a small nuclear RNA U6, RNU48, etc.). For RNU44, U47, or RNU6B, or any combination thereof.

FIG. 6 is a computer system (as used herein) that is programmed to implement the methods of the present disclosure, eg, for generating selector sets and / or for data analysis, or is configured in other ways. Also referred to as "system") 601. System 601 includes a central processing unit (CPU, also referred to herein as "processor" and "computer processor") 605, which can be a single core or multi-core processor, or multiple processors for parallel processing. The system 601 also has a memory 610 (eg, random access memory, read-only memory, flash memory), an electronic storage unit 615 (eg, a hard disk), and a communication interface 620 (eg, for example) for communicating with one or more other systems. , Network adapters), as well as peripheral devices 625, such as caches, other memory, data storage and / or electronic display adapters. The memory 610, the storage unit 615, the interface 620, and the peripheral device 625 are in communication with the CPU 605 through a communication bus (solid line), for example, a motherboard. The storage unit 615 can be a data storage unit (or data repository) for storing data. The system 601 is operably connected to a computer network (“network”) 630 using a communication interface 620. The network 630 can be the Internet, the Internet and / or an extranet, or an intranet and / or an extranet that is in communication with the Internet. In some cases, the network 630 is a telecommunications and / or data network. The network 630 may include one or more computer servers that can enable distributed computing such as cloud computing. In some cases, the network 630 may utilize the system 601 to implement a peer-to-peer network that may allow the device attached to the system 601 to act as a client or server.

The system 601 may be in communication with the processing system 635. The processing system 635 may be configured to implement the methods disclosed herein. In some examples, the processing system 635 is a multiplex fluorescent plate reader, qPCR machine, or nucleic acid sequencing system, such as a next generation sequencing system (eg, Illumina sequencer, Ion Torrent sequencer, Pacific Biosciences sequencer). The processing system 635 may be in communication with the system 601 via network 630 or by direct (eg, wired, wireless) connection. The processing system 635 may be configured for analysis, eg, nucleic acid sequence analysis.

The method described herein uses a machine (or computer processor) executable code (or software) stored in an electronic storage location of system 601 such as memory 610 or electronic storage unit 615. Can be implemented. In use, the code may be executed by processor 605. In some examples, the code may be retrieved from storage unit 615 and stored in memory 610 for immediate access by processor 605. In some situations, the electronic storage unit 615 may be excluded and machine executable instructions are stored in memory 610.

Digital Processing Devices The methods, kits, and systems disclosed herein may include digital processing devices, or their use. In a further embodiment, the digital processing device includes one or more hardware central processing units (CPUs) that perform the functions of the device. Still in further embodiments, the digital processing device further comprises an operating system configured to give executable instructions. In some embodiments, the digital processing device is connected to a computer network as needed. In a further embodiment, the digital processing device is connected to the Internet as needed and, as a result, accesses the World Wide Web. Still in a further embodiment, the digital processing device is connected to the cloud computing infrastructure as needed. In another embodiment, the digital processing device is connected to the intranet as needed. In another embodiment, the digital processing device is connected to the data storage device as needed.

As described herein, suitable digital processing devices include, but are not limited to, server computers, desktop computers, laptop computers, notebook computers, sub-notebook computers, netbook computers, netpad computers, and the like. These include set-top computers, handheld computers, internet appliances, mobile smartphones, tablet computers, mobile information terminals, video game consoles, and vehicles. Those of skill in the art will recognize that many smartphones are suitable for use in the systems described herein. Those skilled in the art will also recognize that selective televisions, video players, and digital music players with computer network connectivity, if desired, are suitable for use in the systems described herein. Let's do it. Suitable tablet computers include those having a booklet, slate, and convertible configuration known to those of skill in the art.

Digital processing devices may generally include an operating system configured to give executable instructions. The operating system can be, for example, software, including programs and data, that manages the hardware of the device and provides services for running applications. Suitable server operating systems for those skilled in the art include, as non-limiting examples, FreeBSD, OpenBSD, NetBSD®, Linux®, Apple® Mac OS X Server®. , Oracle® Solaris®, Windows® Server®, and Novell® NetWare®. As a suitable personal computer operating system for those skilled in the art, as non-limiting examples, Microsoft (registered trademark) Windows (registered trademark), Apple (registered trademark) Mac OS X (registered trademark), UNIX (registered trademark). , And UNIX®-like operating systems such as GNU / Windows®. In some embodiments, the operating system is provided by cloud computing. Suitable mobile smartphone operating systems for those of skill in the art include, as non-limiting examples, Nokia® Symbian® OS, Apple® iOS®, Research In Motion®. ) BlackBerry OS®, Google® Android®, Microsoft® Windows® Phone OS, Microsoft® Windows® Mobile®. ) OS, Linux®, and Palm® WebOS®.

Devices typically include storage and / or memory devices. The storage and / or memory device can be one or more physical devices used to store data or programs transiently or permanently. In some embodiments, the device is volatile memory and requires power to maintain the stored information. In some embodiments, the device is non-volatile memory and retains the stored information even when the digital processing device is unpowered. In a further embodiment, the non-volatile memory includes flash memory. In some embodiments, the non-volatile memory includes a dynamic random access memory (DRAM). In some embodiments, the non-volatile memory includes a ferroelectric random access memory (FRAM®). In some embodiments, the non-volatile memory includes a phase change random access memory (PRAM). In other embodiments, the device is a storage device, which includes, but is not limited to, CD-ROMs, DVDs, flash memory devices, magnetic disk drives, magnetic tape drives, optical disk drives, and cloud computing. Includes storage based on ing. In a further embodiment, the storage and / or memory device is a combination of devices, such as those disclosed herein.

The display for sending visual information to the user can generally be initialized. Examples of displays include cathode line tubes (CRTs, liquid crystal displays (LCDs), thin film liquid crystal displays (TFT-LCDs, organic light emitting diode (OLED) displays. In various further embodiments, the OLED display is a passive matrix OLED. (PMOLED) or active matrix OLED (AMOLED) display. In some embodiments, the display can be a combination of devices such as a plasma display, a video projector, or those disclosed herein.

Digital processing devices may generally include input devices for receiving information from the user. Input devices include, for example, keyboards, non-limiting examples of pointing devices including mice, trackballs, trackpads, joysticks, game controllers, or stylus, touch screens or multi-touch screens, voice or other sound input. It can be a combination of devices such as a microphone for capture, a video camera for capturing motion or visual input, or those disclosed herein.

Non-transient computer readable storage media The methods, kits, and systems disclosed herein include a program that contains instructions that can be run to perform and analyze the tests described herein by the operating system. Is coded and may preferably include one or more non-transient computer readable storage media connected to a digital processing device in the network. The computer-readable storage medium can be a digital tangible component that can be removed from the digital processing device if desired. Computer-readable storage media include, but are not limited to, CD-ROMs, DVDs, flash memory devices, solid state memory, magnetic disk drives, magnetic tape drives, optical disk drives, cloud computing systems and services. .. In some cases, programs and instructions are permanently, substantially permanent, semi-permanent, or non-transiently encoded in the medium.

A non-transient computer-readable storage medium may be encoded by a computer program containing instructions that can be executed by the processor to create or use a classification system. The storage medium is (a) a database of one or more clinical characteristics of two or more control samples, and (i) two or more control samples of two or more. A database in computer memory that can be derived from a subject and (ii) two or more control samples can be classified differentially based on a classification system containing three or more classes. Based on (b) a first software module configured to compare the clinical characteristics of one or more control samples with two or more, and (c) a comparison of one or more clinical characteristics. It may include a second software module configured to generate a classifier set.

At least two of the classes can be selected from endometriosis, non-endometriosis, and healthy.

Web Application In some embodiments, the computer program includes a web application. In light of the disclosure provided herein, one of ordinary skill in the art recognizes that web applications utilize one or more software frameworks and one or more database systems in various embodiments. Will do. In some embodiments, the web application is Microsoft®. Created with a software framework such as NET or Ruby on Rails (RoR). In some embodiments, the web application utilizes one or more database systems, including non-limiting examples of relationship, non-relationship, object-oriented, associative, and XML database systems. In a further embodiment, suitable relationship database systems include, as non-limiting examples, Microsoft® SQL Server, mySQL®, and Oracle®. Those skilled in the art will also recognize that web applications are written in one or more versions of one or more languages in various embodiments. Web applications can be written in one or more markup languages, presentation definition languages, client-side scripting languages, server-side coding languages, database query languages, or a combination thereof. In some embodiments, the web application is described to some extent in a markup language, such as hypertext markup language (HTML), extensible hypertext markup language (XHTML), or extensible markup language (XML). Will be done. In some embodiments, the web application is, to some extent, written in a presentation definition language such as a Cascade Style Sheet (CSS). In some embodiments, the web application is, to some extent, a client-side scripting language such as asynchronous JavaScript® and XML (AJAX), Flash® Actionscript, Javascript®, or Silverlight®. ). In some embodiments, the web application is, to some extent, a server-side coding language, such as Active Server Pages (ASP), ColdFusion®, Perl, Java®, Java Server Pages (JSP), Hypertext Pre. Described in Processor (PHP), Python®, Ruby, Tcl, Smalltalk, WebDNA®, or Groovy. In some embodiments, the web application is written to some extent in a database query language, such as a structured query language (SQL). In some embodiments, the web application incorporates a corporate server product, such as IBM® Lotus Domino®. In some embodiments, the web application comprises a media player element. In various further embodiments, the media player element is, as a non-limiting example, Adobe® Flash®, HTML 5, Apple® QuickTime®, Microsoft® Silverlight. Utilize one or more of a number of suitable multimedia technologies, including (Registered Trademarks), Java®, and Unity®.

Mobile Application In some embodiments, the computer program comprises a mobile application provided to a mobile digital processing device. In some embodiments, the mobile application is provided to the mobile digital processing device as it is manufactured. In another embodiment, the mobile application is provided to the mobile digital processing device via the computer network described herein.

Given the disclosures provided herein, mobile applications may be created using techniques known to those of skill in the art using hardware, languages, and development environments known in the art. Those skilled in the art will recognize that mobile applications are written in several languages. Suitable programming languages include, but are not limited to, C, C ++, C #, Objective-C, Java®, Javascript®, Pascal, Object Pascal, Python®, Ruby, VB. .. Includes XHTML / HTML with or without NET, WML, and CSS, or a combination thereof.

Suitable mobile application development environments are available from several sources. Commercially available development environments include, but are not limited to, Airplay SDK, archeMo, Appcelerator®, Celsius, Bedrock, Flash Lite ,. NET Compact Framework, Rhomobile, and WorkLight Mobile Platform can be mentioned. Other development environments are available free of charge, including, as non-limiting examples, Lazarus, MobiFlex, MoSync, and Phonegap. Mobile device manufacturers also distribute software developer kits, which include, but are not limited to, the iPhone® and iPad® (iOS) SDKs, Android®. ) SDK, BlackBerry® SDK, BREW SDK, Palm® OS SDK, Symbian SDK, webOS SDK, and Windows® Mobile SDK.

As a non-limiting example, if you are a person in the art, Apple (registered trademark) App Store, Android (registered trademark) Market, BlackBerry (registered trademark) App World, App Store for Palm devices, App Catalog (Registered) Several commercial forums are available for distribution of mobile applications, including the Marketplace for Mobile, Ovi Store for Nokia® devices, Samsung® App, and Nintendo® DSi Shop. You will recognize that there is.

Stand-alone application In some embodiments, the computer program includes a stand-alone application that is not like an add-on to an existing process and, for example, is a program that operates as an independent computer process rather than a plug-in. Those of skill in the art will recognize that standalone applications are often compiled. A compiler is a computer program that converts source code written in a programming language into binary object code such as assembly language or machine code. Suitable programming languages to be compiled include, but are not limited to, C, C ++, COBOL, Delphi, Eiffel, Java®, Lisp, Python ™, Visual Basic, and VB. NET, or a combination thereof can be mentioned. Compiling is often done, at least in part, to create an executable program. In some embodiments, the computer program comprises one or more executable compiled applications.

Web Browser Plug-in In some embodiments, the computer program includes a web browser plug-in. In computing, a plug-in is one or more software components that add specific functionality to a larger software application. Software application manufacturers enable third-party developers to bring the ability to extend their applications, help them easily add new characteristics, and reduce the size of their applications. For that, it supports plugins. If supported, plugins allow you to customize the functionality of your software application. For example, plugins are commonly used in web browsers to play video, generate interactivity, scan for viruses, and display specific file types. Several web browser plug-ins for those of skill in the art, including Adobe® Flash® Player, Microsoft® Silverlight®, and Apple® QuickTime®. Will be familiar with. In some embodiments, the toolbar comprises one or more web browser extensions, add-ins, or add-ons. In some embodiments, the toolbar includes one or more explorer bars, tool bands, or desk bands.

Given the disclosures provided herein, one of ordinary skill in the art will use C ++, Delphi, Java®, PHP, Python ™, and VB. You will recognize that there are several plugin frameworks available that allow the development of plugins in various programming languages, including NET, or a combination thereof.

Web browsers (also known as Internet browsers) are software designed for use with networked digital processing devices to extract, display, and traverse information resources on the World Wide Web. It is an application. Suitable web browsers include, but are not limited to, Microsoft® Internet Explorer®, Firefox® Firefox®, Google® Chrome, Apple® Safari. (Registered Trademarks), Opera Safari® Opera®, and KDE Konqueror. In some embodiments, the web browser is a mobile web browser. Mobile web browsers (also known as micro-browsers, mini-browsers, and wireless browsers) are non-limiting examples of handheld computers, tablet computers, netbook computers, sub-notebook computers, smartphones, music players, and mobile information terminals. (PDA), and designed for use in mobile digital processing devices, including handheld video gaming systems. Suitable mobile web browsers include, as a non-limiting example, Google® Android® browser, RIM BlackBerry® Brownser, Apple® Safari®, Palm®. ) Blazer, Palm (registered trademark) WebOS (registered trademark) Browser, Mozilla (registered trademark) Filefox (registered trademark) for mobile, Microsoft (registered trademark) Internet Explorer (registered trademark) Mobile, Amazon (registered trademark) Kindle (registered trademark). ) Basic Web, Nokia® Browner, Opera Software® Opera® Mobile, and Sony® PSP® browsers.

Software Modules The methods, kits, and systems disclosed herein may include software, servers, and / or database modules, or their use. In light of the disclosures provided herein, software modules will be created by techniques known to those of skill in the art using machines, software, and languages known in the art. The software modules disclosed herein are implemented in a number of ways. In various embodiments, software modules include files, code sections, programming objects, programming structures, or combinations thereof. In a further variety of embodiments, the software module comprises a plurality of files, a plurality of code sections, a plurality of programming objects, a plurality of programming structures, or a combination thereof. In various embodiments, the software module may include web applications, mobile applications, and stand-alone applications, as non-limiting examples. In some embodiments, the software module is in one computer program or application. In other embodiments, the software module is in more than one computer program or application. In some embodiments, the software module is hosted on one machine. In other embodiments, the software module is hosted on more than one machine. In a further embodiment, the software module is hosted on a cloud computing platform. In some embodiments, the software module is hosted on one or more machines in one location. In other embodiments, the software module is hosted on one or more machines in more than one location.

Databases The methods, kits, and systems disclosed herein may include one or more databases, or their use. Given the disclosures provided herein, a number of databases of skill in the art will store information about miRNA or ncRNA expression profiles, sequencing data, classifiers, classification systems, treatment regimens, or combinations thereof. And will recognize that it is suitable for retrieval. In various embodiments, suitable databases include, for example, relationship databases, non-relationship databases, object-oriented databases, object databases, entity-related model databases, associative databases, and XML databases. In some embodiments, the database is based on the Internet. In a further embodiment, the database is web based. Still in a further embodiment, the database is based on cloud computing. In other embodiments, the database is based on one or more local computer storage devices.

Data Transmission The methods, kits, and systems disclosed herein can be used to transmit one or more reports. One or more reports may contain information regarding the classification and / or identification of one or more samples from one or more subjects. One or more reports may include information about the disease state (eg, endometriosis or non-endometriosis). One or more reports may include information about a therapeutic regimen for use in subjects who need it to treat endometriosis. One or more reports may be sent to the subject or the subject's medical personnel. The subject's healthcare professional can be a physician, physician's assistant, nurse, or other healthcare provider. The subject's medical personnel may be the subject's family. The subject's family can be a parent, guardian, child, sibling, aunt, uncle, cousin, or spouse. The subject's medical personnel may be the subject's legal representative.

Although preferred embodiments of the invention are shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided merely by way of illustration. The present invention is not intended to be limited by the specific examples provided herein. Although the present invention has been described with reference to the specification described above, the description and illustration of embodiments herein are not meant to be construed in a limited sense. One of ordinary skill in the art will recall a number of variants, modifications, and substitutions without departing from the invention. Further, it should be understood that all aspects of the invention are not limited to the specific depictions, configurations, or relative ratios described herein, depending on various conditions and variables. .. It should be appreciated that various alternatives to the embodiments of the invention described herein can be used in the practice of the invention. Accordingly, the invention is intended to cover any such alternative, modified, modified, or equivalent. It is intended that the following claims define the scope of the invention, as well as the methods and structures within these claims, as well as their equivalents.

(Example 1)
Study group / patient selection

Presented herein are studies testing miRNA biomarkers for endometriosis in a more diverse patient population than previous studies. Serum samples were taken from women prior to undergoing laparoscopic surgery for benign gynecological conditions and miRNA expression analysis of 6 target miRNAs was performed without knowledge of surgical findings. This study design enables testing of whether preoperative evaluation of these miRNA biomarkers can distinguish endometriosis from other benign conditions in a diverse patient population of early and late stage diseases. And said.

Institutional Review Board (IRB) approval was obtained from Yale University School of Medicine (New Haven, CT) for biomedical research on human subjects. Written informed outlets are admitted to Yale-New Haven Hospital and undergoing laparoscopic or laparotomy for suspected benign signs such as pelvic mass, pelvic pain, infertility, and endometriosis. Obtained from. The selection criteria were women aged 18-49 years. Exclusion criteria consisted of postmenopausal patients, pregnancy, severe anemia, hyperplasia or polyps, or malignant tumors. Subjects should be in the disease group if visual findings (and pathology, if necessary) during surgery confirm the presence of endometriosis, or if surgery reveals other benign pathologies. It was stratified into a control group. A full range of diseases resulting in various miRNA levels was obtained, including all stages of endometriosis, as well as untreated and treated subjects. Hormonal drug treatment was recorded for each patient. The phase of the menstrual cycle was determined based on the date of the patient's last menstrual period. The staging was performed by a medical record review using the revised American Society for Reproductive Medicine (rASRM) classification by assessing surgical comments and pathological outcomes (explicitly herein by reference). Revised American Society for Reproductive Medicine classification of endometriosis: 1996. Fertility and sterility. 1997; 67 (5): 817-821).

Sampling Serum samples were taken from 103 women who underwent laparoscopic or open gynecological procedures at Yale New Haven Hospital between September 2016 and October 2017.

Blood (5-10 ml) is taken from the subject prior to surgery (laparoscopic or open) to collect serum samples and in sterile tubes (BD, Franklin Lakes, NJ, USA) without additives. Collected in. Serum was immediately collected by centrifugation at 2500 rpm for 15 minutes at 4 ° C. and a fixed volume of 300 μl was stored at −80 ° C.

Of the 103 patients, 3 were excluded due to unexpected comorbidities including malignant tumors in the pathology. Of the remaining 100 patients, 41 were categorized as the endometriosis study population and 59 were categorized as the control study population. The endometriosis group was categorized based on visual diagnosis and histological verification of the disease. Control groups were categorized based on the lack of visible disease at the time of surgery.

Table 1 summarizes the demographics and clinical characteristics of the study subjects. The mean age of the study population was 34.1 ± 7.1 years in the endometriosis group and 36.9 ± 8.2 years in the control group. The classification of obesity index was 28.1 ± 7.5 in the endometriosis group and 30.4 ± 7.5 in the control group. There was no statistically significant difference between the ages of the women in the two groups (Student's t-test) or between BMI values (Student's t-test). Study subjects were identified primarily as Caucasian, followed by Black / African American and Hispanic.

The endometriosis group, when categorized by the rASRM stage, consisted of varying degrees of disease. Forty-one endometriosis subjects were included in 11 (29%) Stage I, 7 (17%) Stage II, 15 (36%) Stage III, and 8 (19%). Divided into stage IV. Endometrioma was reported in 13 patients. The control group consisted of various benign pathologies. Fifty-nine control subjects were divided into the following categories: 23 (39%) smooth myomas, 4 (7%) cyst adenomas, 5 (8%) chronic pelvic infections, 3 No teratomas in humans (5%), paraovarian cysts in 6 (10%), and no abnormal pathology observed in 18 (31%) (Table 1).

The phases of the menstrual cycle and the presence of hormonal drug treatment are recorded and can also be seen in Table 1. In about half of the study subjects, the phase of the menstrual cycle could not be accurately determined either based on the history of irregular cycles (no data) or for the use of hormonal drug therapy. Of the total number of subjects, including those that could be determined, 8 (19%) in the endometriosis group were in the proliferative phase and 15 (36%) were in the secretory phase and 14 in the control group. Humans (24%) were in the proliferative phase and 13 (22%) were in the secretory phase. Many study subjects used hormonal agents at the time of serum collection. In the endometriosis group, the hormone category is 10 (24%) combined oral contraception, 5 (12%) progesterone only, 0 (0%) estrogen only, 6 (15%). ) GnRH agonist, 1 (2%) aromatase inhibitor, and the remaining 19 (46%) did not use hormonal agents. In the control group, the hormone category was 10 (17%) combined oral contraception, 16 (27%) progesterone only, 1 (2%) estrogen only, 5 (8%) GnRH. Agonists included one (2%) aromatase inhibitor, and the remaining 26 (44%) did not use hormonal agents.

(Example 2)
Analysis of miRNA expression from serum and saliva samples

Serum sample

Total miRNAs were extracted from 300 μl of serum samples taken as in Example 1 using the miRNeasy minikit from Qiagen (Valencia, CA, USA) and from Applied Biosystems of Life Technologies (Carlsbad, CA). The Applied miRNA cDNA synthesis kit was used and reverse transcribed according to the manufacturer's specifications. MicroRNA levels were quantified by qRT-PCR using SYBR Green (Bio-Rad Laboratories, Hercules, CA) with MyQ Single Color Real-Time PCR Detection System (Bio-Rad). The specificity of the amplified transcript and the absence of primer dimers were confirmed by melting curve analysis. Primers for miRNA and U6 genes are described in W. M. Obtained from the Keck Oligonucleotide Synthesis Facility (Yale University, New Haven, CT) and universal reverse primers were obtained from Applied Biosystems. The primer sequences are as described in previous studies ⁴⁰ , 45 and are listed in Table 2. For the qRT-PCR reaction, the following cycling conditions were used: 95 ° C. for 3 minutes, 40 cycles (95 ° C. for 15 seconds, 59 ° C. for 5 seconds, 72 ° C. for 55 seconds). Expression of each miRNA was normalized to expression of human U6 small nuclear RNA. Relative expression was calculated for each miRNA using the delta (Ct) method, also known as the 2 ^-ΔCt method. All experiments were performed at least twice each with two replica wells on each plate.

(Example 3)
Expression analysis of individual miRNAs from serum samples

In this case-control study, the presence of endometriosis was not known prior to surgery / sampling and serum miRNA analysis was performed without knowledge of the disease state. Expression levels of 6 miRNAs (miR-125b-5p, miR-150-5p, miR-342-3p, miR-451a, miR-3613-5p, and let-7b) were measured by quantitative real-time polymerase chain reaction (miR-425-3p, and let-7b). It was measured prospectively using qRT-PCR) and quantified for small nuclear RNA U6. FIG. 1 shows the average expression level of these circulating miRNAs in a subject with endometriosis vs. a control subject with other benign gynecological pathologies. In patients with endometriosis, a significant increase in the expression of miR-125b, miR-150-5p, miR-342-3p, and miR-451a was observed compared to controls, miR-3613-5p and A significant decrease in the expression of let-7b was observed.

Subgroup analyzes were performed to assess whether the timing of serum collection during the menstrual cycle affects miRNA expression and may affect the accuracy of diagnosis by these biomarkers. Control patients were divided based on menstrual period and no significant difference in expression was observed at miRNA levels between patients sampled during the proliferative phase (14,24%) vs. secretory phase (13,22%) (Figure). See 2). Similar analyzes were performed on women during the cycle identified as having endometriosis and found no difference in mean miRNA levels between phases of the menstrual cycle.

Compared to other studies, this study was designed to reflect a more realistic patient population with respect to demographics and ongoing hormone therapy. There were multiple types of hormone therapy (Table 1), with the majority of patients receiving combined oral contraceptives (10, 24%) or GnRH agonists (6, 14%). To determine if hormonal medications performed at the time of surgery affected miRNA expression, we analyzed expression data from patients with endometriosis and treated women with hormones (22, The expression of these miRNAs was compared between 54%) and women who did not receive hormonal drug treatment (19, 46%). As can be seen in FIG. 3, the presence of hormone treatment did not significantly affect the average expression levels of the 6 target miRNAs tested.

To assess whether the expression of these miRNAs correlates with the stage of endometriosis, minimal / mild (stage I / II) endometriosis is moderate / severe (stage III /) in the analysis. Separated from IV). Using the Kruskal-Wallis test (nonparametric one-way ANOVA), all 6 miRNAs were significantly different variances (p <0) among the three groups of control, stage I / II, and stage III / IV. It was found to have 0.05) (Fig. 4). However, after using Dunn's multiple comparison test for the three pairwise comparisons, each subgroup of endometriosis showed significantly different miRNA levels compared to the control group, but minimal / mild vs. moderate. Not shown between / severe (Fig. 4). We also compared miRNA levels in patients with ovarian endometriosis (endometriosis) and those without endometriosis, but no significant difference was observed in this subgroup analysis.

Recipient behavioral characteristics (ROC) analysis of each miRNA was performed to assess the ability of individual miRNA expression levels to be used as biomarkers for the presence of endometriosis. Individual miRNAs had ROC subcurve area (AUC) scores ranging from a minimum of 0.68 for miR-150-5p to a maximum of 0.92 for miR-342-3p (Table 3). For up-regulated miRNAs (miR-125b-5p, miR-150-5p, miR-342-3p, miR-451a), levels above the cutoff value indicate endometriosis but are down-regulated. For miRNAs (let-7b and miR-3613-5p), levels below the cutoff are associated with endometriosis.

In addition, individual miRNA models were analyzed to determine cutoffs that resulted in at least 80% sensitivity or at least 80% specificity. These values are presented in Table 3A.

(Example 4)
Construction and analysis of multivariate models based on miRNA expression in serum

First, the diagnostic performance of the previously reported combinations of miR-125b-5p, miR-451a, and miR-3613-5p was evaluated (Cosar et al. Fertility and sterility. 2016; 106 (2): 402- See 409). When tested against this dataset, this combination gave an AUC score of 0.8. Previous studies included only women with moderate / severe (stage III / IV) endometriosis, and this study included cases of minimal / mild (stage I / II) disease. Due to their higher inclusion, machine learning-based optimization analysis was performed to identify the highest scoring combination of miRNA biomarkers in this more diverse and representative population of patients. Using machine learning with a random forest approach, we developed a new machine learning model using 6 miRNAs. This model was validated in two ways: by random subsampling, which divides the entire data set into training and test subsets, and an independent data set (n = 48, 24 endometriosis and 24 control subjects). According to the test for the body). The AUC score of the model performance of the training and test data sets is shown in FIG. An AUC of 0.939 for the 6-marker classifier algorithm was achieved in an independent validation experiment after performing an independent requantification of miRNAs from patient samples, followed by application of the algorithm.

(Example 5)
Analysis of individual miRNA results from multivariate models and sera

The studies of Examples 1-4 show the ability of circulating miRNA to reliably distinguish endometriosis from other gynecological pathologies with robust diagnostic performance in independent test datasets of diverse (real-world-like) populations. Designed to demonstrate. Patients who have been surgically confirmed to have endometriosis have significantly lower expression of serum miRNA miR-3613-5p and let-7b compared to controls with various gynecological conditions. It was found that the expression of serum miRNAs miR-125b-5p, miR-150-5p, miR-451a and miR-342-3p was significantly high. The clinical features of this study population included diverse racial population statistics, early and late illnesses, and the presence of hormonal treatments, reflecting real-world patients with endometriosis. Control cases also included a wider variety of diseases than in our previous studies, with all women in the control group having different types of cysts (skin, ovary, paraovarian cyst, and serous or mucinous cystadenoma). ) Was diagnosed. Here, uterine fibroids (leiomyomas) were the major pathology seen in control patients (n = 23), and the second most common was the lack of abnormal pathology (Table 1). Evaluation of these markers in a cohort of patients with various pelvic pathologies supports the usefulness of using these markers in the general population to distinguish endometriosis from other conditions. ..

By evaluating the combination of miR-125b-5p, miR-451a, and miR-3613-5p using a logistic regression model, this combination was applied to this data to give an AUC of 0.8. Since this performance is only moderate, the data generated herein can be used in a random forest machine learning paradigm to build an optimal classifier based on the six miRNAs evaluated in the subject herein. However, when applied to an independent dataset from our previous study, this classifier yielded an AUC of 0.939 (Fig. 5). Endometriosis is not a life-threatening condition, so optimizing the classifier for specificity (avoiding false positives) can help prevent overdiagnosis, and women will relapse if symptoms persist. Can be inspected. Therefore, by selecting different cutoff points on the ROC curve, the 6-miRNA random forest model was optimized for specificity to give a model with 96% specificity and 83% sensitivity. Alternatively, different cutoffs on the ROC curve may be used to optimize both sensitivity and specificity values closer to 90%. Higher sensitivity (and lower false negative rates) may be appropriate for the use of biomarker panels as screening tests.

(Example 6)
Generation of algorithmic models trained based on the data derived herein

Key characteristics of the multi-mRNA expression data generated in previous examples by first applying a random forest and penalized linear regression approach to build an improved model for predicting endometriosis. It was determined.

For this procedure, 1000 replicas were generated, each containing a pair of training and test sets. The stratified bootstrap approach was used to maintain the same parts of the case and control in the bootstrap training set. For each replica, a training set was generated using the bootstrap method (resampling with replacement) and the unselected samples were used as the test set. Random division was repeated K = 1000 times.

Random forest (using tree = 500) or linear regression with penalties (alpha level = 0.4 (using elastic nets), 0.7 (using elastic nets), and 1 (using elastic nets), for each pair of datasets. LASSO); as well as the optimal lambda selected if the minimum MSE at each alpha level was included) was applied to the training set. Models constructed from each method were used to estimate the probabilities of having the disease and the resulting values were bisected with a cutoff of 0.5 (50%). Then, calculated weighted accuracy measurements such as AUC, specificity, sensitivity, and misclassification rate were calculated. The predicted disease status was compared to the actual disease status of the test set and training set. A total accuracy measurement, which is the sum of the estimated accuracy measurements from the training set and the (properly weighted) estimation accuracy from the test set, was calculated.

For each approach (random forest or penalized regression), the average of total accuracy and error measurements was reported for 1000 iterations. Similarly, the importance scale (for random forests) or the number of times each variable was selected out of 1000 (for penalized regression) was reported. These data are reported in Table 4.

Based on this data, logistic regression and random forest models were constructed using various numbers of markers selected using the RF importance ranking from the model building steps above. These models were then validated with retrospective data from Cosar et al. Fertility and sterility. 2016; 106 (2): 402-409. The data are presented in Tables 7 and 8.

Based on the data in Tables 7 and 8, Random Forest was determined to be a good training algorithm approach for classifying the data from this study (providing excellent results for all combinations of miRNA properties). Because).

Therefore, we applied a random forest learning approach to either this dataset or a dataset from a previous study and determined the cutoff to optimize the performance of the model. As an example of the demonstration, Figure 7 shows the distribution of voting percentages from the RF model of the retrospective dataset (Cosar); black bars are subjects with surgically defined endometriosis, white bars are uterus. Shown is a histogram showing subjects without endometriosis. The 43% diagnostic threshold (cutoff) shown by the vertical dashed line provides 96% specificity and 83% sensitivity for the RF model of this dataset.

The classifiers constructed from all the biomarkers used in the models herein (miR-125b, miR-451a, miR-3613, miR-150, miR-342, and let-7b) vary widely. We extended this analysis to determine if it could be generalized to a sample.

Importantly, the 6-marker model (D) showed high performance (greater than 0.93 AUC) in all datasets.

(Example 7)
Detection, diagnosis and treatment of endometriosis

Blood, plasma, serum, menstrual blood, menstrual effluent, urine, or saliva samples are taken from a female patient suspected of having endometriosis. The amount of microRNAs associated with endometriosis (eg, miR-125b, miR-451a, miR-3613, miR-150, miR-342, and let-7b) was then determined in the sample and described herein. Detect endometriosis using one of the trained algorithm classifiers used in the book. If endometriosis is detected, the patient is treated with a therapeutically effective dose of GnRH antagonist or agonist therapy (eg, Elagolix). The compound results in a reduction in the symptoms of endometriosis. After 1 month of treatment, 6 months of treatment, and 1 year of treatment, patients are evaluated for the level of microRNA signature associated with endometriosis. If the microRNA signature associated with endometriosis indicates the presence of endometriosis, the dose of GnRH agonist or antagonist therapy (eg, Elagolix) is adjusted upwards and the biomarker is non-endometriosis. Repeat the treatment / test process until present.

Although preferred embodiments of the invention are shown and described herein, it will be apparent to those skilled in the art that such embodiments are provided merely by way of illustration. One of ordinary skill in the art will recall a number of variants, modifications, and substitutions without departing from the invention. It should be appreciated that various alternatives to the embodiments of the invention described herein can be used in the practice of the invention. It is intended that the following claims define the scope of the invention, as well as the methods and structures within these claims, as well as their equivalents.

Claims (144)

A method of detecting the presence or absence of endometriosis in a female subject.
(A) A step of detecting the expression profile of a panel of miRNAs associated with endometriosis in a body fluid sample from the female subject, wherein the panel of miRNAs associated with endometriosis is miR-342. Or with steps, including miR451a,
(B) A step of applying a machine learning algorithm to the expression profile of the panel of the miRNA associated with endometriosis, wherein the machine learning algorithm has a materiality measure assigned to the properties of the miRNA.
i. The importance scale is assigned to miR-342 and the importance scale assigned to miR-342 is selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. Greater than the importance scale assigned to at least one miRNA, or ii. The importance scale is assigned to miR-451a and the importance scale assigned to miR-451a is assigned to at least one miRNA selected from the group consisting of miR-3613, miR-125b, and let-7b. Steps and steps that are greater than the importance scale
(C) A method comprising the step of detecting the presence or absence of endometriosis in the female subject using the machine learning algorithm. The method of claim 1, wherein the female subject has symptoms of endometriosis. The method of claim 1, wherein the female subject has symptoms of endometriosis selected from the group consisting of abdominal pain, gastrointestinal upset, excessive bleeding, infertility, and irregular menstruation. The method of any one of the preceding claims, wherein said subject has not previously been diagnosed with endometriosis. The method of any one of the preceding claims, wherein the subject has previously been diagnosed with endometriosis and the method confirms the presence of endometriosis in the female subject. The method of any one of the preceding claims, further comprising the step of diagnosing endometriosis in said female subject if the presence of endometriosis is detected. The method of any one of the preceding claims, further comprising the step of predicting or monitoring endometriosis in said female subject if the presence of endometriosis is detected. The method of any one of the preceding claims, further comprising the step of treating the female subject with endometriosis if the presence of endometriosis is detected. The method of any one of the preceding claims, further comprising the step of determining that the female subject has a non-endometriosis condition if the absence of endometriosis is detected. The method of any one of the preceding claims, further comprising the step of treating the female subject with a non-endometriotic condition if the absence of endometriosis is detected. The importance of the importance scale assigned to miR-342 assigned to at least one miRNA selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. The method of any one of the preceding claims, which is greater than the measure. 11. The method of claim 11, wherein the importance scale assigned to miR-342 is smaller than the importance scale assigned to at least one other miRNA. The importance of the importance scale assigned to miR-342 assigned to at least two miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. The method of any one of the preceding claims, which is greater than the measure. The importance of the importance scale assigned to miR-342 assigned to at least three miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. The method of any one of the preceding claims, which is greater than the measure. The importance of the importance scale assigned to miR-342 assigned to at least 4 miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. The method of any one of the preceding claims, which is greater than the measure. Any of the preceding claims in which the importance scale assigned to miR-342 is greater than the importance scale assigned to miR-150, miR-3613, miR-451a, let-7b, and miR-125b. The method described in one paragraph. Claim 1 the importance scale assigned to miR-451a is greater than the importance scale assigned to at least one miRNA selected from the group consisting of miR-3613, miR-125b and let-7b. The method described. Claim 1 the importance scale assigned to miR-451a is greater than the importance scale assigned to at least two miRNAs selected from the group consisting of miR-3613, miR-125b and let-7b. The method described. The method of claim 1, wherein the importance scale assigned to miR-451a is greater than the importance scale assigned to miR-3613, miR-125b, and let-7b. The importance scale is assigned such that the highest to lowest rankings of the importance scale are miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. The method according to any one of the claims. The method according to any one of the preceding claims, wherein the body fluid sample comprises cells. The method according to any one of the preceding claims, wherein the body fluid sample is a cell-free sample. The method according to any one of the preceding claims, wherein the body fluid sample is a blood sample, a plasma sample, a saliva sample, or a serum sample. The method of any one of the preceding claims, wherein the panel of miRNA is cell-free miRNA. The method of any one of the preceding claims, wherein the panel of miRNAs is a cell-associated miRNA or an exosome-associated miRNA. The step of applying the machine learning algorithm to the expression profile comprises applying a machine learning algorithm having a specific importance scale ranking assigned to the miRNA property, and the highest to lowest ranking is miR-. 342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150, the method of any one of the preceding claims. The method according to any one of the preceding claims, wherein the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. The method according to any one of the preceding claims, wherein the machine learning algorithm is a random forest algorithm. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a specificity greater than 80%. 29. The method of claim 29, wherein said population of women is premenopausal women. 29. The method of claim 29, wherein said population of women comprises women with smooth myoma, cystadenoma, chronic pelvic infections, teratomas, endometrial tumors, or paraovarian cysts. 29. The method of claim 29, wherein said population of women comprises women with stage I / II endometriosis. 29. The population of women comprises a woman with stage III / IV endometriosis or a woman with all four stages of endometriosis (stage I / II / III / IV). The method described. 29. The method of claim 29, wherein said population of women comprises women who received hormone therapy within 3 months from the date the sample was obtained, or women of any phase of the menstrual cycle. 29. The method of claim 29, wherein said population of women comprises a cohort comprising at least 100 women, at least 500 women, or at least 1000 women. The method of any one of the preceding claims, wherein the machine learning algorithm is trained with expression data from at least 100 samples, at least 500 samples, or at least 1000 samples. The method of any one of the preceding claims, wherein the machine learning algorithm is trained for a population of women, including women with stage I-IV endometriosis. A preceding claim having an AUC for detecting endometriosis greater than 0.85 in the female population, greater than 0.90 in the female population, or greater than 0.92 in the female population. The method described in any one of the above. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a specificity greater than 85%, greater than 90%, or greater than 95%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a sensitivity greater than 80%, greater than 85%, or greater than 90%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a specificity greater than 90% and a sensitivity of less than 85%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a sensitivity of less than 90%, less than 85%, or less than 80%. The method of any one of the preceding claims, further comprising the step of treating the female subject in a procedure without surgery if the absence of endometriosis is detected. If the presence of endometriosis is detected, it further comprises the step of treating the female subject, the treatment being hormonal treatment, surgery, laparoscopic surgery, statins, non-steroidal anti-inflammatory drug (NSAID). , Oral contraceptives, progestin, gonadotrophin release (GnRH) agonists, GnRH antagonists, androgen, anti-progesterone, selective estrogen receptor modulator (SERM), selective estrogen receptor modulator (SPRM), atorvastatin, seribastatin, flubustatin, lovastatin , The method of any one of the preceding claims comprising mevastatin, pitavastatin, pravastatin, rosbatatin, simvastatin, parasetamol, COX-2 inhibitor, or aspirin. The method of any one of the preceding claims, wherein the miRNA expression level is detected by quantitative real-time polymerase chain reaction (RT-PCR), microarray, sequencing, or next-generation sequencing. A method of classifying endometriosis in female subjects
(A) A step of obtaining a body fluid sample containing miRNA, wherein the body fluid sample is from a female subject.
(B) A step of performing a quantitative real-time polymerase chain reaction, microarray assay or sequencing assay on a set of miRNAs in the body fluid sample, wherein the set of miRNAs is associated with two endometriosis. Steps and steps that contain more or more different miRNAs,
(C) The amount of the two or more different miRNAs associated with endometriosis in the biological sample is higher than the two or more in the body fluid sample compared to the amount of control RNA. Steps to determine normalized miRNA levels for different miRNAs,
(D) A step of classifying the female subject as positive or negative for endometriosis by inputting the normalized miRNA level into a trained algorithm, wherein the trained algorithm is described. Weighting the importance rankings of two or more different miRNAs, the trained algorithm has the receiver operating characteristic (ROC) curve or the two or more different miRNAs associated with endometriosis. By selecting the optimal cutoff point in the voting percentage distribution of, the steps are optimized for higher specificity than sensitivity.
(E) A report identifying the female subject as either positive or negative for endometriosis based on the step of classifying the female subject as positive or negative for endometriosis in (d). Methods, including steps to output on the computer screen. 46. The method of claim 46, wherein the trained algorithm is optimized to detect endometriosis with a specificity greater than 80% in a female population. 46. The method of claim 46, wherein the trained algorithm is optimized to detect endometriosis in a female population with a specificity greater than 90% and a sensitivity less than 85%. 46. The method of claim 46, wherein the trained algorithm detects stage I / II endometriosis with a specificity greater than 90% or greater than 95% in a female population. The method of any one of the preceding claims, comprising an AUC for detecting endometriosis greater than 0.85 in a female population. The method of any one of the preceding claims, comprising an AUC for detecting endometriosis greater than 0.9 in a female population. The method of any one of the preceding claims, comprising an AUC for detecting endometriosis greater than 0.92 in a female population. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a specificity greater than 85%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a specificity greater than 90%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a specificity greater than 95%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a sensitivity greater than 80%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a sensitivity greater than 90%. The method of any one of the preceding claims, wherein endometriosis is detected in a female population with a sensitivity of less than 90%, less than 85%, or less than 80%. The method of any one of the preceding claims, wherein said population of women comprises at least 100 women. 59. The method of claim 59, wherein said population of at least 100 women comprises a woman with leiomyoma. The method of any one of the preceding claims, which has an area under the curve (AUC) value greater than 0.85, regardless of the stage of endometriosis or hormonal treatment. 46. The method of claim 46, further comprising treating the female subject with endometriosis after the report identifies the female subject as negative for endometriosis. 46. The method of claim 46, further comprising treating the female subject with a non-endometriotic condition after the report identifies the female subject as negative for endometriosis. The preceding claim further comprises repeating steps (a)-(e) for additional fluid samples obtained after at least 3 months, wherein the report identified the female subject as negative for endometriosis. The method described in any one of the above. The importance of the trained algorithm over the importance scale assigned to at least one miRNA selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. The method of any one of the preceding claims, wherein the scale is assigned to miR-342. The method of any one of the preceding claims, wherein the importance scale assigned to miR-342 is smaller than the importance scale assigned to at least one other miRNA. The importance of the importance scale assigned to miR-342 assigned to at least two miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. The method of any one of the preceding claims, which is greater than the measure. The importance of the importance scale assigned to miR-342 assigned to at least three miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, or miR-125b. The method of any one of the preceding claims, which is greater than the measure. The importance of the importance scale assigned to miR-342 assigned to at least 4 miRNAs selected from the group consisting of miR-150, miR-3613, miR-451a, let-7b, and miR-125b. The method of any one of the preceding claims, which is greater than the measure. Any of the preceding claims in which the importance scale assigned to miR-342 is greater than the importance scale assigned to miR-150, miR-3613, miR-451a, let-7b, and miR-125b. The method described in one paragraph. The preceding claim, wherein the importance scale assigned to miR-451a is greater than the importance scale assigned to at least one miRNA selected from the group consisting of miR-3613, miR-125b and let-7b. The method according to any one of the above. The preceding claim, wherein the importance scale assigned to miR-451a is greater than the importance scale assigned to at least two miRNAs selected from the group consisting of miR-3613, miR-125b and let-7b. The method according to any one of the above. The method according to any one of the preceding claims, wherein the importance scale assigned to miR-451a is greater than the importance scale assigned to miR-3613, miR-125b, and let-7b. The importance scale is assigned such that the ranking from highest to lowest of the importance scale is miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. , The method according to any one of the preceding claims. The method according to any one of the preceding claims, wherein the body fluid sample comprises cells. The method according to any one of the preceding claims, wherein the body fluid sample is a cell-free sample. The method according to any one of the preceding claims, wherein the body fluid sample is a blood sample, a plasma sample, a saliva sample, or a serum sample. The step of applying the machine learning algorithm to the expression profile comprises applying a machine learning algorithm having a specific importance scale ranking assigned to the miRNA property, and the highest to lowest ranking is miR-. 342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150, the method of any one of the preceding claims. The method according to any one of the preceding claims, wherein the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. The method according to any one of the preceding claims, wherein the machine learning algorithm is a random forest algorithm. The method of any one of the preceding claims, wherein said population of women is premenopausal women. In any one of the preceding claims, said population of women, in any combination, comprises women with leiomyomas, cystadenomas, chronic pelvic infections, teratomas, endometrial tumors, or paraovarian cysts. The method described. The method of any one of the preceding claims, wherein said population of women comprises women with stage I / II endometriosis. A prior claim, wherein said population of women comprises a woman with stage III / IV endometriosis, or a woman with all four stages of endometriosis (stage I / II / III / IV). The method described in any one of the above. The method of any one of the preceding claims, wherein said population of women comprises women who received hormone therapy within 3 months from the date the sample was obtained, or women of any phase of the menstrual cycle. .. The method of any one of the preceding claims, wherein said population of women comprises a cohort comprising at least 100 women. The method of any one of the preceding claims, wherein said population of women comprises a cohort comprising at least 500 women. The method of any one of the preceding claims, wherein said population of women comprises a cohort comprising at least 1000 women. The method of any one of the preceding claims, wherein the machine learning algorithm is trained with expression data from at least 100 samples. The method of any one of the preceding claims, wherein the machine learning algorithm is trained with expression data from at least 1000 samples. The method of any one of the preceding claims, wherein the machine learning algorithm is trained for a population of women, including women with stage I-IV endometriosis. The method of any one of the preceding claims, further comprising the step of treating the female subject in a procedure without surgery if the absence of endometriosis is detected. If the presence of endometriosis is detected, it further comprises the step of treating the female subject, the treatment being hormonal treatment, surgery, laparoscopic surgery, statins, non-steroidal anti-inflammatory drug (NSAID). , Oral contraceptives, progestin, gonadotrophin release (GnRH) agonists, GnRH antagonists, androgen, anti-progesterone, selective estrogen receptor modulator (SERM), selective estrogen receptor modulator (SPRM), atorvastatin, seribastatin, flubustatin, lovastatin , The method of any one of the preceding claims comprising mevastatin, pitavastatin, pravastatin, rosbatatin, simvastatin, parasetamol, COX-2 inhibitor, or aspirin. A method of diagnosing and treating endometriosis in female subjects.
(A) A step of detecting the expression profile of a panel of miRNAs associated with endometriosis in a saliva sample from the female subject, wherein the panel of miRNAs associated with endometriosis is miR-125b. And a step containing at least one other miRNA,
(B) A step of applying a machine learning algorithm to the expression profile of the panel of the miRNA associated with endometriosis, wherein the machine learning algorithm has a materiality measure assigned to the properties of the miRNA. The step and step, wherein the importance scale of miR-125b is greater than the importance scale of miR-150, miR-3613, miR-451a, let-7b, or miR-342.
(C) A step of diagnosing endometriosis of the female subject using the machine learning algorithm.
(D) A method comprising treating the endometriosis diagnosed in the female subject with a procedure for endometriosis. The step of applying the machine learning algorithm to the expression profile comprises applying the machine learning algorithm having a specific importance scale assigned to the property of the miRNA, and the importance scale of miR-125b is miR-. The method of claim 94, which is greater than at least one of 150, let-7b, miR-451a, or miR-3613. The step of applying the machine learning algorithm to the expression profile comprises applying the machine learning algorithm having a specific importance scale assigned to the property of the miRNA, and the importance scale of miR-125b is miR-. The method of claim 94, which is greater than at least two of 150, let-7b, miR-451a, or miR-3613. The step of applying the machine learning algorithm to the expression profile comprises applying the machine learning algorithm having a specific importance scale assigned to the property of the miRNA, and the importance scale of miR-125b is miR-. The method of claim 94, which is greater than at least three of 150, let-7b, miR-451a, or miR-3613. The step of applying the machine learning algorithm to the expression profile comprises applying the machine learning algorithm having a specific importance scale assigned to the property of the miRNA, and the importance scale of miR-125b is miR-. The method of claim 94, which is greater than 150, let-7b, miR-451a, and miR-3613. A method of characterizing a female subject as having endometriosis,
(A) A step of obtaining a body fluid sample containing miRNA, wherein the body fluid sample is from a female subject.
(B) A step of performing a quantitative real-time polymerase chain reaction, microarray assay, or sequencing of a set of miRNAs in the body fluid sample, wherein the set of miRNAs is two or the same associated with endometriosis. With steps, which contain more different miRNAs,
(C) The amount of the two or more different miRNAs associated with endometriosis in the biological sample is higher than the two or more in the body fluid sample compared to the amount in the control RNA. Steps to determine normalized miRNA levels for different miRNAs,
(D) The step of classifying the female subject as positive or negative for endometriosis by inputting the normalized miRNA level into a trained algorithm, wherein the trained algorithm is the uterus. By selecting the optimal cutoff point in the recipient motion characteristic (ROC) curve of the two or more different miRNAs associated with endometriosis, the sensitivity is optimized and trained to at least 80%. The algorithm weights the importance rankings of the two or more different miRNAs, with the step.
(E) A report identifying the female subject as either positive or negative for endometriosis based on the step of classifying the female subject as positive or negative for endometriosis in (d). Methods, including steps to output on the computer screen. A method of characterizing a female subject as having no endometriosis,
(A) A step of obtaining a body fluid sample containing miRNA, wherein the body fluid sample is from a female subject having abdominal pain, gastrointestinal upset, excessive bleeding, infertility or irregular menstruation. But, with steps that have not been previously diagnosed with endometriosis,
(B) The step of performing a quantitative real-time polymerase chain reaction or sequencing of a set of miRNAs in the body fluid sample, wherein the set of miRNAs differs by two or more associated with endometriosis. With steps, including miRNA,
(C) The amount of the two or more different miRNAs associated with endometriosis in the biological sample is higher than the two or more in the body fluid sample compared to the amount in the control RNA. Steps to determine normalized miRNA levels for different miRNAs,
(D) A step of classifying the female subject as positive or negative for endometriosis by inputting the normalized miRNA level into a trained algorithm, wherein the trained algorithm is the uterus. By selecting the optimal cutoff point in the receiver operating characteristic (ROC) curves of the two or more different miRNAs associated with endometriosis, the training is optimized for at least 80% specificity. The algorithm used to weight the importance rankings of the two or more different miRNAs, with the step.
(E) A report identifying the female subject as either positive or negative for endometriosis based on the step of classifying the female subject as positive or negative for endometriosis in (d). Methods, including steps to output on the computer screen. The method of claim 100, wherein the trained algorithm calculates a voting score indicating the likelihood of having endometriosis. The method of claim 100, wherein the trained algorithm classifies the female subject as having endometriosis if the voting score is greater than 36%. The method of claim 100, which has an area under the curve (AUC) value greater than 0.85 for stage I / II endometriosis. The method of claim 100, which has an area under the curve (AUC) value greater than 0.85 for distinguishing between endometriosis and leiomyoma. 100. The method of claim 100, which has an area under the curve (AUC) value greater than 0.85, regardless of the stage of endometriosis or hormonal treatment. 100. The method of claim 100, optimized for specificity greater than 90% and sensitivity less than 85%. A method for detecting endometriosis in female subjects
(A) A step of detecting the expression profile of a panel of miRNAs associated with endometriosis in a sample containing miRNAs from said subject.
(B) A step of applying a machine learning algorithm to the expression profile to detect endometriosis in the sample from the subject, wherein the machine learning algorithm
(I) miR-342, miR-451a, and miR-3613,
(Ii) miR-342, miR-451a, miR-3613, and miR-125b,
(Iii) miR-342, miR-451a, miR-3613, miR-125b, and let-7b,
(Iv) miR-342, miR-451a, let-7b, and miR-125b,
(V) miR-342, miR-451a, let-7b, and miR-3613,
(Vi) miR-342, miR-451a, and let-7b,
(Vii) miR-125b, miR-150, miR-342, miR-451a, and let-7b, and (viii) miR-125b, miR-150, miR-342, miR-3613, miR-451a, and let. -7b
A method comprising steps, which is trained on a group of miRNA properties selected from the group consisting of. 10. The method of claim 107, further comprising the step of obtaining the sample containing miRNA from the subject prior to (a). The method of claim 107 or 108, wherein the sample is a saliva sample. The method of claim 107 or 108, wherein the sample is a serum sample. The method according to any one of the preceding claims, wherein the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. The method of any one of the preceding claims, comprising an AUC for detecting endometriosis greater than 0.85 in a female population. The method of claim 112, wherein said population of women is premenopausal women. The method of claim 112, wherein said population of women is premenopausal and is older than 18 years of age. The method of claim 112, wherein said population of women is premenopausal and under 49 years of age. The method of any one of the preceding claims, wherein said population of women is negative for severe anemia, hyperplasia, polyps, and malignant tumors. 12. The method of claim 112, wherein the population of women comprises women who received hormone therapy within 3 months from the date the sample was obtained. 17. The method of claim 117, wherein the hormone therapy comprises an oral contraceptive or a GnRH agonist. The method of any one of the preceding claims, wherein the machine learning algorithm is trained with expression data of at least 100 samples. The method of any one of the preceding claims, wherein the machine learning algorithm is trained on a population of women with surgically confirmed endometriosis. The method of any one of the preceding claims, wherein the machine learning algorithm is trained for a population of women, including women with stage I or II endometriosis. The method of any one of the preceding claims, wherein the machine learning algorithm is trained for a population of women, including women with stage I-IV endometriosis. The step of applying the machine learning algorithm to the expression profile comprises applying the machine learning algorithm having a specific importance scale assigned to the property of the miRNA, and the importance scale of miR-342 is miR-. The method of any one of the preceding claims, greater than at least one of 150, let-7b, or miR-125. The step of applying the machine learning algorithm to the expression profile comprises applying the machine learning algorithm having a specific importance scale assigned to the property of the miRNA, and the importance scale of miR-125b is miR-. The method of any one of the preceding claims, which is greater than one of 150, let-7b, miR-451a, or miR-3613. The steps of applying the machine learning algorithm to the expression profile are the steps of training the machine learning algorithm for a group of (viii) miRNA properties and machine learning with a specific importance measure assigned to the miRNA properties. The ranking from highest to lowest of the properties of the miRNA, including the step of applying the algorithm, is miR-125b, let-7b, miR-3613, miR-150, miR-342, and miR-451a. The method according to any one of the preceding claims. The step of applying the machine learning algorithm to the expression profile is the step of training the machine learning algorithm for the group of miRNA properties of (viii) and the machine having a specific importance scale ranking assigned to the miRNA properties. The preceding claim comprises the steps of applying the learning algorithm, wherein the ranking from highest to lowest is miR-342, miR-451a, miR-3613, miR-125b, let-7b, and miR-150. The method described in any one of the above. The steps of applying the machine learning algorithm to the expression profile are the steps of training the machine learning algorithm for a group of (viii) miRNA properties and machine learning with a specific importance measure assigned to the miRNA properties. The method according to any one of the preceding claims, comprising the step of applying an algorithm, wherein the importance scale ranking of the characteristics of the miRNA follows any column of Table 9. (A) Steps to store information related to the condition of a female patient in a standardized format on multiple network-based non-transient storage devices.
(B) A step of providing remote access to a user over a network so that at least one user can update the information related to the female patient's condition in real time via a graphical user interface. The step and the step, wherein the at least one user provides the updated information in the form of a miRNA expression profile from the female patient.
(C) Using a machine learning algorithm application to transform the expression profile of the miRNA from the female patient into the possibility that the female patient has endometriosis by means of a content server. ,
(D) A step of remembering the possibility that the female patient has endometriosis, and
(E) Each time the updated information is stored, the content server automatically generates a message containing the possibility that the female patient has endometriosis.
(F) Each user sends the message in real time to all of the users over the computer network so that the female patient has immediate access to the possibility of having endometriosis. And including methods. The machine learning algorithm
(I) miR-342, miR-451a, and miR-3613,
(Ii) miR-342, miR-451a, miR-3613, and miR-125b,
(Iii) miR-342, miR-451a, miR-3613, miR-125b, and let-7b,
(Iv) miR-342, miR-451a, let-7b, and miR-125b,
(V) miR-342, miR-451a, let-7b, and miR-3613,
(Vi) miR-342, miR-451a, and let-7b,
(Vii) miR-125b, miR-150, miR-342, miR-451a, and let-7b, and (viii) miR-125b, miR-150, miR-342, miR-3613, miR-451a, and let. -7b
The method of claim 128, wherein the group is trained on a group of miRNA properties selected from the group consisting of. The method of claim 128 or 129, further comprising the step of obtaining a sample containing miRNA from said subject. The method of claim 130, wherein the sample is a blood sample, plasma sample, or serum sample. The method of claim 130, wherein the sample is a saliva sample. The method of claim 130, wherein the sample is a serum sample. The method according to any one of the preceding claims, wherein the machine learning algorithm is a random forest algorithm, k-nearest neighbor algorithm (KNN), support vector machine (SVM), or naive bays. The method of any one of the preceding claims, comprising an AUC for detecting endometriosis greater than 0.85 in a female population. The method of any one of the preceding claims, wherein said population of women comprises women who received hormone therapy within 3 months from the date the sample was obtained. The method of any one of the preceding claims, wherein the hormone therapy comprises an oral contraceptive or a GnRH agonist. The method according to any one of the preceding claims, wherein the sample is a cell-free serum sample. The method according to any one of the preceding claims, wherein the sample is a cell-free saliva sample. The method of any one of the preceding claims, further comprising treating the subject to treat the endometriosis based on the possibility reported in (e). The method of claim 140, wherein the treatment comprises a hormonal treatment, a statin, or a non-steroidal anti-inflammatory drug (NSAID). The method of claim 140, wherein the hormone treatment comprises an oral contraceptive, a progestin, a GnRH agonist, a GnRH antagonist, an androgen, an antiprogesterone, SERM, or SPRM. The method of claim 140, wherein the statin comprises atorvastatin, cerivastatin, fluvastatin, lovastatin, mevastatin, pitavastatin, pravastatin, rosuvastatin, or simvastatin. The method of claim 140, wherein the NSAID comprises paracetamol, a COX-2 inhibitor, or aspirin.

Images