Glucagon-like peptide-1 receptor related biomarkers and uses thereof

Abstract

Provided are methods and systems for detecting a polymorphism in one or more glucagon-like protein-1 receptor (GLP-1R) related markers, and optionally detecting one or more metabolic markers or neural markers, in a subject. Also provided are methods and systems for predicting a risk for developing GLP-1R associated condition or a response to a GLP-1R agonist therapy in a subject by detecting a polymorphism in one or more GLP-1R related markers, and optionally detecting one or more metabolic markers or neural markers. The methods and systems can be used for selecting a treatment regimen, monitoring a response to the treatment, or adjusting the treatment regimen in the subject.

Description

Docket No.107976.00039 GLUCAGON-LIKE PEPTIDE-1 RECEPTOR RELATED BIOMARKERS AND USES THEREOF TECHNICAL FIELD

[0001] The present disclosure relates generally to the field of evaluating disease risk and / or response to treatment, and providing decision support for appropriate treatment regimens. CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application claims priority to U.S. provisional application No.63 / 655,529, filed on June 3, 2024, the entire contents of which are hereby incorporated by reference in this application. BACKGROUND

[0003] Glucagon-like peptide-1 (GLP-1) is an endogenous incretin hormone that enhances the secretion of insulin in response to eating and inhibits the release of glucagon, thereby helping to regulate blood glucose levels. GLP-1 is released by the gut after eating and reduce blood sugar and energy intake by activating the GLP-1 receptor (GLP-1R).

[0004] GLP-1R agonists, interchangeably referred to as GLP-1 agonists, are a class of drugs that mimic the actions of GLP-1. GLP-1R agonists work by activating the GLP-1 receptor. GLP-1R agonists slow gastric emptying, inhibit the release of glucagon, and stimulate insulin production, thereby reducing hyperglycemia in people with type 2 diabetes (T2D). GLP-1R agonists also reduce food intake and therefore body weight, making them an effective treatment for obesity. GLP-1R agonists were initially developed for treatment of type 2 diabetes. The 2022 American Diabetes Association standards of medical care recommend GLP-1R agonists as a first line therapy for type 2 diabetes, specifically in patients with atherosclerotic cardiovascular disease or obesity. Additionally, some GLP-1R agonists have been approved to treat obesity in the absence of diabetes.

[0005] GLP-1R agonists are an important advancement in the management of GLP-1R associated conditions (such as type 2 diabetes and obesity), offering multiple benefits beyond glucose control. However, there is considerable heterogeneity in how well patients respond to GLP-1R agonist treatment, with some subjects having a marked response and others having no improvement. Further, genetic factors are heavily associated with glucose homeostasis and diseases such as type 2 diabetesDocket No.107976.00039 and obesity. Accordingly, there is a need for efficiently and accurately determining each patient’s susceptibility to GLP-1 receptor associated condition and / or response to GLP-1R agonist therapy. SUMMARY

[0006] Genetic factors are important determinants of glucose homeostasis and type 2 diabetes susceptibility. Heritability of both fasting glucose and type 2 diabetes is at 35–40% and 30–60%, respectively. Identifying genetic markers related to the GLP-1R and their association with type 2 diabetes, obesity, or other GLP-1R associated conditions can help understand individual variability in the risk for developing GLP-1R associated condition and / or individual variability in the response to GLP-1R agonists, and potentially provide personalized treatment or prevention approaches.

[0007] Applicant recognized the problems noted above and has conceived and developed systems and associated methods for predicting risk of a GLP-1R associated condition or response to a GLP- 1R therapy, and providing decision support for appropriate treatment regimens.

[0008] In one aspect of the present disclosure, a method for detecting a polymorphism in one or more GLP-1R related markers in a subject is provided. The method includes genotyping a sample obtained from the subject to detect the presence of the polymorphism in the one or more GLP-1R related markers. The one or more GLP-1R related markers include one or more of rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467.

[0009] In some embodiments, the marker rs7997012 is located at position rs7997012 in the 5- HTR2A gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs453 is located at position rs4532 in the DRD1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1799971is located atDocket No.107976.00039 position rs1799971 in the OPRM1 gene; the marker rs9479757 is located at position rs9479757 in the MUOR gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs10305492 is located at position rs10305492 in the GLP-1R gene; the marker rs10305420 is located at position rs10305420 in the GLP-1R gene; the marker rs2268639 is located at position rs2268639 in the GLP-1R gene; the marker rs2268640 is located at position rs2268640 in the GLP-1R gene; the marker rs57922 is located at position rs57922 in the GLP-1R gene; the marker rs58428187 is located at position rs58428187 in the GLP-1R gene; the marker rs6923761 is located at position rs6923761 in the GLP-1R gene; the marker rs9299870 is located at position rs9299870 in the GLP-1R gene; the marker rs140226575 is located at position rs140226575 in the ARRB1 gene; the marker rs78979036 is located at position rs78979036 in the GLP-1R gene; the marker rs78052828 is located at position rs78052828 in the GLP-1R gene; the marker rs77501730 is located at position rs77501730 in the G6PC2 gene; the marker rs77501730 is located at position rs56100844 in the G6PC2 gene; the marker rs76895963 is located at position rs76895963 in the CCND2 gene; the marker rs115545608 is located at position rs115545608 in the G6PC2 gene; the marker rs16856115 is located at position rs16856115 in the G6PC2gene; the marker rs73174306 is located at position rs73174306 in the MECOM gene; the marker rs11708067 is located at position rs11708067 in the ADCY5 gene; the marker rs138917529 is located at position rs138917529 in the GCK gene; the marker rs1514895 is located at position rs1514895 in the SLC2A2 gene; the marker rs2168101 is located at position rs2168101 in the LM01 gene; the marker rs34222465 is located at position rs34222465 in the CACNA2D3 gene; the marker rs348330 is located at position rs348330 in the ABCB10 gene; the marker rs9379084 is located at position rs9379084 in the RREB1 gene; the marker rs10830963 is located at position rs10830963 in the MTNR1B gene; the marker rs183606969 is located at position rs183606969 in the GCK gene; the marker rs12692596 is located at position rs12692596 in the RBMS1 gene; the marker rs8192556 is located at position rs8192556 in the NEUROD1 gene; the marker rs118126621 is located at position rs118126621 in the ARMC2 and / or SESN1 gene; the marker rs78444298 is located at position rs78444298 in the EDEM3 gene; the marker rs35889227 is located at position rs35889227 in the FOXN3 gene; the marker rs11257655 is located at position rs11257655 in the CDC123 and / or CAMK1D gene; the marker rs6538804 is located at position rs6538804 in the RMST gene; the marker rs2255805 is located at position rs2255805 in the TSHZ2 gene; the marker rs17168486 is located at position rs17168486 in the DGKB and / or AGMO genes; the marker rs115128825 is located at position rs115128825 in the G6PC2 gene; the marker rs34814128 is located at position rs34814128Docket No.107976.00039 in the OR4C5 and / or OR4A47 genes; the marker rs10501320 is located at position rs10501320 in the MADD gene; the marker rs11592309 is located at position rs11592309 in the ADRA2A gene; the marker rs3842753 is located at position rs3842753 in the INS gene; the marker rs231362 is located at position rs231362 in the KCNQ1 gene; and / or the marker rs3765467 is located at position rs3765467 in the GLP-1R gene, in a human genome.

[0010] In some embodiments, the method further includes detecting a status of one or more metabolic markers, and / or a polymorphism in one or more neural response markers, and / or a polymorphism in one or more additional GLP-1R related markers in the subject. Certain embodiments include detecting any combinations of the foregoing. In certain embodiments, the one or more metabolic markers include one or more of age, sex, a lipid species profile, blood glucose level, blood hemoglobin A1c level, blood pressure, and waist circumference. In further embodiments, the one or more neural response markers include one or more of rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757. In certain embodiments, the one or more additional GLP-1R related markers can be located in one or more of the following genes: ANKK, FASN, GRIK1, HTR2C, INSR, LEPR, PCSK1, PLXNA4, POMC, TAAR1, TCF7L2.

[0011] In one aspect of the present disclosure, a method of predicting a risk for developing GLP- 1R associated condition or a response to a GLP-1R agonist therapy in a subject is provided. The method includes detecting a polymorphism in one or more GLP-1R related markers in a sample obtained from the subject; providing a risk score based on the polymorphism in the one or more GLP- 1R related markers; and selecting the subject as a candidate for a treatment regimen based on the risk score. The risk score indicates the risk of developing the GLP-1R associated condition and / or the likelihood of a poor response to a GLP-1R agonist therapy. The one or more GLP-1R related markers include one or more of rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467.Docket No.107976.00039

[0012] In some embodiments, the GLP-1R associated condition is one or more of type 2 diabetes, obesity, atherosclerotic cardiovascular disease, non-alcoholic fatty liver disease, polycystic ovary syndrome, food addiction, and addiction to cocaine, amphetamine, opioids, alcohol, or nicotine, or other substances that can lead to an abuse disorder.

[0013] In some embodiments, the marker rs7997012 is located at position rs7997012 in the 5- HTR2A gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs453 is located at position rs4532 in the DRD1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1799971is located at position rs1799971 in the OPRM1 gene; the marker rs9479757 is located at position rs9479757 in the MUOR gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs10305492 is located at position rs10305492 in the GLP-1R gene; the marker rs10305420 is located at position rs10305420 in the GLP-1R gene; the marker rs2268639 is located at position rs2268639 in the GLP-1R gene; the marker rs2268640 is located at position rs2268640 in the GLP-1R gene; the marker rs57922 is located at position rs57922 in the GLP-1R gene; the marker rs58428187 is located at position rs58428187 in the GLP-1R gene; the marker rs6923761 is located at position rs6923761 in the GLP-1R gene; the marker rs9299870 is located at position rs9299870 in the GLP-1R gene; the marker rs140226575 is located at position rs140226575 in the ARRB1 gene; the marker rs78979036 is located at position rs78979036 in the GLP-1R gene; the marker rs78052828 is located at position rs78052828 in the GLP-1R gene; the marker rs77501730 is located at position rs77501730 in the G6PC2 gene; the marker rs77501730 is located at position rs56100844 in the G6PC2 gene; the marker rs76895963 is located at position rs76895963 in the CCND2 gene; the marker rs115545608 is located at position rs115545608 in the G6PC2 gene; the marker rs16856115 is located at position rs16856115 in the G6PC2gene; the marker rs73174306 is located at position rs73174306 in the MECOM gene; the marker rs11708067 is located at position rs11708067 in the ADCY5 gene; the marker rs138917529 is located at position rs138917529 in the GCK gene; the marker rs1514895 is located at position rs1514895 in the SLC2A2 gene; the marker rs2168101 is located at position rs2168101 in the LM01 gene; the markerDocket No.107976.00039 rs34222465 is located at position rs34222465 in the CACNA2D3 gene; the marker rs348330 is located at position rs348330 in the ABCB10 gene; the marker rs9379084 is located at position rs9379084 in the RREB1 gene; the marker rs10830963 is located at position rs10830963 in the MTNR1B gene; the marker rs183606969 is located at position rs183606969 in the GCK gene; the marker rs12692596 is located at position rs12692596 in the RBMS1 gene; the marker rs8192556 is located at position rs8192556 in the NEUROD1 gene; the marker rs118126621 is located at position rs118126621 in the ARMC2 and / or SESN1 gene; the marker rs78444298 is located at position rs78444298 in the EDEM3 gene; the marker rs35889227 is located at position rs35889227 in the FOXN3 gene; the marker rs11257655 is located at position rs11257655 in the CDC123 and / or CAMK1D gene; the marker rs6538804 is located at position rs6538804 in the RMST gene; the marker rs2255805 is located at position rs2255805 in the TSHZ2 gene; the marker rs17168486 is located at position rs17168486 in the DGKB and / or AGMO genes; the marker rs115128825 is located at position rs115128825 in the G6PC2 gene; the marker rs34814128 is located at position rs34814128 in the OR4C5 and / or OR4A47 genes; the marker rs10501320 is located at position rs10501320 in the MADD gene; the marker rs11592309 is located at position rs11592309 in the ADRA2A gene; the marker rs3842753 is located at position rs3842753 in the INS gene; the marker rs231362 is located at position rs231362 in the KCNQ1 gene; and / or the marker rs3765467 is located at position rs3765467 in the GLP-1R gene, in a human genome.

[0014] In some embodiments, the risk score is provided using a risk model. In some embodiments, the risk model includes a machine learning model. In certain embodiments, the machine learning model integrates heterogenous data modalities to produce a holistic subject profile, generates an individualized treatment path, classifies a subject into a behavioral phenotype or a drug response phenotype matched to a treatment regimen, adapts to temporal trends through intake of longitudinal subject data, and / or produces outputs, such as payer-facing risk stratifications. In certain embodiments, the heterogenous data includes genotyping results, metabolic biomarker status, behavioral data, and environmental data. In certain embodiments, the outputs include one or more of ranked treatment options, dosing suggestions, behavioral intervention parings, and / or monitoring frequency guidance.

[0015] In some embodiments, the method includes providing an individual risk score for each of a plurality of polymorphisms in the one or more GLP-1R related markers, thereby providing individual risk scores; and providing the risk score as a composite of the individual risk scores. In some embodiments, the individual risk scores are weighted based on the risk of developing type 2 diabetesDocket No.107976.00039 or food addiction, to provide the risk score as the composite of the individual risk scores that reflects disease susceptibility. In some embodiments, the individual risk scores are weighted based on the likelihood of response to GLP-1 agonists, or the likelihood of having a poor response to the GLP- 1R agonist therapy. Methods include providing the risk score as the composite of the individual risk scores that reflect therapeutic response potential. In some embodiments, the individual risk scores are weighted based on the risk of developing type 2 diabetes or food addiction, and optionally also on the likelihood of responding to GLP-1R agonist therapy. The resulting composite risk score can reflect disease susceptibility, therapeutic response potential, or both.

[0016] In some embodiments, the risk score being greater than a predetermined threshold indicates the subject has an increased risk for developing type 2 diabetes or food addiction, or having a poor response to the GLP-1R agonist therapy relative to a reference population. In some embodiments, the risk score being lesser than the predetermined threshold indicates the subject has a reduced risk for developing type 2 diabetes or food addiction, or having a poor response to the GLP-1R agonist therapy relative to a reference population.

[0017] In some embodiments, the method further includes determining a status of one or more metabolic markers, a polymorphism in one or more neural response markers, and / or a polymorphism in one or more additional GLP-1R related markers in the subject; and providing the risk score based on the polymorphism in the one or more GLP-1R related markers, the status of the one or more metabolic markers, the polymorphism in the one or more neural response markers, the polymorphism in one or more additional GLP-1R related markers, or combinations of two or more of the foregoing. In some embodiments, the method further includes administering a treatment regimen to the subject. In some embodiments, the one or more metabolic markers include one or more of age, sex, a lipid species profile, blood glucose level, blood hemoglobin A1c level, blood pressure, and waist circumference. In some embodiments, the status of the one or more metabolic markers includes data from continuous monitoring of blood glucose level in the subject. In some embodiments, the status of the one or more metabolic markers includes data from continuous monitoring of blood hemoglobin A1c level in the subject.

[0018] In embodiments, the one or more neural response markers include one or more of rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757. In certain embodiments, the additional GLP-1R related markers are one or more of ANKK, FASN, GRIK1, HTR2C, INSR, LEPR, PCSK1, PLXNA4, POMC, TAAR1, TCF7L2.Docket No.107976.00039

[0019] In embodiments, the marker rs7997012 is located at position rs7997012 in the 5-HTR2A gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs4532 is located at position rs4532 in the DRD1 gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1799971 is located at position rs1799971 in the OPRM1 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; and the marker rs9479757 is located at position rs9479757 in the MUOR gene, in the human genome.

[0020] In some embodiments, the treatment regimen includes a GLP-1R agonist therapy, a DPP-4 inhibitor therapy, a SGLT2 inhibitor therapy, an anti-obesity agent therapy, a diet therapy, an exercise therapy, insulin therapy, oral diabetic medication, cognitive therapy, behavioral therapy, or combinations of two or more of the foregoing. In certain embodiments, the GLP-1R agonist therapy includes one or more of semaglutide, liraglutide, dulaglutide, exenatide, or tirzepatide. In certain embodiments, the DPP-4 inhibitor therapy includes one or more of sitagliptin, saxagliptin, alogliptin, or linagliptin. In certain embodiments, the SGLT2 inhibitor therapy includes one or more of empagliflozin or canagliflozin. In certain embodiments, an anti-obesity agent therapy includes one or more of orlistat, phentermine-topiramate, naltrexone-bupropion, liraglutide, semaglutide, tirzepatide, and setmelanotide

[0021] In some embodiments, the method further includes providing a status of the one or more metabolic markers in the subject at two or more time points to provide a series of metabolic marker status, and monitoring the risk for developing the GLP-1R associated condition, monitoring a response to the treatment regimen, and / or adjusting or changing the treatment regimen in the subject based on the risk score and the series of metabolic marker status. In some embodiments, the polymorphism includes a single nucleotide polymorphism (SNP). In some embodiments, the method provided herein is implemented by a computer. In certain aspects, provided herein is a non-transitory machine-readable storage medium storing processor-executable instructions that, when executed by at least one processor, cause the at least one processor to perform the method provided herein.Docket No.107976.00039

[0022] Embodiments include systems for predicting a risk for developing a GLP-1R associated condition or a response to a GLP-1R agonist therapy in a subject. One such system includes a polymorphism analyzer to analyze a sample from the subject and to provide a first plurality of polymorphism profiles of each of a set of specified allelic variants in the sample; and a processor and a machine-readable storage medium storing (i) a trained ensemble model, the trained ensemble model being trained with inputs that include a second plurality of polymorphism profiles associated with the specified set of allelic variants from a first plurality of test subjects and a third plurality of polymorphism profiles associated with the specified set of allelic variants from a second plurality of test subjects and outputs that include a first plurality of outputs specifying that the first plurality of test subjects have been diagnosed with the GLP-1R associated condition and / or having a good response to a GLP-1R agonist therapy, and a second plurality of outputs specifying that the second plurality of test subjects have not been diagnosed with a GLP-1R associated condition and / or having a poor response to a GLP-1R agonist therapy, and (ii) instructions, when executed by the processor, configured to: provide the trained ensemble model with the first plurality of polymorphism profiles to generate a score indicative of risk of developing a GLP-1R associated condition or responding poorly to a GLP-1R agonist therapy in the subject based on the first plurality of polymorphism profiles, in response to the score based on the first plurality of polymorphism profiles being greater than a pre-determined threshold, transmit to a user interface an output indicating a higher likelihood of the subject to develop the GLP-1R associated condition or a higher likelihood of the subject to respond to the GLP-1R agonist therapy, and in response to the score based on the first plurality of polymorphism profiles being less than or equal to the pre-determined threshold, transmit to the user interface an output indicating a lower likelihood of the subject to develop the GLP-1R associated condition or a lower likelihood of the subject to respond to the GLP-1R agonist therapy.

[0023] In some embodiments, the machine-readable storage medium further has instructions to: in response to the score based on the first plurality of polymorphism profiles being greater than the pre-determined threshold, retrieve a treatment regimen recommendation from the database; and transmit to the user interface an output indicating the higher likelihood of the subject to develop the GLP-1R associated condition or the higher likelihood of the subject to respond to the GLP-1R agonist therapy, and the treatment regimen recommendation.Docket No.107976.00039

[0024] In some embodiments, the trained ensemble model is generated by at least one of a random tree model and a support vector machine model.

[0025] In some embodiments, the first plurality of polymorphism profiles comprise polymorphism in one or more of rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467. In some embodiments, the first plurality of polymorphism profiles further comprise polymorphism in one or more of GLP-1R, ANKK, FASN, GRIK1, HTR2C, INSR, LEPR, PCSK1, PLXNA4, POMC, TAAR1, and TCF7L2.

[0026] In certain embodiments, the system further includes a database storing a treatment regimen recommendation. In certain embodiments, the system further includes a device to collect and prepare a sample for the polymorphism analysis.

[0027] Other aspects and features of the present disclosure will become apparent to those of ordinary skill in the art after reading the detailed description herein and the accompanying figures. BRIEF DESCRIPTION OF DRAWINGS

[0028] The foregoing aspects, features, and advantages of the present disclosure will be further appreciated when considered with reference to the following drawing:

[0029] FIGs. 1A through 1E are diagrammatic representations of a stacked machine learning model, according to an embodiment of the present disclosure.

[0030] FIGs.2A through 2C are diagrammatic representations of systems containing a stacked machine learning model, according to an embodiment of the present disclosure.

[0031] FIGs.3A through 3E are flowcharts of different embodiments of methods of predicting risk of developing GLP-1R associated condition or poor response to GLP-1R agonist therapy.

[0032] FIG.4 is an illustration of the effect of GLP-1R agonists on the mesolimbic reward pathways and reward-related behavior.Docket No.107976.00039

[0033] FIG. 5 schematically depicts a clinical decision pipeline based on detection of GLP-1R related markers, according to embodiments of the present disclosure.

[0034] FIG. 6 is a flowchart associated with a clinical decision support system for GLP-1R associated conditions, according to embodiments of the present disclosure. DETAILED DESCRIPTION

[0035] The foregoing aspects, features, and advantages of the present disclosure will be further appreciated when considered with reference to the following description of the embodiments and accompanying drawings. In describing the embodiments of the disclosure illustrated in the appended drawings, specific terminology will be used for the sake of clarity. The disclosure, however, is not intended to be limited to the specific terms used, and it is to be understood that each specific term includes equivalents that operate in a similar manner to accomplish a similar purpose. Numerous specific details, examples, and embodiments are set forth and described to provide a thorough understanding of various embodiments of the present disclosure. However, in certain instances, well- known or conventional details are not described in order to provide a concise discussion of embodiments of the present disclosure.

[0036] When introducing elements of various embodiments of the present disclosure, the articles “a,” “an,” “the,” and “said” are intended to mean that there are one or more of the elements. The terms “comprising,” “including,” and “having” are intended to be inclusive and mean that there may be additional elements other than the listed elements. Any examples of operating parameters and / or environmental conditions are not exclusive of other parameters / conditions of the disclosed embodiments. Additionally, it should be understood that references to “one embodiment”, “an embodiment,” “certain embodiments,” or “other embodiments” of the present disclosure are not intended to be interpreted as excluding the existence of additional embodiments that also incorporate the recited features.

[0037] As used herein, “SNP” or “single nucleotide polymorphism” means a sequence variation that occurs when a single nucleotide (A, T, C, or G) in the genome sequence is altered or variable.

[0038] As used herein, “marker,” or “molecular marker,” or “marker locus” is a term used to denote a nucleic acid or amino acid sequence that is sufficiently unique to characterize a specific locus on the genome.Docket No.107976.00039

[0039] As used herein, a centimorgan (“cM”) is a unit of measure of recombination frequency and genetic distance between two loci. One cM is equal to a 1% chance that a marker at one genetic locus will be separated from a marker at, a second locus due to crossing over in a single generation.

[0040] A “subject” refers an animal, such as a mammal, including a primate (such as a human, a non-human primate, such as a monkey) and a non-primate (such as a mouse). In some aspects of the disclosure, the subject is a human. In some aspects, the subject is a pediatric subject, such as a neonate, an infant, or a child. In other aspects, the subject is an adult subject.

[0041] A “patient” refers to a subject who shows symptoms and / or signs of a disease, is under treatment for disease, has been diagnosed with a disease, and / or is at risk of developing a disease. A “patient” can be a human or veterinary subject. Any reference to subjects in the present disclosure should be understood to include the possibility that the subject is a “patient” unless clearly dictated otherwise by context. More specifically, the subject in certain aspects is a patient who has, suspected to have, or is at risk of having a GLP-1R associated condition (such as type 2 diabetes mellitus, obesity, atherosclerotic cardiovascular disease, non-alcoholic fatty liver disease, polycystic ovary syndrome, or food addiction).

[0042] As used herein, the terms “treating,” “treatment” and the like shall include the management and care of a subject or patient for the purpose of combating a disease, condition, or disorder and includes the administration of a composition to prevent the onset of the symptoms or complications, alleviate the symptoms or complications, reduce at least one associated sign, symptom, or condition, or eliminate the disease, condition, or disorder. Treatment also refers to a prophylactic treatment, such as prevention of a disease (such as type 2 diabetes mellitus, obesity, atherosclerotic cardiovascular disease, non-alcoholic fatty liver disease, polycystic ovary syndrome, or food addiction) or prevention of at least one sign, symptom, or condition associated with the disease (such as type 2 diabetes mellitus, obesity, atherosclerotic cardiovascular disease, non-alcoholic fatty liver disease, polycystic ovary syndrome, or food addiction). Treatment can also mean prolonging survival as compared to expected survival in the absence of treatment.

[0043] As used herein with respect to a parameter, the term “decreased” or “decreasing” or “decrease” or “reduced” or “reducing” or “reduce” or “lower” refers to a detectable (such as at least about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 96%, 97%, 98%, 99%, or 100%) negative change in the parameter from a comparison control, such as an established normal or reference level of the parameter, or anDocket No.107976.00039 established standard control. Accordingly, the terms “decreased,” “reduced,” and the like encompass both a partial reduction and a complete reduction compared to a control.

[0044] As used herein with respect to a parameter, the term “increased” or “increasing” or “increase” or “enhanced” or “enhancing” or “enhance” refers to a detectable (such as at least about 5%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 100%, 200%, 300%, 400%, 500%, 600%, 700%, 800%, 900%, or 1000%) positive change in the parameter from a comparison control, such as an established normal or reference level of the parameter, or an established standard control. A. Glucagon-Like Peptide-1 Receptor (GLP-1R)

[0045] G protein-coupled receptors (GPCRs) form a large group of evolutionarily related proteins that are cell surface receptors that detect molecules outside the cell and activate cellular responses. They are coupled with G proteins. They pass through the cell membrane seven times in the form of six loops (three extracellular loops interacting with ligand molecules, three intracellular loops interacting with G proteins, an N-terminal extracellular region, and a C-terminal intracellular region of amino acid residues, which is why they are sometimes referred to as seven-transmembrane receptors. Ligands can bind either to the extracellular N-terminus and loops (e.g., glutamate receptors) or to the binding site within transmembrane helices (rhodopsin-like family). They are all activated by agonists, although a spontaneous auto-activation of an empty receptor has also been observed.

[0046] GPCRs are an important drug target and approximately 34% of all Food and Drug Administration (FDA) approved drugs target 108 members of this family. The global sales volume for these drugs is estimated to be 180 billion US dollars as of 2018. It is estimated that GPCRs are targets for about 50% of drugs currently on the market, mainly due to their involvement in signaling pathways related to many diseases i.e., mental, metabolic including endocrinological disorders, immunological including viral infections, cardiovascular, inflammatory, senses disorders, and cancer.

[0047] The exact size of the GPCR superfamily is unknown, but at least 831 different human genes (or about 4% of the entire protein-coding genome) have been predicted to code for them from genome sequence analysis. Although numerous classification schemes have been proposed, the superfamily was classically divided into three main classes (A, B, and C) with no detectable shared sequence homology between classes.

[0048] According to the classical A-F system, GPCRs can be grouped into six classes based on sequence homology and functional similarity:Docket No.107976.00039 ^ Class A (Rhodopsin-like) ^ Class B (B1: Secretin receptor family; B2: Adhesion receptor family) ^ Class C (Metabotropic glutamate / pheromone) ^ Class D (Fungal mating pheromone receptors) ^ Class E (Cyclic AMP receptors) ^ Class F (Frizzled / Smoothened)

[0049] Class B1 GPCRs are targeted clinically for a wide range of diseases including diabetes (glucagon-like peptide 1 receptor [GLP-1R] and amylin receptors [AMYRs]), cardiovascular complications (GLP-1R), obesity (GLP-1R and gastric inhibitory polypeptide receptor [GIPR]), migraine headaches (calcitonin gene–related peptide receptor [CGRPR]), hypoglycemia (glucagon receptor [GCGR]), osteoporosis (parathyroid hormone 1 receptor [PTH1R]), and short bowel syndrome (glucagon-like peptide-2 receptor [GLP-2R]). Beyond approved therapeutics, adenylate cyclase activating polypeptide 1 receptor (PAC1R) is implicated in migraine and post-traumatic stress disorder and the vasoactive intestinal peptide receptors 1 and 2 (VPAC1R, VPAC2R) are involved in inflammation. Parathyroid hormone 2 receptor (PTH2R) is implicated in nociception and fear response, and the corticotropin receptors (CRF1R and CRF2R) are implicated in stress responses.

[0050] Cross talk between ligands interacting with GPCR group B1 receptors, such as GLP-1R, can occur through several mechanisms including the following:

[0051] Heterodimerization: GPCRs can form heterodimers with other GPCRs, including those within the same subgroup. Heterodimerization can influence ligand binding, signaling pathways, and cellular responses. For example, the glucagon receptor and the GLP-1 receptor have been shown to form heterodimers, which can modulate their respective signaling pathways.

[0052] Shared signaling pathways: Ligands that interact with different GPCR group B1 receptors may activate common downstream signaling pathways. For instance, both glucagon and GLP-1 can activate adenylate cyclase through G protein signaling, leading to an increase in intracellular cyclic AMP (cAMP) levels. This shared signaling pathway can result in cross talk between these ligands and their respective receptors.

[0053] Cross-regulation: Ligand binding to one receptor within GPCR group B1 can influence the activity or expression of other receptors within the same subgroup. For example, activation of the glucagon receptor by its ligand may affect the expression or function of other receptors involved in glucose metabolism or energy homeostasis.Docket No.107976.00039

[0054] Functional antagonism: Some ligands may exert opposing effects on receptors within GPCR group B1. For example, glucagon stimulates gluconeogenesis and glycogenolysis, whereas GLP-1 promotes glucose-dependent insulin secretion and inhibits glucagon secretion. These opposing actions can result in functional antagonism between ligands and their receptors within the same subgroup.

[0055] Fifteen members of the GPCR class B1 family (GPCRB1) have been identified in humans, with high sequence homology across vertebrate species. As provided herein, Applicant identified polymorphisms in the GPCRB1 receptors that can be used as markers to identify subjects that have poor outcomes with GLP-1R agonists. The variable responses to these ligands may be associated with polymorphisms within the GLP-1 receptor or other GPCRB1 receptors. Cross talk between ligands could impact the response to GLP-1 and this may be associated with polymorphisms in GLP-1R and other GPCRB1 receptors.

[0056] The part of GPCRs that typically binds to a ligand is the extracellular domain. The extracellular domain of a GPCR contains amino acid residues that form a ligand-binding pocket or site. This pocket is often formed by loops and helices within the extracellular domain, and it has a shape and chemical environment that is complementary to the structure of the ligand it binds. This specificity allows GPCRs to selectively recognize and interact with their ligands. A test to detect polymorphisms in GPCRs may target the extracellular domain as this portion of the receptor has a large number of polymorphisms and directly interacts with GLP-1. A test for polymorphisms in GPCRB1 receptors is used to screen a cohort of subjects with various phenotypic responses to GLP- 1R agonist treatment. Once the data is obtained, machine learning algorithms are deployed to develop a genetic profile (GLP-1R biomarkers) to predict patient outcomes. B. GLP-1R Related Markers

[0057] Provide herein are GLP-1R related markers. A “GLP-1 related marker” as used herein refers to any genetic marker related to the GLP-1R pathway or GLP-1R activity, and any genetic marker related to risk of developing a GLP-1R associated condition or treatment outcomes of a GLP-1R associated condition. A GLP-1R related marker can include a genetic marker that is located outside the GLP-1R gene. GLP-1R markers can be used to assess genetic preposition of a subject to GLP-1R associated conditions and / or predictive treatment responses to targeted GLP-1R agonist therapeutics. A “GLP-1R associated condition,” “GLP-1R associated disease,” “GLP-1R associated disorder” and the like interchangeably refer to any disease, disorder, or condition that is directly or indirectly associated with the GLP-1R activity and / or that may benefit from modifying the GLP-1R activity,Docket No.107976.00039 such as by GLP-1R agonist therapy. GLP-1R associated conditions include type 2 diabetes, obesity, atherosclerotic cardiovascular disease, non-alcoholic fatty liver disease, polycystic ovary syndrome, for all of which a GLP-1R therapy may be indicated. GLP-1R associated conditions also include addiction and food-related disorder such as food addiction. The neural response pathways (also referred to as mesolimbic reward pathways) primarily involve the mesolimbic and mesocortical circuits and are integral to processing rewards and reinforcing behaviors. The neural response pathways involve the dopaminergic system, which plays a crucial role in the development and perpetuation of addictive behaviors. Without wishing to be bound by any particular theory, the same neural mechanisms (for example involving the dopaminergic system) are implicated in food-related disorders, such as food addiction, type 2 diabetes, and obesity, where the reward system’s response to food intake can lead to compulsive eating patterns. The neural response pathways can also be important to determine responses to GLP-1 and / or GLP-1R agonists in subjects. GLP-1Rs are expressed in the mesolimbic reward pathways, involving in nucleus accumbers (NAc), lateral spectum (LS), bed nucleus of the stria terminalis (BNST), hippocampal formation (HPF), lateral hypothalamus (LH), ventral tegmental area (VTA), stratum lacunosum-moleculare (SLM), and nucleus tractus solitarius (NTS). GLP-1R agonists may be associated with the reversed reward responses to the drugs of abuse like cocaine, amphetamines, opioids, alcohol and nicotine. The effect of GLP-1R agonists on the neural response pathways (mesolimbic reward pathways) and reward- related behavior is illustrated in FIG. 4. As shown in FIG. 4, stimulation of GLP-1R in the neural response pathways results in changes in reward-related behavior in food addiction, such as decreased palatable food intake, frequency and consumption rate; decreased rewarding food cue responses; and / or decreased active behavior. Stimulation of GLP-1R in the neural response pathways also results in changes in reward-related behavior in cocaine and amphetamine addiction, such as decreased acute and chronic cocaine self-administration, decreased cocaine and amphetamine-induced hyperlocomotion, conditioned place preference (CPP) and cocaine priming-induced reinstatement of cocaine-seeking behavior, and decreased cocaine and amphetamine-induced dopamine release in NAc. Stimulation of GLP-1R in the neural response pathways also results in changes in reward- related behavior in opioid addiction, such as decreased morphine-induced expression and acquisition of CPP, and accelerated extinction and reduced reinstatement of rewarding effects of morphine. Stimulation of GLP-1R in the neural response pathways also results in changes in reward-related behavior in alcohol or nicotine addiction, such as reduced alcohol intake without changing water intake or causing emesis or nausea, decreased nicotine intake, decreased alcohol or nicotine-induced dopamine replace in the NAc, decreased alcohol and nicotine-induced CPP, decreased memoryDocket No.107976.00039 consolidation and retrieval of alcohol reward in CPP and hyperlocomotion, or prevented alcohol deprivation induced drinking and alcohol preference.

[0058] Examples of GLP-1R related markers are set forth in Table 1, including genetic variations in the GLP-1R gene (Table 2 and further described below) and genetic markers associated with random glucose testing glucose levels (Table 3 and further described below). GLP-1R related markers including those set forth in Table 1 (including those set forth in Tables 2 and 3) can be associated with the risk of developing a GLP-1R associated condition in a subject and / or how well the subject responds to GLP-1R agonists.

[0059] The GLP-1R related markers and associations provided herein can be used to enhance personalized treatment strategies, improving the management of GLP-1R associated condition (such as type 2 diabetes) and patient outcomes.

[0060] Table 1. Genetic markers associated with predisposition to food addiction, predisposition to type 2 diabetes, and / or response to GLP-1R agonist therapy. No. Marker Gene Description Posterior name / locus probabilityDocket No.107976.00039 No. Marker Gene Description Posterior name / locus probabilityDocket No.107976.00039 No. Marker Gene Description Posterior name / locus probabilityDocket No.107976.00039 No. Marker Gene Description Posterior name / locus probabilityDocket No.107976.00039 No. Marker Gene Description Posterior name / locus probability[ ] na ys s o genome w e assoc a on s u y ( ) en e oc n - and ARRB1 that were significantly associated with HbA1c reduction from four prospective observational studies: Diabetes Research on Patient Stratification (DIRECT), Predicting Response to Incretin Based Agents in Type 2 Diabetes (PRIBA), Prospective Cohort MRC ABPI Stratification and Extreme Response Mechanism in Diabetes (PROMASTER), and Genetics of Diabetes Audit and Research in Tayside Scotland (GoDARTS09), and two clinical trial datasets from the HARMONY phase 3 trials. These markers are set forth in Table 2, and also are identified as the GLP-1R related markers in Table 1. These identified loci are GLP-1R agonists specific, with no association with effect of other glucose lowering drugs to reduce glucose or HbA1c. A combined genotype (haplotype) derived from GLP- 1R and ARRB1 identified 4% of the population who had a 30% greater reduction in HbA1c relative to the 9% of the population with the worst response to the GLP-1R agonist.

[0062] Table 2. GLP-1R or ARRB1 SNPs associated with HbA1c reduction Marker Gene SNPs (good response → poor response name / locus to GLP-1R agonists)

[0063] A separate GWAS identified 32 genetic markers associated with random glucose testing glucose levels (RG) in non-diabetes population. These markers are set forth in Table 3 below, and also are identified as the GLP-1R related markers in Table 1. These markers enabled investigation of the relationships of type 2 diabetes with other possible GLP-1R associated conditions, and provided pathways for T2D treatment stratification.

[0064] Table 3. Genetic markers associated with random glucose testing glucose levels in the European ancestry meta-analysisDocket No.107976.00039 Marker e / locus G posterior nam ene probability

[0065] In one aspect of the present disclosure, a method for detecting a polymorphism in one or more GLP-1R related markers in a subject is provided. The method includes genotyping a sample obtained from the subject to detect the presence of the polymorphism in the one or more GLP-1R related markers. These one or more GLP-1R related markers are set forth in Table 1, and are one or more of rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115,Docket No.107976.00039 rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467. In certain embodiments, these markers can be located within 10, 5, 4, 3, 2, 1, or 0.5 cM from the genetic positions rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467, respectively, in the human genome. In specific embodiments, the marker rs7997012 is located at position rs7997012 in the 5-HTR2A gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs453 is located at position rs4532 in the DRD1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1799971is located at position rs1799971 in the OPRM1 gene; the marker rs9479757 is located at position rs9479757 in the MUOR gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs10305492 is located at position rs10305492 in the GLP-1R gene; the marker rs10305420 is located at position rs10305420 in the GLP-1R gene; the marker rs2268639 is located at position rs2268639 in the GLP-1R gene; the marker rs2268640 is located at position rs2268640 in the GLP-1R gene; the marker rs57922 is located at position rs57922 in the GLP-1R gene; the marker rs58428187 is located at position rs58428187 in the GLP-1R gene; the marker rs6923761 is located at position rs6923761 in the GLP-1R gene; the marker rs9299870 is located at position rs9299870 in the GLP-1R gene; the marker rs140226575 is located at position rs140226575 in the ARRB1 gene; the marker rs78979036 is located at position rs78979036 in the GLP-1R gene; the marker rs78052828 is located at position rs78052828 in the GLP-1R gene; the marker rs77501730 is located at position rs77501730 in the G6PC2 gene; theDocket No.107976.00039 marker rs77501730 is located at position rs56100844 in the G6PC2 gene; the marker rs76895963 is located at position rs76895963 in the CCND2 gene; the marker rs115545608 is located at position rs115545608 in the G6PC2 gene; the marker rs16856115 is located at position rs16856115 in the G6PC2gene; the marker rs73174306 is located at position rs73174306 in the MECOM gene; the marker rs11708067 is located at position rs11708067 in the ADCY5 gene; the marker rs138917529 is located at position rs138917529 in the GCK gene; the marker rs1514895 is located at position rs1514895 in the SLC2A2 gene; the marker rs2168101 is located at position rs2168101 in the LM01 gene; the marker rs34222465 is located at position rs34222465 in the CACNA2D3 gene; the marker rs348330 is located at position rs348330 in the ABCB10 gene; the marker rs9379084 is located at position rs9379084 in the RREB1 gene; the marker rs10830963 is located at position rs10830963 in the MTNR1B gene; the marker rs183606969 is located at position rs183606969 in the GCK gene; the marker rs12692596 is located at position rs12692596 in the RBMS1 gene; the marker rs8192556 is located at position rs8192556 in the NEUROD1 gene; the marker rs118126621 is located at position rs118126621 in the ARMC2 and / or SESN1 gene; the marker rs78444298 is located at position rs78444298 in the EDEM3 gene; the marker rs35889227 is located at position rs35889227 in the FOXN3 gene; the marker rs11257655 is located at position rs11257655 in the CDC123 and / or CAMK1D gene; the marker rs6538804 is located at position rs6538804 in the RMST gene; the marker rs2255805 is located at position rs2255805 in the TSHZ2 gene; the marker rs17168486 is located at position rs17168486 in the DGKB and / or AGMO genes; the marker rs115128825 is located at position rs115128825 in the G6PC2 gene; the marker rs34814128 is located at position rs34814128 in the OR4C5 and / or OR4A47 genes; the marker rs10501320 is located at position rs10501320 in the MADD gene; the marker rs11592309 is located at position rs11592309 in the ADRA2A gene; the marker rs3842753 is located at position rs3842753 in the INS gene; the marker rs231362 is located at position rs231362 in the KCNQ1 gene; and / or the marker rs3765467 is located at position rs3765467 in the GLP-1R gene, in a human genome.

[0066] Genotyping can be performed in any method known in the art. For example, genomic DNA samples can be genotyped by PCR, real-time PCR, and / or quantitative PCR using a microarray platform or other types of high-throughput format. Sanger sequencing can be used to confirm genotyping accuracy. The potential importance of the polymorphisms at different loci (in different markers) on predicting the risk of developing a GLP-1R associated condition or the likelihood or responding to a GLP-1R agonist therapy can be analyzed using a probabilistic model, with normalization based on the SNP with the highest relative predictive importance.Docket No.107976.00039

[0067] The GLP-1R related markers can include rs6923761, rs140226575, rs10305420, rs2268640, rs78979036, rs58428187, and / or rs78052828 as set forth in Table 2, which can be located within 10, 5, 4, 3, 2, 1, or 0.5 cM from genetic positions rs6923761, rs140226575, rs10305420, rs2268640, rs78979036, rs58428187, and rs78052828 respectively, in the human genome.

[0068] The GLP-1R related markers can include rs77501730, rs56100844, rs76895963, rs10305492, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, and / or rs231362 as set forth in Table 3, which can be located within 10, 5, 4, 3, 2, 1, or 0.5 cM from genetic positions rs77501730, rs56100844, rs76895963, rs10305492, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, and rs231362, respectively, in the human genome.

[0069] The GLP-1R related markers can include rs3765467, rs6923761, rs10305420, rs2268639, rs57922, and / or rs9299870, set forth in Table 1, which can be located within 10, 5, 4, 3, 2, 1, or 0.5 cM from genetic positions rs3765467, rs6923761, rs10305420, rs2268639, rs57922, and / or rs9299870, respectively, in the human genome.

[0070] In some embodiments, the method further includes detecting a status of one or more metabolic markers in the subject. In certain embodiments, the metabolic markers include one or more of age, sex, a lipid species profile, blood glucose level, blood hemoglobin A1c level, blood pressure, and waist circumference. The lipid species profile can include any of the lipid species or lipidome analysis as described for example in Beyene et al.2023 Nature Communications 14:6280, the content of which is incorporated by reference herein in its entirety.

[0071] In some embodiments, the method further includes detecting a status of one or more neural response markers. The “neural response markers” refer to markers of brain reward pathways, and examples of such markers are described in US Patent No. 10,998,105, the content of which is incorporated herein by reference. The example neural response markers set forth in Table 4 below have demonstrated significant predictive capabilities in understanding complex human behaviors and their associated risks and responses to various stimuli. These pathways, primarily involving the mesolimbic and mesocortical circuits, are integral to processing rewards and reinforcing behaviors.Docket No.107976.00039 By analyzing neural activity within these circuits, the Neural Response Panels can accurately forecast subject susceptibility to conditions such as addiction (such as food addiction), where the dopaminergic system plays a crucial role in the development and perpetuation of addictive behaviors. The same neural mechanisms are implicated in food-related disorders, such as type 2 obesity, where the reward system’s response to food intake can lead to compulsive eating patterns.

[0072] The example neural response markers in Table 4 were developed using GWAS and machine learning techniques. GWAS enables the identification of genetic variants associated with specific traits, including those influencing the reward pathways in the brain. Markers rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757 can be located within 10, 5, 4, 3, 2, 1, or 0.5 cM from the genetic positions rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757, respectively, in the human genome. In specific embodiments, the marker rs7997012 is located at position rs7997012 in the 5-HTR2A gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs4532 is located at position rs4532 in the DRD1 gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1799971 is located at position rs1799971 in the OPRM1 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; and the marker rs9479757 is located at position rs9479757 in the MUOR gene, in the human genome.

[0073] By analyzing large datasets, GWAS can pinpoint single nucleotide polymorphisms (SNPs) and other genetic markers that correlate with variations in reward-related behaviors and responses to drugs. This genetic information is crucial for understanding the biological underpinnings of these pathways and their role in various disorders.

[0074] Table 4. Example neural response markers Marker Gene Gene description Impact, Impact, dDocket No.107976.00039 Marker Gene Gene description Impact, Impact, name / locus unnormalized normalizedC. Clinical Decision Pipeline Based on Detection of GLP-1R Related Markers

[0075] The methods provided herein, such as methods of predicting risk of developing GLP- 1R associated condition, predicting response to GLP-1R agonist therapy, providing treatment recommendations, treating, or making clinical decisions can include the following integrated pipeline. The example clinical decision pipeline is depicted in FIG.5.

[0076] Sample Collection & Preparation. As shown in step 502 of the clinical decision-making process 500 of FIG.5, a biological sample (e.g., saliva, buccal swab, or blood) is collected from the subject using standardized clinical-grade kits ensuring preservation of DNA quality. The sample is then subjected to pre-processing protocols including cell lysis, DNA extraction, quantification, and normalization. Sample integrity and concentration thresholds are verified prior to downstream processing.

[0077] Genotyping / Sequencing Process. As shown in step 504 of the clinical decision-making process 500 of FIG.5, the prepared DNA sample undergoes genotyping or sequencing to detect polymorphisms in specific GLP-1R related markers. Methods may include SNP microarray, next-generation sequencing (NGS), or targeted sequencing panels tailored to detect known GLP- 1R and associated loci (see Tables 1–4). High-throughput technologies ensure high coverage and sensitivity. Quality metrics such as call rate, depth, and allele frequency are monitored.Docket No.107976.00039 Confirmatory methods such as Sanger sequencing may be used for variants of uncertain significance.

[0078] Data Processing & Quality Control. As shown in step 506 of the clinical decision- making process 500 of FIG.5, raw sequence or array data is processed through a bioinformatics pipeline including base calling, alignment to a reference genome, and variant calling. Quality control steps assess data integrity, contamination, and sequencing artifacts. Only variants passing defined thresholds for quality metrics (e.g., Phred quality score, genotype call rate, Hardy- Weinberg equilibrium) are included in further analysis.

[0079] Risk Score Generation. As shown in step 508 of the clinical decision-making process 500 of FIG. 5, validated variants are integrated into a predictive model—such as a machine learning-based polygenic risk score model—which assigns weighted contributions to each SNP based on its association strength and posterior probability (as seen in Tables 1–3). The final composite risk score represents the subject’s likelihood of (a) developing type-2 diabetes and / or other GLP-1R associated condition or (b) responding positively or negatively to GLP-1R agonist therapy. Risk thresholds can be calibrated against population cohorts, and scores may be stratified into actionable risk categories (e.g., high, moderate, low). The output can be further integrated into clinical decision support systems.

[0080] Importantly, GLP-1 receptor agonists (GLP-1RAs)—such as liraglutide, semaglutide, and exendin-4—have demonstrated promise beyond glycemic control and weight loss. Preclinical studies indicate that GLP-1RAs modulate dopaminergic activity in mesolimbic brain regions (e.g., the ventral tegmental area, nucleus accumbens, and lateral septum), thereby reducing reward, drug-seeking behavior, and relapse-like responses in animal models. These findings support a mechanistic link between GLP-1R activation and the suppression of reward- driven behaviors, including those associated with opioids, alcohol, nicotine, and food addiction. Although clinical trials are currently limited, the biological rationale and ongoing investigations in addiction medicine provide a compelling basis for stratifying patients based on GLP-1R- related addiction risk markers and guiding use of GLP-1R agonists accordingly.

[0081] Clinical Recommendation and Treatment. As shown in step 510 of the clinical decision- making process 500 of FIG. 5, certain neural biomarkers provided herein, such as single nucleotide polymorphisms (SNPs), can play a critical role in mediating the response to GLP-1R agonist therapy, particularly with respect to behavioral phenotypes such as weight management, food craving, and food addiction. These markers influence signaling in brain regions responsibleDocket No.107976.00039 for reward processing and satiety, including the mesolimbic and mesocortical pathways. Thus, these biomarkers can be used to recommend or select a treatment regimen for a subject. Biomarkers provided herein can also be used to recommend or select as subject as a candidate for a treatment regimen. For example, GLP-1Rs are expressed in neural structures involved in reward regulation such as the ventral tegmental area (VTA), nucleus accumbens (NAc), lateral hypothalamus (LH), and hippocampus. Activation of these receptors has been shown to modulate food intake, reward-seeking behaviors, and responses to addictive behaviors or substances. Therefore, genetic variation in neurotransmitter pathways can affect the efficacy of GLP-1R- targeted interventions.

[0082] Example SNPs and associated phenotypes are as follows:

[0083] rs7997012 (5-HTR2A): Serotonin receptor involved in appetite regulation and mood. The A allele is associated with altered serotonergic tone, impacting craving behaviors. Subjects with this variant may exhibit differential satiety responses to GLP-1R agonists.

[0084] rs1800497 (DRD2): Located in the dopamine D2 receptor gene, a key node in the reward pathway. The T allele is linked with reduced D2 receptor availability and increased vulnerability to addictive behaviors. GLP-1R agonists may attenuate food and drug-related reward signaling more effectively in carriers.

[0085] rs3758653 (DRD4) and rs4532 (DRD1): Dopamine receptor SNPs affecting neural signaling strength. Both are associated with novelty seeking and impulsivity, which influence weight and food cue reactivity. DRD4 VNTR alleles have also been linked to attention and impulsivity traits relevant in eating behavior disorders.

[0086] rs948854 (GAL): This polymorphism in the galanin gene, a neuropeptide linked to fat preference and feeding behavior, is associated with susceptibility to obesity and binge-eating phenotypes. Galanin modulates the hypothalamic control of energy intake, and variants may predict which patients exhibit enhanced weight loss when treated with GLP-1R therapy.

[0087] rs1045642 (ABCB1): Impacts P-glycoprotein expression at the blood-brain barrier. This can affect the central penetration of GLP-1R agonists and, consequently, central appetite control. Individuals with the TT genotype may experience altered central nervous system drug exposure.

[0088] rs2236861 (DOR): Delta opioid receptor SNPs affect hedonic feeding behavior. Variants at this locus may modulate the reinforcing effects of palatable food and opioids, and their interaction with GLP-1R pathways can impact treatment outcomes.Docket No.107976.00039

[0089] rs1051660 (OPRK1): The kappa opioid receptor is implicated in dysphoria and stress- induced craving. This marker may interact with GLP-1R agonist effects on stress eating and reward suppression.

[0090] rs1799971 (OPRM1): A well-established marker in opioid reward and addiction studies. This SNP has been associated with differential outcomes in behavioral therapies and could serve as a predictor of GLP-1R-induced suppression of drug or food-related urges.

[0091] The biomarker polymorphisms (such as SNPs) provided herein can be integrated into a composite risk-response model. For example, machine learning methods can assign weights to each SNP based on known associations with neural, metabolic, and behavioral outcomes. For example, individuals with a high-risk neural SNP profile (e.g., DRD2-TT, OPRM1-GG) may respond more robustly to GLP-1R agonists in terms of craving reduction and weight loss, while others may need adjunctive behavioral interventions.

[0092] For example, a method of predicting a risk for developing GLP-1R associated condition or a response to a GLP-1R agonist therapy in a subject can include detecting a polymorphism in one or more GLP-1R related markers in a sample obtained from the subject; providing a risk score based on the polymorphism in the one or more GLP-1R related markers; and selecting the subject as a candidate for a treatment regimen based on the risk score. The risk score indicates the risk of developing the GLP-1R associated condition or the likelihood of a poor response to a GLP-1R agonist therapy. The GLP-1R related markers can include any one of the GLP-1R related markers set forth in Tables 1-3, such as rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467. As described herein, these markers may be located within 10, 5, 4, 3, 2, 1, or 0.5 cM from the genetic positions rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895,Docket No.107976.00039 rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467, respectively, in the human genome.

[0093] In specific embodiments, the marker rs7997012 is located at position rs7997012 in the 5- HTR2A gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs453 is located at position rs4532 in the DRD1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1799971is located at position rs1799971 in the OPRM1 gene; the marker rs9479757 is located at position rs9479757 in the MUOR gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs10305492 is located at position rs10305492 in the GLP-1R gene; the marker rs10305420 is located at position rs10305420 in the GLP-1R gene; the marker rs2268639 is located at position rs2268639 in the GLP-1R gene; the marker rs2268640 is located at position rs2268640 in the GLP-1R gene; the marker rs57922 is located at position rs57922 in the GLP-1R gene; the marker rs58428187 is located at position rs58428187 in the GLP-1R gene; the marker rs6923761 is located at position rs6923761 in the GLP-1R gene; the marker rs9299870 is located at position rs9299870 in the GLP-1R gene; the marker rs140226575 is located at position rs140226575 in the ARRB1 gene; the marker rs78979036 is located at position rs78979036 in the GLP-1R gene; the marker rs78052828 is located at position rs78052828 in the GLP-1R gene; the marker rs77501730 is located at position rs77501730 in the G6PC2 gene; the marker rs77501730 is located at position rs56100844 in the G6PC2 gene; the marker rs76895963 is located at position rs76895963 in the CCND2 gene; the marker rs115545608 is located at position rs115545608 in the G6PC2 gene; the marker rs16856115 is located at position rs16856115 in the G6PC2gene; the marker rs73174306 is located at position rs73174306 in the MECOM gene; the marker rs11708067 is located at position rs11708067 in the ADCY5 gene; the marker rs138917529 is located at position rs138917529 in the GCK gene; the marker rs1514895 is located at position rs1514895 in the SLC2A2 gene; the marker rs2168101 is located at position rs2168101 in the LM01 gene; the markerDocket No.107976.00039 rs34222465 is located at position rs34222465 in the CACNA2D3 gene; the marker rs348330 is located at position rs348330 in the ABCB10 gene; the marker rs9379084 is located at position rs9379084 in the RREB1 gene; the marker rs10830963 is located at position rs10830963 in the MTNR1B gene; the marker rs183606969 is located at position rs183606969 in the GCK gene; the marker rs12692596 is located at position rs12692596 in the RBMS1 gene; the marker rs8192556 is located at position rs8192556 in the NEUROD1 gene; the marker rs118126621 is located at position rs118126621 in the ARMC2 and / or SESN1 gene; the marker rs78444298 is located at position rs78444298 in the EDEM3 gene; the marker rs35889227 is located at position rs35889227 in the FOXN3 gene; the marker rs11257655 is located at position rs11257655 in the CDC123 and / or CAMK1D gene; the marker rs6538804 is located at position rs6538804 in the RMST gene; the marker rs2255805 is located at position rs2255805 in the TSHZ2 gene; the marker rs17168486 is located at position rs17168486 in the DGKB and / or AGMO genes; the marker rs115128825 is located at position rs115128825 in the G6PC2 gene; the marker rs34814128 is located at position rs34814128 in the OR4C5 and / or OR4A47 genes; the marker rs10501320 is located at position rs10501320 in the MADD gene; the marker rs11592309 is located at position rs11592309 in the ADRA2A gene; the marker rs3842753 is located at position rs3842753 in the INS gene; the marker rs231362 is located at position rs231362 in the KCNQ1 gene; and / or the marker rs3765467 is located at position rs3765467 in the GLP-1R gene, in a human genome.

[0094] The GLP-1R associated condition can be one or more of type 2 diabetes, obesity, atherosclerotic cardiovascular disease, non-alcoholic fatty liver disease, polycystic ovary syndrome, food addiction, and addiction to cocaine, amphetamine, opioids, alcohol, or nicotine. In some embodiments, the risk score is provided using a risk model. In some embodiments, the risk model includes a machine learning model, as further discussed below.

[0095] The method can include providing an individual risk score for each of a plurality of polymorphisms in the one or more GLP-1R related markers, thereby providing individual risk scores; and providing the risk score as a composite of the individual risk scores. The individual risk scores can be weighted based on relevance to the risk of developing the GLP-1 associated condition, or the likelihood of having a poor response to the GLP-1R agonist therapy, to provide the risk score as the composite of the individual risk scores.

[0096] In some embodiments, the risk score being greater than a predetermined threshold can indicate the subject has an increased risk for developing type 2 diabetes or food addiction, or having a poor response to the GLP-1R agonist therapy relative to a reference population. Conversely, the riskDocket No.107976.00039 score being lesser than the predetermined threshold can indicate the subject has a reduced risk for developing type 2 diabetes or food addiction, or having a poor response to the GLP-1R agonist therapy relative to a reference population.

[0097] The method can further include determining a status of one or more metabolic markers and / or a polymorphism in one or more neural response markers in in the subject; and providing the risk score based on the polymorphism in the one or more GLP-1R related markers, the status of the one or more metabolic markers, the polymorphism in the one or more neural response markers, or combination of two or more of the foregoing. The metabolic markers can include one or more of age, sex, a lipid species profile, blood glucose level, blood hemoglobin A1c level, blood pressure, and waist circumference. The status of the metabolic markers can include data from continuous monitoring of blood glucose level in the subject. The status of the metabolic markers can also include data from continuous monitoring of blood hemoglobin A1c level in the subject.

[0098] The neural response markers can include one or more of rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757, located within 10, 5, 4, 3, 2, 1, or 0.5 cM from the genetic positions rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757, respectively, in the human genome. In specific embodiments, the marker rs7997012 is located at position rs7997012 in the 5-HTR2A gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs4532 is located at position rs4532 in the DRD1 gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1799971 is located at position rs1799971 in the OPRM1 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; and the marker rs9479757 is located at position rs9479757 in the MUOR gene, in the human genome.

[0099] By incorporating neural biomarker genotyping into the clinical decision pipeline, it is possible to tailor GLP-1R agonist therapy not just for glycemic control, but also for maximizing impact on neurobehavioral dimensions such as food addiction, compulsive eating, and substanceDocket No.107976.00039 use comorbidity. This stratification enables more personalized treatment and may inform preventive strategies in high-risk populations.

[0100] The methods provided herein, such as methods of predicting risk of developing GLP- 1R associated condition, predicting response to GLP-1R agonist therapy, providing treatment recommendations, treating, or making clinical decisions, can be performed without machine learning. Alternatively, the methods provided herein can be performed with machine learning. Machine learning methods that can be used in the methods provided herein, such as methods of risk prediction, risk-response determination, and treatment provided herein are further described below. D. ML Frameworks and Personalized Stratification Method

[0101] Machine learning (ML) frameworks serve a central role in interpreting diverse patient data and generating individualized recommendations for GLP-1R associated conditions. The following subsections describe components of certain embodiments of end-to-end ML-driven stratification and treatment optimization systems.

[0102] Multimodal Input Integration: The ML models integrate heterogeneous data modalities—genetic, behavioral, and environmental—to produce a holistic patient profile. Genetic data include SNP variants (e.g., GLP-1R, DRD2, OPRM1); behavioral inputs include physical activity patterns, dietary adherence, sleep cycles, and reward-driven behaviors; environmental variables may encompass socioeconomic status, food access, and treatment adherence. This integration enables models to account for gene-environment interactions and epigenetic modulation, yielding higher-resolution predictions of treatment response.

[0103] Personalized Treatment Recommendations: Using supervised and unsupervised learning methods (e.g., gradient boosting, Bayesian optimization, clustering algorithms), the system generates individualized treatment paths. These include: ^ Optimized GLP-1R agonist dosing based on metabolic and neural genotype ^ Combination regimens (e.g., GLP-1R therapy + behavioral CBT for compulsive eating) ^ Behavioral interventions tailored to neural and behavioral responsiveness ^ Adherence-promoting strategies based on historical user engagement profiles

[0104] Neural Biomarker Stratification: SNPs associated with neural reward sensitivity (e.g., DRD2 rs1800497, OPRM1 rs1799971, COMT rs4680) are leveraged to predict reward-driven behaviors and potential for addiction-like responses to food or substances. ML models clusterDocket No.107976.00039 subjects into stratified behavioral phenotypes such as high impulsivity, low satiety response, or reward-seeking dominance. These clusters are then matched to specific behavioral protocols (e.g., dopamine pathway-targeted cognitive-behavioral therapy) or pharmacogenomic dosing strategies.

[0105] Adaptive Learning Models: ML algorithms are designed to adapt to temporal trends through continuous intake of longitudinal patient data such as: ^ Weight trajectories ^ HbA1c or blood glucose via CGM ^ Mood or craving logs ^ Medication adherence over time

[0106] The models update parameters in near-real-time or batch cycles to refine predictions and recommendations. This allows dynamic re-stratification and dose adjustment, effectively enabling a closed-loop precision treatment system.

[0107] Reimbursement-Tier Assignment: To ensure clinical and economic viability, outputs from the ML system include payer-facing risk stratifications. Patients are categorized based on projected cost-effectiveness of treatment regimens, which are mapped to reimbursement tiers. For example: ^ High predicted HbA1c reduction + low cost per QALY improvement = Tier 1 ^ Marginal expected benefit or low adherence = Lower reimbursement priority The model may also estimate potential reduction in healthcare utilization, enabling value- based care models.

[0108] Example machine learning models that can be used in the methods provided herein are described in US Patent No.10,998,105, the content of which is incorporated herein by reference.

[0109] Embodiments include computer implemented methods for predicting risk of developing GLP-1R associated condition or poorly responding to GLP-1R agonist therapy in a patient using a computer system having one or more processors coupled to a memory storing one or more computer readable instructions for execution by the one or more processors. One such method includes the steps of storing a set of data containing a plurality of subject records. The subject records can include clinical data of each of the subject, such as age, sex, race, prior medical history, and ethnicity. In an embodiment, the clinical data includes demographic data, socioeconomic data, and any data that about a subject that can be obtained by observation or oral or written communication. The subject records can also include data about health outcomes andDocket No.107976.00039 healthcare resource utilization, such as quantification or description of the use of services by a subject for the purpose of preventing and curing GLP-1R associated conditions or underlying disease, promoting maintenance of health and well-being, or obtaining information about one's health status and prognosis. For example, an increased risk of a GLP-1R associated condition may include poorer health outcomes and greater utilization of healthcare resources. Each subject record includes a SNP (single nucleotide polymorphism) profile for each individual of a plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition. Each subject record can also include a plurality of physical characteristics for each individual of a plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition. The subject record for each of the plurality of test subjects also includes a first substance use indicator indicating that the test subject has been diagnosed with a GLP-1R associated condition. The subject record for each of the plurality of subjects not diagnosed with a GLP-1R associated condition includes a second substance use indicator indicating the subject has not been diagnosed with a GLP-1R associated condition. The method further includes the step of selecting a first subset of the subject records for the plurality of test subjects and a second subset of the subject records for the plurality of subjects not diagnosed with a GLP-1R associated condition. The first and second subsets do not include the first substance use indicator and the second substance use indicator, respectively. The first and second subsets serve as inputs into one or more initial machine learning models. The method further includes the step of determining whether the one or more initial machine learning models meet a predetermined sensitivity, a predetermined specificity, and a predetermined accuracy for predicting the first substance use indicator for the test subjects and the second substance use indicator of the subjects not diagnosed with a GLP- 1R associated condition. The method further includes the step of generating an ensemble machine learning model responsive to the one or more initial machine learning models that meet the predetermined sensitivity, the predetermined specificity, and the predetermined accuracy. The method further includes the step of supplying a SNP profile of a patient to the ensemble machine learning model as an input to obtain a score indicative of the patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy. The method further includes the step of presenting the score or other indicator indicating the patient’s risk of developing a GLP-1R associated condition to a healthcare professional for decision support to implement a treatment regimen and mitigate the patient’s risk of substance abuse or substance dependence.Docket No.107976.00039

[0110] In another embodiment, a non-transitory machine-readable storage medium may be encoded with instructions executable by a processing resource. The non-transitory machine- readable storage medium includes instructions to store one or more sets of data. Each set of data may include one or more allelic variants. Each of the one or more allelic variants are associated with a subject. Each of the one or more allelic variants includes one or more SNP profiles of the subject and a value indicating whether the subject is diagnosed with a GLP-1R associated condition. The non-transitory machine-readable storage medium can include instructions to generate one or more subsets of data based on the sets of data. Each subset of data may include a plurality of SNP profiles. Each of the plurality of SNP profiles may include a same value indicating whether the subjects of the plurality of SNP profiles are diagnosed with GLP-1R associated condition. The non-transitory machine-readable storage medium can include instructions to train a stacked model with training data generated via creation of the subsets of data. The stacked model may include, at least, a random forest model and a support vector machine model. Each training data may correspond to one of the one or more subsets of data. Each of the one or more subsets of data may correspond to the same value indicating whether the subject is diagnosed with GLP-1R associated condition, such that the training of the stacked model provides a score or other indicator that informs a patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy.

[0111] In another embodiment, a non-transitory machine-readable storage medium may be encoded with instructions executable by a processing resource. The non-transitory machine- readable storage medium can include instructions to retrieve from memory, a trained stacked model. The stacked model may include, at least, a random forest model and a support vector machine model. The stacked model may be trained with one or more inputs and outputs derived from one or more sets of data. The inputs may include one or more SNP profiles. Each of the SNP profiles may be associated with one or more allelic variants of the subject. The outputs may include a value indicating whether the subject has a risk of developing a GLP-1R associated condition. The non-transitory machine-readable storage medium can include instructions to, in response to receipt of an input of a sample, determine if the sample includes the one or more alleles associated with the SNP profiles. The non-transitory machine-readable storage medium can include instructions to, in response to a determination that the sample does not include one of the one or more alleles associated with the SNP profiles, send a response to a user indicating that a prediction is not available. The non-transitory machine-readable storage medium canDocket No.107976.00039 include instructions to, in response to a determination that the sample does include the one or more alleles associated with the SNP profiles, determine, using the retrieved trained stacked model and an average of the results of the random tree model and support vector machine model, a prediction of disposition to GLP-1R associated condition relative to SNP profiles of the one or more alleles. A plurality of inputs to the retrieved trained stacked model may include all possible SNP profiles associated to the plurality of allelic variants.

[0112] Another embodiment of the disclosure is directed to a method of determining a subject’s risk of developing a GLP-1R associated condition. This risk may be expressed as a score or a probability or other indicator that informs a patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy. In certain embodiments, the risk is presented as a binary determination, such as one of Yes or No or one of Low Risk or High Risk. In certain embodiments, the risk is presented as one of small set of possibilities, such as one of low risk, intermediate risk, or high risk. In certain embodiments, the risk is presented as a score ranging from 0 to 1 or from 0 to 100. The method includes the steps of receiving a sample of the subject; analyzing the sample, via a processor of a computing device, to obtain a SNP profile of each of a set of specified allelic variants. In response to the SNP profiles of each of the set of specified allelic variants, the method further includes the step of determining, via the processor and a trained stacked model of the computing device, a score indicating the subject’s risk of developing a GLP-1R associated condition based on the SNP profiles of each of the set of specified allelic variants. The method further includes the step of determining, via the processor, a treatment regimen recommendation for the subject based on the score; and transmitting, via the processor, the score and treatment regimen recommendation to a user device.

[0113] In an embodiment, the score is a value output from the trained stacked model between 0 and 1 and the trained stacked model includes a pre-determined threshold. In an embodiment, the score with a value greater than the threshold indicates the subject has a high risk of developing a GLP-1R associated condition, and a score with a value less than or equal to the threshold indicates the subject has a low risk of developing a GLP-1R associated condition. In an embodiment, the set of specified allelic variants includes two or more of the fifteen allelic variants. The fifteen allelic variants include allelic variants of the following genes: serotonin 2A receptor, galanin, ATP binding cassette transporter 1, catechol-O-methyltransferase, dopamine transporter, dopamine D2 receptor, dopamine D1 receptor, methylene tetrahydrofolate reductase, dopamine beta hydroxylase, delta opioid receptor, a first mu opioid receptor (OPRM1),Docket No.107976.00039 dopamine D4 receptor, gamma-aminobutyric acid, kappa opioid receptor, and a second mu opioid receptor (MUOR). In an embodiment, the set of specified allelic variants includes all fifteen allelic variants.

[0114] The method may include transmitting, via one or more processors of a computing device, a request for a sample of the subject. The method may include determining, via the one or more processors, a treatment regimen recommendation for the subject based on the score and transmitting, via the one or more processors, the score and treatment regimen recommendation to a user device and to the subject’s electronic medical records (EMR). In an embodiment, the score is a value output from the trained stacked model between 0 and 1 and the score with a value greater than the pre-determined threshold indicates the subject has a high risk of developing a GLP-1R associated condition, and a score with a value less than or equal to the pre-determined threshold indicates the subject has a low risk of developing a GLP-1R associated condition.

[0115] Another embodiment of the disclosure is directed to a system to determine a subject’s risk of developing a GLP-1R associated condition. The system can include an analyzer with an input for a sample. In an embodiment, the analyzer conducts an allelic analysis of the sample received at the input to generate the SNP profiles for the trained ensemble model. The analysis, based on the received sample, may provide SNP profiles of each of a set of specified allelic variants. The system can include a trained ensemble model, such as a trained stacked model. The trained ensemble model may include one or more inputs and an output. The inputs may accept one or more SNP profiles from the analyzer. Based on the SNP profiles, the trained ensemble model may provide, at the output, a score or other indicator that informs a patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy. The system may include a user interface input / output in communication with a processor and machine-readable storage medium. The machine-readable storage medium may store instructions and the instructions may be executable by the processor. The executed instruction may, in response to a risk output from the trained ensemble model, determine an indicator to indicate a subject’s risk of developing GLP-1R associated condition, the indicator based on the risk or probability. Further, the instructions may include transmission of the indicator to a user interface associated with a user device.

[0116] The embodiments described herein may utilize a trained ensemble or trained stacked model to determine a predicted predisposition to a GLP-1R associated condition of a subject.Docket No.107976.00039 This approach may allow for prevention or modification of prescription of specific medications to individuals predisposed to a GLP-1R associated condition, thus preventing potential negative implications through prescription. For example, upon receipt of a sample, a system may provide a prediction of the likelihood of a patient developing a GLP-1R associated condition within, depending on various factors, hours or days. Certain embodiments include methods and systems for determining health outcomes and / or healthcare resource utilization based on evaluating the risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy using machine learning models. For example, an increased risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy may include poorer health outcomes and greater utilization of healthcare resources. Certain embodiments include methods and systems for determining health or wellness of the subject based on the analysis of the allelic variants and / or clinical data of the subject. Certain embodiments include methods and systems for determining health outcomes of the subject based on the analysis of the allelic variants and / or clinical data of the subject. Certain embodiments include methods and systems for determining healthcare resource utilization by a subject based on the analysis of the allelic variants and / or clinical data of the subject.

[0117] In an embodiment, a system 201 (as illustrated in FIGS.2B, and 2C) for predicting risk of developing GLP-1R associated condition or poorly responding to GLP-1R agonist therapy and / or for providing decision support to healthcare professionals to implement a treatment regimen and mitigate substance use includes a user device 220, an electronic communications network, a SNP analyzer 222, a patient record database 224, a processor 204 to execute instructions stored on a non-transitory machine readable storage medium 206, the instructions to, when executed by the processor 204, may retrieve and / or utilize a trained ensemble and / or trained stacked model 100 stored in a memory 214 (e.g., the computing device 202 may receive or retrieve the trained stacked model 100 and store the trained stacked model 100 in the machine readable storage medium 206). In an embodiment, the system 201 for predicting risk of developing GLP-1R associated condition or poorly responding to GLP-1R agonist therapy can include decision support modules (for example, included as machine readable instructions stored in machine readable storage medium 206) coupled to a patient record database 224 and / or one or more of a prescription drug database, the patient’s electronic health records, and drug utilization records. Decision support modules may include providing evidence-based information to the healthcare professional regarding drug selection and dosing, potential adverseDocket No.107976.00039 drug reactions and drug allergies, duration or form of prescribed drugs, or stratification of treatments. The computing device 202 may include an input / output to connect to a user device 220 or user interface. The user device 220 or user interface may include a display 216. The display 216 may be configured to display a graphical user interface (GUI) for receiving a SNP profile and one or more electronic medical records (EMR) or electronic health records (EHR) associated with a patient and displaying a risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy. The GUI may display one or more recommendations for treatment regimens for the patient. The electronic communications network is in communication with the user device 220, the SNP analyzer 222, the patient record database 224, and the processor 204 that executes the trained ensemble and / or trained stacked model 100. In another example as shown in FIG. 2C, the computing device 202 may be or include a SNP analyzer 222 or another genetic information analyzer. Further, rather than connect to the trained stacked model 100, the computing device 202 can include the trained stacked model 100 (e.g., the trained stacked model 100 may be stored in the machine-readable storage medium 206).

[0118] In an embodiment, a system 200 (as illustrated in FIG. 2A) may include a computing device 203 including a machine-readable storage medium 207 and a processor 205. The machine readable storage medium 207 includes instructions, when executed by the processor 205, to receive a plurality of SNP profiles and / or GLP-1R associated condition profiles associated to the plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition from the user device 221 or patient record database 224, train one or more predictive machine learning models (for example, stacked model 210 or a first random forest model 105, second random forest model 107, and a SVM model 109) using the plurality of physical characteristics and associated SNP profiles, thereby generating an ensemble predictive model (for example, a trained stacked model 100) for determining risk of GLP-1R associated condition. The machine- readable storage medium 207 may include instructions, when executed by the processor 205, to receive a plurality of physical and / or clinical data from a plurality of EMR associated to a plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition from a user device 221 or patient record database 224. The machine-readable storage medium 207 may also include instructions to retrieve one or more treatment regimen recommendations associated with a particular risk of GLP-1R associated condition. The data (as in, the training data 102) used to train the one or more predictive machine learning models (e.g., a stacked model 210)Docket No.107976.00039 may include a GLP-1R associated condition profile. The GLP-1R associated condition profile may include one or more of a plurality of physical or clinical data from a plurality of EMR associated to a plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition, a plurality of SNP profiles associated to the plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition, a plurality of medical diagnoses regarding GLP- 1R associated conditions associated to the plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition, and / or a plurality of treatment regimen recommendations.

[0119] The machine-readable storage medium 207 may include instructions to assign a plurality of weight values to the training data 102 within a plurality of GLP-1R associated condition profiles, such as to a plurality of associated SNP profiles of the training data, determine the one or more weighted data inputs of the training data, thereby to form an initial set of decision- making rules. The machine-readable storage medium 207 may include instructions to compare the initial set of decision-making rules to the one or more weighted data inputs of the training data and re-weight the initial set of decision-making rules based on the comparison to create a re-weighted set of decision-making rules. The re-weighted set of decision-making rules and the one or more weighted data inputs of the training data may be used to update the initial predictive model to generate an ensemble predictive model for determining risk of GLP-1R associated condition.

[0120] For example, the training data 102 may be comprised of a set of data (as illustrated in FIG.1B) including data associated with a group of individuals or subjects known to have a GLP- 1R associated condition and a group of individuals or subjects with no known GLP-1R associated conditions. Various models may be utilized to create a classification, prediction, or probability of developing GLP-1R associated condition, based on SNP profiles and / or physical and / or clinical data. Models and methods may include decision trees, random forest models, random forests utilizing bagging or boosting (as in, gradient boosting), neural network methods, support vector machines (SVM), other supervised learning models, other semi-supervised learning models, and / or other unsupervised learning models, as will be readily understood by one having ordinary skill in the art. Further, a stacked model 210 or an ensemble machine learning model may be utilized. A stacked model 210 may include two or more learning models, for example, a first random forest model 105, a second random forest model 107, a SVM model 109, and / or other models as will be readily understood by one having ordinary skill in the art.Docket No.107976.00039

[0121] As illustrated in FIGS. 1A and 1C through 1E, the trained stacked models 100 may include a first random forest classifier 106 based on results from a first random forest model 105, a second random classifier 108 based on a second random forest model 107, and a SVM classifier 110 based on a SVM model 109. The first random forest model 105 and second random forest model 107 may be trained utilizing a re-sampling or bootstrapping technique. In other words, samples from the training data 102 may be re-used or left out for new decision trees or sets of decision trees. As illustrated in FIG 1B, the training data 102 may include an 80% training 116 and 20% testing 118 split (other split percentages, such as 70 / 30 may be utilized, as will be readily understood by one having ordinary skill in the art). In other words, the random forest may be trained with the 80% 116 of the training data 102, while the 20% 118 of the training data 102 may be held out for testing the newly trained model. In another example, the stacked model 210 may include a gradient boosted tree model. The gradient boosted tree model may include similar training methods. The gradient boosted tree model may include logic that measures error based on small changes to predictions. In other words, the target outcome for each case may depend on how changing a case’s prediction impacts prediction error. Other models may be utilized, for example, linear regression models, logistic regression models, naïve Bayes models, kNN models, k-Means models, dimensionality reduction models, other gradient boosting models, and / or neural networks, as will be readily understood by one having ordinary skill in the art. In another example, ensemble model training may utilize n fold cross-validation. In other words, the set of data may be split n times (for example, 5-fold or 20-fold cross-validation). While each split trains a model, another split may be leftover. The trained model may test against the leftover split. The model may be trained again, but a different split may be held out for testing. This may occur until all the training data 102 is utilized for training and testing. Such a training and testing method may prevent or lessen the likelihood of over-fitting. Over-fitting may occur when a model exhibits random error or noise instead of an underlying relationship. As noted, cross-validation may prevent or lessen the likelihood of over-fitting. Other methods to avoid over-fitting may be utilized, as will be readily understood by one having ordinary skill in the art.

[0122] Training a model, may result in a classifier, predictor, score, or probability. A sample 101 (for example, a SNP profile obtained from a biological material, such as from a buccal swab or a blood sample (or other method of obtaining DNA or RNA), other allelic variations, RNA expression data, and / or physical or clinical data may be entered via an input device 218 and / orDocket No.107976.00039 a SNP analyzer 222 into a computing device) may be applied to the predictor to produce a probability or prediction score 114 between 0 and 1 or some other arbitrary numerical range. Multiple models or ensemble models may be stacked to further prevent over-fitting or over confident predictions. The average 112 of multiple models may, when those models disagree, tend to average out towards a mean of the original dataset.

[0123] As noted, the machine readable storage medium 206 may include instructions to determine risk or probability of development of GLP-1R associated condition for a particular patient using the trained ensemble and / or trained stacked model 100 responsive to the plurality of physical or clinical data and / or the associated SNP profile of the particular patient, identify the one or more associated recommendations responsive to the risk of GLP-1R associated condition, and transmit the risk of GLP-1R associated condition and the associated treatment regimen recommendations to the user device 220 via the electronic communication network to be displayed on the GUI. This risk may be expressed as a score or a probability or other indicator that informs a patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy. In certain embodiments, the risk is presented as a binary determination, such as one of Yes or No or one of Low Risk or High Risk. In certain embodiments, the risk is presented as one of small set of possibilities, such as one of low risk, intermediate risk, or high risk. In certain embodiments, the risk is presented as a score ranging from 0 to 1 or from 0 to 100. The user device 220 may be configured to display a GUI for displaying the risk of GLP-1R associated condition and the one or more associated treatment regimen recommendations. For example, the user device 220 may include a mobile device (e.g., smart phone, tablet, etc.), a desktop computer, a laptop, a wearable computing device, or other type of computing device, as will be readily understood by one having ordinary skill in the art. The user device 220 may further include a mobile computer application.

[0124] As noted, the output of each of the models may be a value from 0 to 1 or some other arbitrary number. As noted, an average of the outputs may be taken, as a correction for potential disagreements in the models. In a further example, an allocated threshold may indicate predisposition to develop GLP-1R associated condition, such as an output greater than 0.33, while a value less than or equal to 0.33 may indicate no or lower likelihood or predisposition to develop GLP-1R associated condition. Further, the allocated threshold may be set during the training of the stacked model 210. For example, the machine-readable storage medium 207 may include instructions to automatically set the allocated threshold or include a predetermined valueDocket No.107976.00039 for a threshold. In another example, various indicators may be associated with various thresholds. For example, an indicator may include, “YES”, “NO”, “LOW”, “HIGH”, “INTERMEDIATE”, some other indicator to indicate level of risk associated with GLP-1R associated condition, and / or some combination thereof. Further, “YES” may indicate that a patient is predisposed to develop GLP-1R associated condition, while “NO” may indicate that the patient is not predisposed to develop GLP-1R associated condition. “LOW” may indicate that a patient has a low risk of GLP-1R associated condition, “INTERMEDIATE” may indicate that a patient has a level of risk somewhere in between “HIGH” and “LOW” (but not past a threshold to indicate a high risk or low risk), and “HIGH” may indicate that a patient has a high level of risk of GLP-1R associated condition. In another example, the indicator may be presented, via the user device 220, with or without the percentage of risk (in other words, the output of the average of the trained stacked model 100).

[0125] In an embodiment, the machine-readable storage medium further can include instructions the processor to: in response to the score that informs a patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy being greater than the pre-determined threshold, retrieve the non-GLP-1R agonist based treatment regimen recommendation from the database; and transmit to the user interface an output indicating the predisposition of the patient to develop the GLP-1R associated condition and the non-GLP-1R agonist based treatment regimen recommendation. In an embodiment, the machine-readable storage medium further can include instructions to: in response to the score that informs a patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy being less than or equal to the pre-determined threshold, retrieve the GLP-1R agonist based treatment regimen recommendation from the database; and transmit to the user interface an output indicating the predisposition of the patient to develop the GLP-1R associated condition and the GLP-1R agonist based treatment regimen recommendation. In an embodiment, a treatment regimen recommendation can include a reduced dose of the GLP-1R agonist. In an embodiment, a treatment regimen recommendation can include a reduced dose of the GLP-1R agonist in combination with the non-GLP-1R agonist or other interventions. In an embodiment, a treatment regimen recommendation can include a limited time period for use of the GLP-1R agonist followed by extended use of the non-GLP-1R agonist or other interventions. For example, a treatment regimen recommendation can include use of a GLP-1R agonist for no more than six months or a year, then use of a non- GLP-1R agonist.Docket No.107976.00039

[0126] The present disclosure also discloses embodiments directed to a method for predicting a patient’s risk of developing a GLP-1R associated condition or for providing decision support to healthcare professionals to implement a treatment regimen on a graphical user interface (GUI) of a computing system using an ensemble or stacked machine learning model. In an example, training data 102 or the sample 101 may be pre-processed by a pre-processing module 104 or pre-processing instructions stored on the computing device 203 and / or computing device 202, respectively, or at the memory 215 where the stacked model 210 is stored or the memory 214 where the trained stacked model 100 is stored. In one such method, a plurality of physical or clinical data and associated SNP profile of a particular patient may be processed to a format for processing, such as normalization, one-hot encoding, and / or ordinal encoding, by one or more machine learning models, and storing the plurality of physical data and associated SNP profile of the particular patient to a data structure in a patient records database 224. The method may further include training one or more initial predictive models for predicting a patient’s risk of developing a GLP-1R associated condition using the plurality of physical or clinical data and / or associated SNP profile of a plurality of subjects (with and without diagnoses of GLP-1R associated condition) constituting the training data 102, thereby generating an updated predictive model for predicting risk of developing GLP-1R associated condition or poorly responding to GLP-1R agonist therapy or for providing decision support to healthcare professionals to implement a treatment regimen. The method may additionally include assigning a plurality of weight values to the one or more data inputs of the training data 102, determining the one or more weighted data inputs of the training data 102, thereby to form an initial set of decision- making rules, and comparing the initial set of decision-making rules to the one or more weighted data inputs of the training data 102. In an embodiment, the clinical data includes age, sex, race, and ethnicity. In an embodiment, the clinical data includes demographic data, socioeconomic data, and any data that about a subject that can be obtained by observation or oral or written communication.

[0127] Further, the method may include, for example, re-weighting the initial set of decision- making rules based on the comparison to create a re-weighted set of decision-making rules. The re-weighted set of decision-making rules and the one or more weighted data inputs of the training data may be used to update the initial predictive model to generate the updated predictive model for predicting risk of developing GLP-1R associated condition or poorly responding to GLP-1RDocket No.107976.00039 agonist therapy or for providing decision support to healthcare professionals to implement a treatment regimen.

[0128] FIGS. 3A through 3E illustrate flow diagrams, implemented in a computing device, to predict a subject’s or patient’s predisposition to GLP-1R associated condition (in other words, a patient’s risk of developing a GLP-1R associated condition) or the subject’s or patient’s response to GLP-1R agonist therapy, according to an embodiment. The method is detailed with reference to the computing device 202. The actions of methods 300, 322, and 342, may be completed by the computing device 202. Specifically, methods 300, 322, and 342 may be included in one or more programs, protocols, or instructions loaded into the machine-readable storage medium 206 of the computing device 202 and executed on the processor 204 or one or more processors of the computing device 202. The order in which the operations are described is not intended to be construed as a limitation, and any number of the described steps may be combined in any order and / or in parallel to implement the methods.

[0129] FIG. 3A illustrates a flowchart of a method 300 to utilize a stacked machine learning model to predict a subject’s or patient’s predisposition to GLP-1R associated condition (in other words, a patient’s risk of developing a GLP-1R associated condition), or the subject’s or patient’s response to GLP-1R agonist therapy, according to an embodiment. The method 300, at step 302, may include receiving a sample for allelic analysis (e.g., such as from a buccal swab, a blood sample and / or any other sample containing genetic material). The sample is analyzed for one or more SNP profiles associated to a specified set of allelic variants of the subject or patient. In an example, the sample is analyzed for one or more allelic variants (for example, the allelic variants shown in Table 2 below). Information from a biological material or a sample 101 may be entered into a computing device 202 via an input device 218. The input device 218 may be a keyboard, touchscreen, and / or mouse. In another example, a SNP analyzer 222 may send the SNP profiles to the computing device 202 (for example, from the analysis of a buccal swab). In yet another example, the computing device 202 may include the SNP analyzer 222 or other analyzer to obtain genetic information from a biological material.

[0130] In such examples, in response to the receipt of a sample, at step 303, the computing device 202, via the included SNP analyzer 222, may analyze the sample to obtain the one or more SNP profiles of the allelic variants (or other genetic information) from the sample. In an embodiment, the SNP analyzer can include any automated DNA or RNA sequencing system coupled to a software module to call SNPs in genome or transcriptome samples. In anDocket No.107976.00039 embodiment, data from a next generation sequencing system or high-density SNP arrays are aggregated and aligned to call the SNPS of the set of specified allelic variants.

[0131] The method 300, at step 304, may include, in response to receipt of the sample, determining whether the computing device 202 is communicatively connected to the trained ensemble or trained stacked model 100. If the computing device 202 is not connected to the trained ensemble or trained stacked model 100, the computing device 202 may, at step 306, send a prompt to a healthcare professional that an initial risk, prediction, or score of GLP-1R associated condition or response to GLP-1R agonist therapy is not available. After sending the prompt, the computing device 202 may, at step 308, attempt to re-establish connection with the trained ensemble or trained stacked model 100. In another example, the computing device 202 may not send a prompt, but rather attempt to re-establish connection with the trained ensemble or trained stacked model 100, which may take a short period of time (e.g., seconds or minutes). After attempting to re-establish the connection, the method, at step 310, may restart the process with the same sample (for example, by resubmitting the one or more allelic variants for analysis to the computing device 202). In another example and as illustrated in FIG. 2C, the computing device 202, rather than connecting to the trained stacked model 100, may include the trained stacked model (e.g., the trained stacked model 100 may be stored in either the machine-readable storage medium 206 or another memory of the computing device 202). Thus, the computing device 202 may not check for a communicative connection.

[0132] The method 300, at step 312, may include determining, via the computing device 202, whether the sample includes the one or more allelic variants used to train the trained ensemble or trained stacked model 100 (e.g., specified allelic variants). In response to a determination, at step 312, that not all the one or more allelic variants are present (for example, due to an indeterminate allele or no call), the method 300, at step 314, may include prompting, via the computing device 202 to the user device 220 or display 216, a healthcare professional that an initial risk, prediction, or score of a likelihood to develop a GLP-1R associated condition or respond to GLP-1R agonist therapy is not available, at which point, the healthcare professional may decide to input another sample for analysis.

[0133] In response to a determination that the sample includes the one or more allelic variants, the method 300, at step 316, may further determine, via the trained stacked model 100, a score or other indicator that informs a patient’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy. In other words, the SNP profiles of the allelicDocket No.107976.00039 variants may be input into the trained stacked model 201 or applied to the model classifiers, resulting in an output (e.g., 0 to 1 or some other range of numbers). Further, the score or prediction may be compared to a threshold. As noted, the threshold may be pre-determined or input into the computing device 202. If the score or prediction is less than the threshold, at step 318, a score, prediction, and / or indicator indicating no or low potential or likelihood for development of a GLP-1R associated condition or indicating good response to GLP-1R agonist therapy by a patient may be returned or transmitted to the user device 220. Additionally, a recommendation for a treatment regimen may be sent to a healthcare professional, based on the determination that there is no or low risk for GLP-1R associated condition or there is good response to GLP-1R agonist therapy.

[0134] In response to a determination that the score indicator is greater than a threshold, at step 320, a score, prediction, and / or indicator indicating a potential or likelihood for development of a GLP-1R associated condition or response to GLP-1R agonist therapy by a patient may be returned or transmitted to the user device 220. Additionally, the method 300 may include identifying one or more associated recommendations related to the determined score, prediction, and / or indicator, and automatically transmitting (along with the score or prediction) the determined initial risk and the one or more associated treatment regimen recommendations (e.g., a recommendation to utilize an alternate drug or prescription, such as a non-GLP-1R agonist, or a reduced dose GLP-1R agonist drug-based treatment regimen) to the user device 220 or display 216 of a healthcare professional via the electronic communication network in real time or in a timeframe of hours to days. Thus, the healthcare professional may have relevant information on treatment regimens to discuss with the patient, and the treatment regimens would be based on a potential for or risk of developing a GLP-1R associated condition and / or likelihood of good response to GLP-1R agonist therapy. The method 300 may further include updating the EMR of the patient with further clinical data or patient history and updating the prediction of the initial risk of GLP-1R associated condition or response to GLP-1R agonist therapy to reflect the additional clinical data or patient history (e.g., updating the EMR at the patient record database 224).

[0135] FIGS.3B-3E illustrate a flowchart of a method 322 to utilize a stacked machine learning model to predict a subject’s or patient’s predisposition to GLP-1R associated condition (in other words, a patient’s risk of developing a GLP-1R associated condition), or response to GLP-1R agonist therapy, according to another embodiment. In such embodiments, at step 324, aDocket No.107976.00039 healthcare professional may initiate the method 322 prior to an operation (e.g., pre-operation) to determine a pain management regimen for the patient post-surgery. In another example, method 322 may be initiated by a healthcare professional at an appointment or while performing an annual wellness check (e.g., a wellness check for employment or work-related insurance or pharmaceutical insurance purposes or at the request of a subject / patient’s insurer). Further, the initiation may be in response to a determination that the patient requires a prescription of a particular drug or that the patient or a healthcare professional has requested the determination of the patient’s risk of developing a GLP-1R associated condition or likelihood of response to GLP- 1R agonist therapy. In an embodiment, a healthcare professional may initiate the method 322 when a patient requests an elective procedure.

[0136] In response to an initiation of method 322, the computing device 202 may determine whether any prior scores for a particular patient have been determined. In other words, the computing device 202 may determine whether a patient’s genetic data has been utilized in the process to determine the risk of developing a GLP-1R associated condition or likelihood of response to GLP-1R agonist therapy. In such examples, the computing device 202 include an input / output connected to a patient record database 224. The computing device 202 may determine whether the prior scores are stored in the patient record database 224 (e.g., in a patients EMR), and / or in another data storage location.

[0137] If a patient does not have prior scores available, the computing device 202, at step 328, requests a biological sample, genetic data, and / or physical or clinical data of the patient for analysis or data input at the user device 220. The request may be transmitted to a healthcare professional. The prompt may be displayed via the display 216 of a user device 220. In another example, the sample (e.g., genetic information from a DNA or RNA from a buccal swab) may be input into a SNP analyzer 222 (for example, via a biofilm chip) and the data output from the SNP analyzer 222 may be transmitted to the computing device 202. In yet another example, the computing device 202 may include the functionality of the SNP analyzer 222. In another example, the computing device 202 may prompt or initiate shipment of a home swab kit to the patient. The computing device 202, at step 330, analyzes the sample to determine the SNP profiles of the set of specified allelic variants. The computing device 202 may analyze the biological sample or other genetic material or pre-process the physical or clinical data. In response to the analysis of the SNP profiles of the set of specified allelic variants or receipt of analytical data from the swab or other genetic material and / or in response to the pre-processingDocket No.107976.00039 of the physical or clinical data, the computing device 202 may initiate the determination of the score indicating the risk of developing use disorder, as described in FIGS. 3A and 3B (e.g., at steps 312 through 320 or steps 304 through 320).

[0138] In response to a determination that the score indicating the risk of developing GLP-1R associated condition or having poor response to GLP-1R agonist therapy is less than the threshold, at step 332, the computing device 202 may generate a report. The report may include the allelic variants of the patient, the score determined from the trained ensemble or trained stacked model 201, the level of risk associated with developing GLP-1R associated condition or having poor response to GLP-1R agonist therapy based on the score, a treatment regimen recommendation, and / or physical or clinical data. In another example, in response to a determination that the score indicating the risk of developing use disorder is more than the threshold, at step 334, the computing device 202 may generate a report as well, with similar information as described above for step 332.

[0139] At step 338, the computing device 202 may transmit a treatment regimen recommendation based on the report or based on the score indicative of the level of risk associated with developing GLP-1R associated condition or response to GLP-1R agonist therapy. In an embodiment, the report may include the original prescription and the score or indicator for a patient’s likelihood or predisposition to develop a GLP-1R associated condition or responding to GLP-1R agonist therapy. Based on the returned score indicating a high risk of developing a GLP-1R associated condition or poor response to GLP-1R agonist therapy, the computing device 202 may provide suggestions or recommendations for a treatment regimen or possible alterations to the prescription to a drug that may reduce the patient’s likelihood to develop the GLP-1R associated condition or enhance the patient’s likelihood to respond to the therapy provided.

[0140] The computing device 202 may transmit the report and / or treatment regimen recommendation to patient data storage or a patient’s EMR (e.g., patient record database 224). In another example, a field associated with likelihood to develop substance abuse disorder may be altered or updated, based on the results described above. For example, the computing device 202 may include or add a note or indication of the high risk of developing a GLP-1R associated condition or likelihood of response to GLP-1R agonist therapy in the patient’s EMR (e.g., stored medical records at a patient record database 224 or other data storage locations).

[0141] In FIGS. 3D through 3E, the method 342 may be initiated, at step 344, in addition to what is described above for method 322, by a pharmacy benefit manager (PBM) or by an insurerDocket No.107976.00039 (e.g., prescription drug insurance provider). In such examples, prior to funding or paying for a portion of or all of a certain prescription drug, a PBM or insurer may request that a patient provide a report or score indicating the risk of the patient developing a GLP-1R associated condition or response to GLP-1R agonist therapy (as described above). As such a determination may take time, in some examples, the computing device 202, PBM, or insurer may provide a 3- day maximum prescription of the drug prescribed, at step 346. At step 330, in response to receipt of the sample from the patient, the computing device 202 may perform an analysis to obtain one or more SNP profiles of the allelic variants or other patient related information and then initiate the determination of the score indicating the risk of developing use disorder, as described in FIG. 3A (e.g., at steps 312 through 318 or steps 302 through 318). Further, once the score is determined, the computing device 202 may generate reports, recommend possible alterations to a prescription, generate a treatment regimen recommendation, and / or transmit or store the report and prescription data to the patient data storage (e.g., patient record database 224) as described above for FIGS.3B through 3C (e.g., at steps 332 through 340).

[0142] As used herein, an electronic communications network may be any type of network configured to provide communications between components of system 200—the user device 220 or user interface, the SNP analyzer 222, the patient record database 224, and the processor 204 that executes the trained ensemble or trained stacked model 100. For example, the electronic communications network may include a network infrastructure that facilitates the exchange of information and provides communication, such as the Internet, a Local Area network (LAN), wired network, cable, Wireless Local Area Network (WLAN), cellular, satellite, or other suitable connections that enable the exchange of information between the components of systems, as will be readily understood by one having ordinary skill in the art. In some other embodiments, however, the components of the system 200 may be connected through a dedicated direct communication link.

[0143] As used herein, a “machine readable storage medium” may be any electronic, magnetic, optical, or other physical storage apparatus to contain or store information such as executable instructions, data, and the like. For example, any machine-readable storage medium described herein may be any of random access memory (RAM), volatile memory, non-volatile memory, flash memory, a storage drive (e.g., hard drive), a solid-state drive, any type of storage disc, and the like, or a combination thereof. As noted, the machine-readable storage medium 206 may store or include instructions executable by the processor 204.The processor 204 or processingDocket No.107976.00039 resource that executes the trained ensemble or trained stacked model 100 can include a server, storage services, cloud provided by an entity such Microsoft Azure™ or Amazon Web Services® cloud, among others, as will be readily understood by one having ordinary skill in the art. The processor or processing resource may also include a plurality of computing devices in communication with the electronic communications network. For example, in some embodiments, the processor may be a plurality of processors connected together in communication with the electronic communications network. In other embodiments, the processor may be a group of graphical processing units configured to work in parallel as a GPU cluster. A processor may include a single processor device and / or a plurality of processor devices (e.g., distributed processors).

[0144] A processor 204 may be any suitable processor capable of executing / performing instructions. A processor 204 may include a central processing unit (CPU) that carries out program instructions to perform the basic arithmetical, logical, and input / output operations required to execute the method of predicting risk of developing GLP-1R associated condition or poorly responding to GLP-1R agonist therapy or for providing decision support to healthcare professionals to implement a treatment regimen. A processor 204 may include code (e.g., processor firmware, a protocol stack, a database management system, an operating system, or a combination thereof) that creates an execution environment for program instructions. Processes and logic flows described herein may be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating corresponding output. The processor is communicatively connected to one or more patient record databases through the electronic communication network. Patient record databases 224 may include one or more memory devices that store information and are accessed and managed through or by the processor 204. For example, databases may include Oracle ™ database or other relational databases or non-relational databases such as Hadoop. Databases can include computing components such as database server configured to receive and process requests for data stored in memory devices of databases and to provide data from the databases. The patient record databases 224 may be located remotely, such as in a different geographical location or the cloud. The patient record databases 224 may include EMR of the patients, training data, and / or initial and updated predictive models. The patient record databases 224 may include other clinical data such as age, sex, race, and ethnicity. In an embodiment, the clinical dataDocket No.107976.00039 includes demographic data, socioeconomic data, and any data that about a subject that can be obtained by observation or oral or written communication. E. Treatment of a GLP-1R Associated Condition

[0145] Also disclosed is a method for treating (including providing preventive measures to) a subject having, suspected to have, or at risk of having a GLP-1R associated condition. The method can further include administering the treatment regimen (which the subject was selected as a candidate for receiving) to the subject. Once the risk for developing GLP-1R associated condition or a response to a GLP-1R agonist therapy in a subject, one of ordinary skill in the art (such as a clinician) can readily select a procedure or a treatment regimen that is compatible with the detection results. The treatment regimen can include a GLP-1R agonist therapy, a diet therapy, an exercise therapy, insulin therapy, oral diabetic medication, cognitive therapy, behavioral therapy, or any therapy suitable for treating the GLP-1R associated condition. One of ordinary skill in the art is aware of appropriate treatments for conditions discussed in relation the methods of detection, prediction, and assessment described herein.

[0146] To facilitate the delivery of optimal and personalized treatment regimens for GLP-1R associated conditions, the present disclosure provides for a decision support framework integrated with the detection and stratification outputs. This decision support system (DSS) is designed to assist clinicians and healthcare systems in selecting, optimizing, and adjusting treatment protocols in response to a subject’s molecular, metabolic, and behavioral profile.

[0147] Clinical Decision Support System (CDSS) Architecture: The CDSS receives inputs from genotyping results (e.g., GLP-1R-related SNPs), metabolic biomarkers (e.g., HbA1c, BMI, lipid profile), behavioral traits (e.g., eating behavior, adherence risk), and environmental context (e.g., socioeconomic status, access to care). These data are processed by rules-based engines and / or predictive ML algorithms to output: ^ Ranked treatment options (e.g., semaglutide vs liraglutide vs lifestyle-only; treatment regimens for GLP-1R associated condition is further described in the present disclosure) ^ Dosing suggestions (e.g., lower starting dose for high-risk adverse response genotypes) ^ Behavioral intervention pairings (e.g., CBT for patients with reward-seeking SNP profile) ^ Monitoring frequency guidance (e.g., CGM or quarterly HbA1c based on metabolic volatility)

[0148] As exemplified in FIG. 6, CDSS 602 can receive inputs from genotyping results 604 (e.g., GLP-1R-related SNPs), metabolic biomarkers 606 (e.g., HbA1c, BMI, lipid profile),Docket No.107976.00039 behavioral traits 608 (e.g., eating behavior, adherence risk, and environmental context 610 (e.g., socioeconomic status, access to care). These data can be processed by rules-based engines and / or predictive ML algorithms to output: ranked treatment options 612 (e.g., semaglutide vs liraglutide vs lifestyle-only). dosing suggestions 614 (e.g., lower starting dose for high-risk adverse response genotypes). behavioral intervention pairings 616 (e.g., CBT for patients with reward-seeking SNP profile), and monitoring frequency guidance 618 (e.g., CGM or quarterly HbA1c based on metabolic volatility).

[0149] Real-Time Integration in Clinical Workflow: The system can be implemented via: EMR-integrated modules to surface recommendations at point-of-care; alerts for contraindications or risk profiles (e.g., SNPs indicating adverse drug response); and / or patient- facing digital tools to support adherence and education, synchronized with clinician dashboards.

[0150] Feedback Loop & Adaptive Updating: Longitudinal patient outcomes (e.g., weight loss, glycemic control, adverse effects) are fed back into the system to refine future recommendations, creating a closed-loop learning system. In one embodiment, data are pooled across multiple patients (in de-identified form) to allow the model to evolve toward population- optimized and subpopulation-specific treatment strategies.

[0151] Support for Combination and Sequencing Decisions: In complex cases, the system may recommend sequential or combination therapies—e.g., starting with GLP-1R agonist plus behavioral therapy, followed by oral glucose-lowering agents if weight plateau occurs. The CDSS may also include eligibility mapping for advanced interventions such as bariatric surgery for high-risk, non-responsive individuals.

[0152] Cross-Stakeholder Utility: Beyond clinicians, the decision support outputs can inform pharmacists (for dispensing support based on genetic contraindications), insurers (via stratified reports supporting cost-effectiveness or tier eligibility), researchers (via anonymized aggregate outcome benchmarking), and patients (via personalized summaries of why a treatment path was selected).

[0153] The present disclosure further provides for expanded gene-drug interaction modeling, wherein both known and novel biomarkers—including those set forth in Tables 1–4—are used in concert to guide individualized therapeutic selection. By analyzing genetic variants across pharmacodynamic and pharmacokinetic pathways, this method supports rational drug choice, dose adjustment, and risk mitigation for GLP-1R-associated conditions.Docket No.107976.00039

[0154] Integration of Known Pharmacogenomic Markers: The present disclosure incorporates commercially recognized pharmacogenetic biomarkers, including but not limited to: ^ CYP2C9, CYP2C19, CYP3A5 – affecting metabolism of oral anti-diabetic drugs (e.g., sulfonylureas, thiazolidinediones) ^ TCF7L2 (rs7903146) – associated with poor sulfonylurea response; presence may favor GLP-1R agonists ^ SLCO1B1 (rs4149056) – influencing statin-induced myopathy risk, relevant in comorbid dyslipidemia ^ KCNJ11, ABCC8 – regulating pancreatic beta-cell activity, modifying response to sulfonylureas or GLP-1 therapies

[0155] These markers can be used in parallel with the GLP-1R pathway-related markers described herein (e.g., rs10305420, rs6923761, rs58428187), to inform whether a patient is likely to respond preferentially to: ^ A GLP-1R agonist (e.g., semaglutide, liraglutide) ^ A DPP-4 inhibitor (e.g., sitagliptin), or ^ Non-GLP options (e.g., metformin, SGLT2 inhibitors)

[0156] Combination Modeling: Predictive Weighting of Marker Panels: The decision model leverages both additive and interaction-based scoring: ^ Additive scoring: each biomarker (e.g., GLP-1R SNP + TCF7L2 SNP) contributes a weighted score toward a treatment axis (e.g., “GLP-1 favorable” vs “non-responder”) ^ Interaction-based models: certain SNP combinations may amplify or suppress treatment response (e.g., DRD2 + OPRM1 variants increasing behavioral response to GLP-1-induced satiety)

[0157] Behavioral + Metabolic Composite Consideration: The method enables multidimensional decision-making, such as: ^ Behaviorally high-risk individuals (e.g., with rs1800497 in DRD2): may be directed toward GLP-1R agonists plus cognitive-behavioral therapy to manage reward sensitivity ^ Metabolically insulin-resistant genotypes (e.g., GCK or G6PC2 variants): may receive GLP-1R agonists over SGLT2 inhibitors due to better postprandial control ^ Neural risk marker carriers (e.g., rs1051660 in OPRK1): may be prioritized for satiety- enhancing therapies over standard diet counselingDocket No.107976.00039

[0158] Expanded Coverage Enables Contraindication Flagging: Known pharmacogenomic liabilities (e.g., poor metabolizer status for CYP2C19 or adverse response predictors for GLP- 1R) can be detected and used to: ^ Avoid drug-induced adverse events (e.g., pancreatitis, hypoglycemia) ^ Guide selection toward safer alternatives ^ Adjust dosing ranges dynamically

[0159] Clinical and Regulatory Harmonization: The method aligns with CPIC, FDA, and EMA guidelines on gene-drug interactions, enabling clinical utility across regulated environments. Moreover, inclusion of novel GLP-1R-related biomarkers complements existing commercial tests by expanding clinical coverage into weight-loss, food addiction, and reward- related phenotypes.

[0160] Table 5 provides example gene-drug interaction pairs. The information provided herein can support treatment stratification, including multi-marker pharmacogenomic panels, biomarker-based dosing optimization, and obesity / metabolic disease stratification. Table 5. Gene–Drug Interaction Pairs Supporting Treatment Stratification Gene Associated Drugs Mechanistic or Clinical Rationale ANKK1 Contrave (naltrexone / bupropion) Dopamine signaling gene; bupropion is gDocket No.107976.00039 PLXNA4 Qsymia Implicated in melanocortin pathway and weight regulation

[0161] The present disclosure contemplates the use of a broad range of drug classes for the treatment or prevention of GLP-1R associated conditions, either alone or in combination with the predictive methods disclosed herein. These drug classes include both currently approved agents and investigational compounds, with emphasis on their mechanisms of action, therapeutic targets, and compatibility with the biomarker-guided stratification described above.

[0162] GLP-1 Receptor Agonists (GLP-1RAs): This class directly targets the glucagon-like peptide-1 receptor and mimics the activity of endogenous GLP-1. GLP-1RAs lower blood glucose by enhancing insulin secretion, suppressing glucagon release, delaying gastric emptying, and promoting satiety. Leading therapies include the following: ^ Semaglutide (Ozempic®, Wegovy®) – once-weekly injection, significant weight loss and glycemic benefits ^ Liraglutide (Victoza®, Saxenda®) – daily injection, approved for both type 2 diabetes and obesity ^ Dulaglutide (Trulicity®) – weekly injectable with cardiovascular outcome data ^ Exenatide (Byetta®, Bydureon®) – short-acting and extended-release formulations ^ Tirzepatide (Mounjaro®) – a dual GLP-1 / GIP receptor agonist showing superior metabolic and weight outcomes

[0163] DPP-4 Inhibitors (Gliptins): These agents inhibit the enzyme dipeptidyl peptidase-4, which degrades endogenous GLP-1, thereby modestly increasing GLP-1 levels. Examples include Sitagliptin (Januvia®), Saxagliptin, Alogliptin, and Linagliptin. DPP-4 inhibitors are generally weight-neutral and less potent than GLP-1RAs, but may be selected in biomarker- defined subgroups (e.g., when specific GLP-1R SNPs predict poor agonist response).

[0164] SGLT2 Inhibitors: Although not directly affecting GLP-1R pathways, SGLT2 inhibitors (e.g., empagliflozin, canagliflozin) are often part of combination regimens, especially in patients with GLP-1R polymorphisms associated with attenuated response to incretin-based therapy.Docket No.107976.00039

[0165] Oral Anti-Diabetic Agents include the following: ^ Metformin – first-line for type 2 diabetes; may be less effective in subjects with certain TCF7L2 or GLP-1R SNPs ^ Sulfonylureas – insulin secretagogues (e.g., glimepiride), often avoided in patients with high hypoglycemia risk or in those identified genetically as low responders ^ Thiazolidinediones (TZDs) – PPARγ agonists, effective in insulin resistance but may cause weight gain

[0166] Anti-Obesity Agents include the following: ^ Naltrexone / Bupropion (Contrave®) – targets reward circuits; response may be modulated by OPRM1 and DRD2 genotypes ^ Phentermine / Topiramate (Qsymia®) – appetite suppression; interaction with dopaminergic and serotonergic pathways may be affected by COMT and 5-HTR2A SNPs ^ Orlistat – lipase inhibitor with minimal systemic absorption

[0167] Adjunctive Behavioral and Cognitive Therapies: Based on stratified genetic and behavioral risk markers, non-pharmacologic interventions such as cognitive behavioral therapy (CBT), motivational interviewing (MI), or digital adherence programs may be recommended.

[0168] Further provided is a method of monitoring the GLP-1R associated condition and / or response to a GLP-1R agonist therapy in a subject. The method can further include providing a status of the one or more metabolic markers in the subject at two or more time points to provide a series of metabolic marker status, and monitoring the risk for developing the GLP-1R associated condition, monitoring a response to the treatment regimen, and / or adjusting or changing the treatment regimen in the subject based on the risk score and the series of metabolic marker status.

[0169] The polymorphism can include a single nucleotide polymorphism (SNP). Additionally or alternatively, the polymorphisms can include an insertion, deletion, translocation, or any variation at the locus.

[0170] In certain embodiments, once the polymorphism in the one or more GLP-1R related markers and optionally the status of the one or more metabolic markers and / or the polymorphism in the one or more neural response markers, the risk score is provided using a machine learning model. Certain embodiments include methods for predicting risk of a GLP-1R associated condition and / or predictingDocket No.107976.00039 a response to a GLP-1R agonist therapy using machine learning models and associated systems for providing decision support to healthcare professionals to implement a treatment regimen and mitigate the risk of the GLP-1R associated condition and enhance response to therapy. Certain embodiments include methods and systems for determining health outcomes and / or healthcare resource utilization based on evaluating the risk of a GLP-1R associated condition or response to GLP-1R agonist therapy using machine learning models. In certain embodiments, the systems and methods described herein are used by an insurance entity to provide insurance coverage for a treatment regimen. These embodiments include storing a subject’s risk of GLP-1R associated condition or response to GLP- 1R agonist therapy as part of the subject’s medical records. A health care provider, such as a physician or a nurse or a pharmacist, can access the subject’s risk of GLP-1R associated condition or likelihood of response to GLP-1R agonist therapy in the medical records to recommend a regimen for the subject for treatment or prevention of GLP-1 associated conditions.

[0171] Embodiments include computer implemented methods for predicting risk of developing a GLP-1R associated condition and / or a response to a GLP-1R agonist therapy in a patient using a computer system having one or more processors coupled to a memory storing one or more computer readable instructions for execution by the one or more processors. One such method includes the steps of storing a set of data containing a plurality of subject records. The subject records can include clinical data of each of the subject, such as age, sex, race, ethnicity, a lipid species profile, blood glucose level, blood hemoglobin A1c level, blood pressure, and waist circumference, and prior medical history. In an embodiment, the clinical data includes demographic data, socioeconomic data, and any data that about a subject that can be obtained by observation or oral or written communication. The subject records can also include data about health outcomes and healthcare resource utilization, such as quantification or description of the use of services by a subject for the purpose of preventing and curing GLP-1R associated condition or underlying disease, promoting maintenance of health and well-being, or obtaining information about one’s health status and prognosis. For example, an increased risk of GLP-1R associated condition may include poorer health outcomes and greater utilization of healthcare resources. Each subject record can include a GLP-1R related marker polymorphism profile for each subject of a plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition. Each subject record can also include a plurality of physical characteristics for each subject of a plurality of test subjects and subjects not diagnosed with a GLP-1R associated condition.Docket No.107976.00039

[0172] The method can further include the step of selecting a first subset of the subject records for the plurality of test subjects and a second subset of the subject records for the plurality of subjects not diagnosed with a GLP-1R associated condition (control subjects). The first and second subsets can serve as inputs into one or more initial machine learning models, where the input data include the GLP-1R related marker polymorphism profiles and optionally the metabolic marker status and / or the neural response marker polymorphism profiles. The method can further include the step of determining whether the one or more initial machine learning models meet a predetermined sensitivity, a predetermined specificity, and a predetermined accuracy for predicting the risk of developing the GLP-1R associated condition and / or response for a GLP-1R agonist therapy in the test subjects and the control subjects. The method can further include the step of generating an ensemble machine learning model responsive to the one or more initial machine learning models that meet the predetermined sensitivity, the predetermined specificity, and the predetermined accuracy. The method further includes the step of supplying a GLP-1R related marker polymorphism profile (and optionally the metabolic marker status and / or the neural response marker polymorphism profile) of a subject to the ensemble machine learning model as an input to obtain a risk score indicative of the subject’s risk of developing a GLP-1R associated condition and / or having a poor response to the GLP-1R agonist therapy. The method can further include the step of presenting the risk score or other indicator indicating the subject’s risk of developing a GLP-1R associated condition and / or having a poor response to the GLP-1R associated therapy to a healthcare professional for decision support to implement a treatment regimen and mitigate the subject’s risk of substance abuse or substance dependence.

[0173] In another embodiment, a non-transitory machine-readable storage medium may be encoded with instructions executable by a processing resource. As used herein, a “machine-readable storage medium” may be any electronic, magnetic, optical, or other physical storage apparatus to contain or store information such as executable instructions, data, and the like. For example, any machine- readable storage medium described herein may be any of random access memory (RAM), volatile memory, non-volatile memory, flash memory, a storage drive (e.g., hard drive), a solid state drive, any type of storage disc, and the like, or a combination thereof. The machine-readable storage medium may store or include instructions executable by the processor. The non-transitory machine-readable storage medium includes instructions to store one or more sets of data. Each set of data may include one or more allelic variants. Each of the one or more allelic variants are associated with a subject. Each of the one or more allelic variants includes one or more GLP-1R associated markerDocket No.107976.00039 polymorphism profiles (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles) of the subject and a value indicating whether the subject is diagnosed with a GLP-1R associated condition or responds to a GLP-1R agonist therapy). The non- transitory machine-readable storage medium can include instructions to generate one or more subsets of data based on the sets of data. Each subset of data may include a plurality of GLP-1R associated marker polymorphism profiles (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles). Each of the plurality of GLP-1R associated marker polymorphism profiles may include a same value indicating whether the subjects of the plurality of GLP-1R associated marker polymorphism profiles are diagnosed with the GLP-1R associated condition or responded well to GLP-1R agonist therapy. The non-transitory machine- readable storage medium can include instructions to train a stacked model with training data generated via creation of the subsets of data. The stacked model may include, at least, a random forest model and a support vector machine model. Each training data may correspond to one of the one or more subsets of data. Each of the one or more subsets of data may correspond to the same value indicating whether the subject is diagnosed with GLP-1R associated condition or responds to a GLP-1R agonist therapy, such that the training of the stacked model provides a score or other indicator that informs a subject’s risk of a GLP-1R associated condition and / or response to a GLP-1R agonist therapy.

[0174] In another embodiment, a non-transitory machine-readable storage medium may be encoded with instructions executable by a processing resource. The non-transitory machine-readable storage medium can include instructions to retrieve from memory, a trained stacked model. The stacked model may include, at least, a random forest model and a support vector machine model. The stacked model may be trained with one or more inputs and outputs derived from one or more sets of data. The inputs may include one or more GLP-1R associated marker polymorphism profiles (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles). Each of the GLP-1R associated marker polymorphism profiles (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles) may be associated with one or more allelic variants of the subject. The outputs may include a value indicating whether the subject has a risk of developing a GLP-1R associated condition or responds to a GLP-1R therapy. The non-transitory machine-readable storage medium can include instructions to, in response to receipt of an input of a sample, determine if the sample includes the one or more alleles associated with the GLP-1R associated marker polymorphism profiles. The non-transitory machine-readable storage medium can include instructions to, in response to a determination that the sample does notDocket No.107976.00039 include one of the one or more alleles associated with the GLP-1R associated marker polymorphism profiles, send a response to a user indicating that a prediction is not available. The non-transitory machine-readable storage medium can include instructions to, in response to a determination that the sample does include the one or more alleles associated with the GLP-1R associated marker polymorphism profiles, determine, using the retrieved trained stacked model and an average of the results of the random tree model and support vector machine model, a prediction of disposition to GLP-1R associated condition or response to GLP-1R agonist therapy relative to GLP-1R associated marker polymorphism profiles of the one or more alleles. A plurality of inputs to the retrieved trained stacked model may include all possible GLP-1R associated marker polymorphism profiles associated to the plurality of allelic variants.

[0175] Another embodiment of the disclosure is directed to a computer-implemented method of determining a subject’s risk of developing a GLP-1R associated condition and / or having a poor response to a GLP-1R agonist therapy. This risk may be expressed as a score or a probability or other indicator that informs a subject’s risk of a GLP-1R associated condition or response to a GLP-1R agonist therapy. In certain embodiments, the risk score is presented as a binary determination, such as one of Yes or No or one of Low Risk or High Risk or one of Poor Response and Good Response. In certain embodiments, the risk is presented as one of small set of possibilities, such as one of low risk, intermediate risk, or high risk, or poor response, intermediate response, or good response. In certain embodiments, the risk is presented as a score ranging from 0 to 1 or from 0 to 100. The method includes the steps of receiving a sample of the subject; analyzing the sample, via a processor of a computing device, to obtain a GLP-1R associated marker polymorphism profile of each of a set of specified allelic variants. In response to the GLP-1R associated marker polymorphism profiles of each of the set of specified allelic variants, the method further includes the step of determining, via the processor and a trained stacked model of the computing device, a score indicating the subject’s risk of developing a GLP-1R associated condition or response to GLP-1R agonist therapy based on the GLP-1R associated marker polymorphism profiles of each of the set of specified allelic variants. The method further includes the step of determining, via the processor, a treatment regimen recommendation for the subject based on the score; and transmitting, via the processor, the score and treatment regimen recommendation to a user device.

[0176] In an embodiment, the score is a value output from the trained stacked model between 0 and 1 and the trained stacked model includes a pre-determined threshold. In an embodiment, the score with a value greater than the threshold indicates the subject has a high risk of developing a GLP-1RDocket No.107976.00039 associated condition or having a poor response to a GLP-1R agonist therapy, and a score with a value less than or equal to the threshold indicates the subject has a low risk of developing a GLP-1R associated condition or having a good response to a GLP-1R agonist therapy. In an embodiment, the set of specified allelic variants includes 2 or more, 3 or more, 4 or more, 5 or more, 6 or more, 7 or more, 8 or more, 9 or more, 10 or more, 11 or more, 12 or more, 13 or more, 14 or more, 15 or more, 16 or more, 17 or more, 18 or more, 19 or more, 20 or more, 21 or more, 22 or more, 23 or more, 24 or more, 25 or more, 26 or more, 27 or more, 28 or more, 29 or more, 30 or more, 31 or more, 32 or more, 33 or more, 34 or more, 35 or more, 36 or more, 37 or more, 38 or more, 39 or more, 40 or more, 41 or more, 42 or more, 43 or more, 44 or more, 45 or more, 46 or more, 47 or more, 48 or more, 49 or more, 50 or more, 51 or more, 52 or more, 53 or more, 54 or more, 55 or more, 6=56 or more, or all 57 variants of the GLP-1R related markers set forth in Table 1. Optionally, the score is determined based also on one or more metabolic marker status and / or one or more neural response marker polymorphism profiles set forth in Table 4.

[0177] The method may include transmitting, via one or more processors of a computing device, a request for a sample of the subject. The method may include determining, via the one or more processors, a treatment regimen recommendation for the subject based on the risk score and transmitting, via the one or more processors, the risk score and treatment regimen recommendation to a user device and to the subject’s electronic medical records (EMR). In an embodiment, the score is a value output from the trained stacked model between 0 and 1 and the score with a value greater than the pre-determined threshold indicates the subject has a high risk of developing a GLP-1R associated condition or a poor response to a GLP-1R agonist therapy, and a score with a value less than or equal to the pre-determined threshold indicates the subject has a low risk of developing a GLP-1R associated condition or a poor response to a GLP-1R agonist therapy.

[0178] Further, the method may include, for example, re-weighting the initial set of decision- making rules based on the comparison to create a re-weighted set of decision-making rules. The re- weighted set of decision-making rules and the one or more weighted data inputs of the training data may be used to update the initial predictive model to generate the updated predictive model for predicting risk of GLP-1R associated condition or response to GLP-1R agonist therapy, or for providing decision support to healthcare professionals to implement a treatment regimen.

[0179] Another embodiment of the disclosure is directed to a system to determine a subject’s risk of developing a GLP-1R associated condition or response to a GLP-1R agonist therapy. The system can include an analyzer with an input for a sample. In an embodiment, the analyzer conducts an allelicDocket No.107976.00039 analysis of the sample received at the input to generate the GLP-1R associated marker polymorphism profiles (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles) for the trained ensemble model. The analysis, based on the received sample, may provide GLP-1R associated marker polymorphism profiles of each of a set of specified allelic variants (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles). The system can include a trained ensemble model, such as a trained stacked model. The trained ensemble model may include one or more inputs and an output. The inputs may accept one or more GLP-1R associated marker polymorphism profiles (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles) from the analyzer. Based on the GLP-1R associated marker polymorphism profiles (and optionally one or more metabolic marker status and / or one or more neural response marker polymorphism profiles), the trained ensemble model may provide, at the output, a score or other indicator that informs a subject’s risk of developing a GLP-1R associated condition or poorly responding to a GLP-1R agonist therapy. The system may include a user interface input / output in communication with a processor and machine-readable storage medium. The machine-readable storage medium may store instructions and the instructions may be executable by the processor. The executed instruction may, in response to a risk output from the trained ensemble model, determine an indicator to indicate a subject’s risk of developing GLP-1R associated condition or a poor response to a GLP-1R agonist therapy, the indicator based on the risk or probability. Further, the instructions may include transmission of the indicator to a user interface associated with a user device.

[0180] The embodiments described herein may utilize a trained ensemble or trained stacked model to determine a predicted predisposition to a GLP-1R associated condition or a predicted response to a GLP-1R agonist therapy of a subject. This approach may allow for prevention or modification of prescription of specific medications to subjects predisposed to a GLP-1R associated condition or predicted response to GLP-1R agonist therapy. For example, upon receipt of a sample, a system may provide a prediction of the likelihood of a subject developing a GLP-1R associated condition or responding to a GLP-1R agonist therapy within, depending on various factors, hours or days. Certain embodiments include methods and systems for determining health outcomes and / or healthcare resource utilization based on evaluating the risk of developing a GLP-1R associated condition or having a poor response to a GLP-1R agonist therapy using machine learning models. For example, an increased risk of GLP-1R associated condition may include poorer health outcomes and greater utilization of healthcare resources. An increased likelihood of a poor response to a GLP-1R agonistDocket No.107976.00039 therapy may indicate poorer health outcomes or utilization of healthcare resources other than a GLP- 1R agonist therapy. Certain embodiments include methods and systems for determining health or wellness of the subject based on the analysis of the allelic variants and / or clinical data of the subject. Certain embodiments include methods and systems for determining health outcomes of the subject based on the analysis of the allelic variants and / or clinical data of the subject. Certain embodiments include methods and systems for determining healthcare resource utilization by a subject based on the analysis of the allelic variants and / or clinical data of the subject.

[0181] Data encryption may be used to comply with business or regulatory rules, such as Health Insurance Portability and Accountability Act (HIPAA) privacy rules or the like. Such data encryption may ensure a secured communication of the determined initial diagnosis and the one or more associated recommendations between the user device and the third parties, thereby preventing the privacy of the communicated data to be compromised.

[0182] The section headings used herein are for organizational purposes only and are not to be construed as limiting the subject matter described in any way. While many embodiments of the present disclosure have been shown and described herein, it will be obvious to those skilled in the art that such embodiments are provided by way of example only. Numerous variations, changes, and substitutions will now occur to those skilled in the art without departing from the present disclosure. It should be understood that various alternatives to the embodiments of the present disclosure described herein may be employed in practicing the present disclosure. It is intended that the following claims define the scope of the present disclosure and that methods and structures within the scope of these claims and their equivalents be covered thereby.

Claims

Docket No.107976.00039 CLAIMS 1. A method for detecting a polymorphism in one or more glucagon-like protein-1 receptor (GLP-1R) related markers in a subject, the method comprising: genotyping a sample obtained from the subject to detect the presence of the polymorphism in the one or more GLP-1R related markers, the one or more GLP-1R related markers comprise one or more of rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467.

2. The method of claim 1, wherein the marker rs7997012 is located at position rs7997012 in the 5-HTR2A gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs453 is located at position rs4532 in the DRD1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1799971is located at position rs1799971 in the OPRM1 gene; the marker rs9479757 is located at position rs9479757 in the MUOR gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs10305492 is located at position rs10305492 in the GLP-1R gene; the marker rs10305420 is located at position rs10305420 in the GLP-1R gene; the marker rs2268639 is located at position rs2268639 in the GLP-1R gene; the marker rs2268640 is located at position rs2268640 in the GLP-1R gene; the marker rs57922 is located at position rs57922 in the GLP-1R gene; the marker rs58428187 is located at position rs58428187 in the GLP-1R gene; the marker rs6923761 is located at position rs6923761 in the GLP-1R gene; the marker rs9299870 is located at position rs9299870 in the GLP-1R gene; theDocket No.107976.00039 marker rs140226575 is located at position rs140226575 in the ARRB1 gene; the marker rs78979036 is located at position rs78979036 in the GLP-1R gene; the marker rs78052828 is located at position rs78052828 in the GLP-1R gene; the marker rs77501730 is located at position rs77501730 in the G6PC2 gene; the marker rs77501730 is located at position rs56100844 in the G6PC2 gene; the marker rs76895963 is located at position rs76895963 in the CCND2 gene; the marker rs115545608 is located at position rs115545608 in the G6PC2 gene; the marker rs16856115 is located at position rs16856115 in the G6PC2gene; the marker rs73174306 is located at position rs73174306 in the MECOM gene; the marker rs11708067 is located at position rs11708067 in the ADCY5 gene; the marker rs138917529 is located at position rs138917529 in the GCK gene; the marker rs1514895 is located at position rs1514895 in the SLC2A2 gene; the marker rs2168101 is located at position rs2168101 in the LM01 gene; the marker rs34222465 is located at position rs34222465 in the CACNA2D3 gene; the marker rs348330 is located at position rs348330 in the ABCB10 gene; the marker rs9379084 is located at position rs9379084 in the RREB1 gene; the marker rs10830963 is located at position rs10830963 in the MTNR1B gene; the marker rs183606969 is located at position rs183606969 in the GCK gene; the marker rs12692596 is located at position rs12692596 in the RBMS1 gene; the marker rs8192556 is located at position rs8192556 in the NEUROD1 gene; the marker rs118126621 is located at position rs118126621 in the ARMC2 and / or SESN1 gene; the marker rs78444298 is located at position rs78444298 in the EDEM3 gene; the marker rs35889227 is located at position rs35889227 in the FOXN3 gene; the marker rs11257655 is located at position rs11257655 in the CDC123 and / or CAMK1D gene; the marker rs6538804 is located at position rs6538804 in the RMST gene; the marker rs2255805 is located at position rs2255805 in the TSHZ2 gene; the marker rs17168486 is located at position rs17168486 in the DGKB and / or AGMO genes; the marker rs115128825 is located at position rs115128825 in the G6PC2 gene; the marker rs34814128 is located at position rs34814128 in the OR4C5 and / or OR4A47 genes; the marker rs10501320 is located at position rs10501320 in the MADD gene; the marker rs11592309 is located at position rs11592309 in the ADRA2A gene; the marker rs3842753 is located at position rs3842753 in the INS gene; the marker rs231362 is located at position rs231362 in the KCNQ1 gene; and / or the marker rs3765467 is located at position rs3765467 in the GLP-1R gene, in a human genome.

3. The method of claim 1 or 2, further comprising detecting a status of one or more metabolic markers, a polymorphism in one or more neural response markers, and / or a polymorphism in one or more additional GLP-1R related markers in the subject.Docket No.107976.00039 4. The method of claim 3, wherein the one or more metabolic markers comprise one or more of age, sex, a lipid species profile, blood glucose level, blood hemoglobin A1c level, blood pressure, and waist circumference.

5. The method of claim 3 or 4, wherein the one or more neural response markers comprise one or more of rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757, and / or wherein the one or more additional GLP-1R related markers are located in a gene selected from the group consisting of ANKK, FASN, GRIK1, HTR2C, INSR, LEPR, PCSK1, PLXNA4, POMC, TAAR1, TCF7L2.

6. A method of evaluating a risk for developing glucagon-like peptide-1 receptor (GLP-1R) associated condition or a response to a GLP-1R agonist therapy in a subject, the method comprising: detecting a polymorphism in one or more GLP-1R related markers in a sample obtained from the subject; providing a risk score based on the polymorphism in the one or more GLP-1R related markers, the risk score indicating the risk of developing the GLP-1R associated condition or the likelihood of a poor response to a GLP-1R agonist therapy; and selecting the subject as a candidate for a treatment regimen based on the risk score, the one or more GLP-1R related markers are one or more of rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467.Docket No.107976.00039 7. The method of claim 6, wherein the GLP-1R associated condition is type 2 diabetes, obesity, atherosclerotic cardiovascular disease, non-alcoholic fatty liver disease, polycystic ovary syndrome, food addiction, or addiction to cocaine, amphetamine, opioids, alcohol, or nicotine.

8. The method of claim 6 or 7, wherein the marker rs7997012 is located at position rs7997012 in the 5-HTR2A gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs453 is located at position rs4532 in the DRD1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1799971is located at position rs1799971 in the OPRM1 gene; the marker rs9479757 is located at position rs9479757 in the MUOR gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs10305492 is located at position rs10305492 in the GLP-1R gene; the marker rs10305420 is located at position rs10305420 in the GLP-1R gene; the marker rs2268639 is located at position rs2268639 in the GLP-1R gene; the marker rs2268640 is located at position rs2268640 in the GLP- 1R gene; the marker rs57922 is located at position rs57922 in the GLP-1R gene; the marker rs58428187 is located at position rs58428187 in the GLP-1R gene; the marker rs6923761 is located at position rs6923761 in the GLP-1R gene; the marker rs9299870 is located at position rs9299870 in the GLP-1R gene; the marker rs140226575 is located at position rs140226575 in the ARRB1 gene; the marker rs78979036 is located at position rs78979036 in the GLP-1R gene; the marker rs78052828 is located at position rs78052828 in the GLP-1R gene; the marker rs77501730 is located at position rs77501730 in the G6PC2 gene; the marker rs77501730 is located at position rs56100844 in the G6PC2 gene; the marker rs76895963 is located at position rs76895963 in the CCND2 gene; the marker rs115545608 is located at position rs115545608 in the G6PC2 gene; the marker rs16856115 is located at position rs16856115 in the G6PC2gene; the marker rs73174306 is located at position rs73174306 in the MECOM gene; the marker rs11708067 is located at position rs11708067 in the ADCY5 gene; the marker rs138917529 is located at position rs138917529 in the GCK gene; the marker rs1514895 is located at position rs1514895 in the SLC2A2 gene; the marker rs2168101 is located at position rs2168101 in the LM01 gene; the marker rs34222465 is located atDocket No.107976.00039 position rs34222465 in the CACNA2D3 gene; the marker rs348330 is located at position rs348330 in the ABCB10 gene; the marker rs9379084 is located at position rs9379084 in the RREB1 gene; the marker rs10830963 is located at position rs10830963 in the MTNR1B gene; the marker rs183606969 is located at position rs183606969 in the GCK gene; the marker rs12692596 is located at position rs12692596 in the RBMS1 gene; the marker rs8192556 is located at position rs8192556 in the NEUROD1 gene; the marker rs118126621 is located at position rs118126621 in the ARMC2 and / or SESN1 gene; the marker rs78444298 is located at position rs78444298 in the EDEM3 gene; the marker rs35889227 is located at position rs35889227 in the FOXN3 gene; the marker rs11257655 is located at position rs11257655 in the CDC123 and / or CAMK1D gene; the marker rs6538804 is located at position rs6538804 in the RMST gene; the marker rs2255805 is located at position rs2255805 in the TSHZ2 gene; the marker rs17168486 is located at position rs17168486 in the DGKB and / or AGMO genes; the marker rs115128825 is located at position rs115128825 in the G6PC2 gene; the marker rs34814128 is located at position rs34814128 in the OR4C5 and / or OR4A47 genes; the marker rs10501320 is located at position rs10501320 in the MADD gene; the marker rs11592309 is located at position rs11592309 in the ADRA2A gene; the marker rs3842753 is located at position rs3842753 in the INS gene; the marker rs231362 is located at position rs231362 in the KCNQ1 gene; and / or the marker rs3765467 is located at position rs3765467 in the GLP-1R gene, in a human genome.

9. The method of any one of claims 6 to 8, wherein the risk score is provided using a risk model.

10. The method of claim 9, wherein the risk model comprises a machine learning model.

11. The method of claim 10, wherein the machine learning model integrates heterogenous data modalities to produce a holistic subject profile, generates an individualized treatment path, classify a subject into a behavioral phenotype or a drug response phenotype matched to a treatment regimen, adapts to temporal trends through intake of longitudinal subject data, and / or produce outputs comprising payer-facing risk stratifications.

12. The method of claim 11, wherein the heterogenous data comprises genotyping results, metabolic biomarker status, behavioral data, and environmental data.Docket No.107976.00039 13. The method of claim 11 or 12, wherein the outputs comprise ranked treatment options, dosing suggestions, behavioral intervention parings, and / or monitoring frequency guidance.

14. The method of any one of claims 6 to 13, the method comprising: providing an individual risk score for each of a plurality of polymorphisms in the one or more GLP-1R related markers, thereby providing individual risk scores; and providing the risk score as a composite of the individual risk scores.

15. The method of claim 14, wherein the individual risk scores are weighted based on relevance to the risk of developing the GLP-1R associated condition, or the likelihood of having a poor response to the GLP-1R agonist therapy, to provide the risk score as the composite of the individual risk scores.

16. The method of any one of claims 6 to 15, wherein the risk score being greater than a predetermined threshold indicates the subject has an increased risk for developing type 2 diabetes or food addiction, or having a good response to the GLP-1R agonist therapy relative to a reference population, and wherein the risk score being lesser than the predetermined threshold indicates the subject has a reduced risk for developing type 2 diabetes or food addiction, or having a poor response to the GLP-1R agonist therapy relative to a reference population.

17. The method of any one of claims 6 to 16, the method further comprising: detecting a status of one or more metabolic markers, a polymorphism in one or more neural response markers, and / or a polymorphism in one or more additional GLP-1R related markers in the subject; and providing the risk score based on (i) the polymorphism in the one or more GLP-1R related markers and (ii) the status of the one or more metabolic markers, the polymorphism in the one or more neural response markers, and / or the polymorphism in one or more additional GLP-1R related markers.

18. The method of claim 17, wherein the one or more metabolic markers comprise one or more of age, sex, a lipid species profile, blood glucose level, blood hemoglobin A1c level, blood pressure, and waist circumference.Docket No.107976.00039 19. The method of claim 17 or 18, wherein the status of the one or more metabolic markers comprises data from continuous monitoring of blood glucose level in the subject.

20. The method of any one of claims 17 to 19, wherein the status of the one or more metabolic markers comprises data from continuous monitoring of blood hemoglobin A1c level in the subject.

21. The method of any one of claims 17 to 20, wherein the one or more neural response markers comprise one or more of rs7997012, rs948854, rs1045642, rs4680, rs6347, rs1800497, rs4532, rs1801133, rs1611115, rs2236861, rs1799971, rs3758653, rs211014, rs1051660, and rs9479757, and / or wherein the one or more additional GLP-1R related markers are located in a gene selected from the group consisting of ANKK, FASN, GRIK1, HTR2C, INSR, LEPR, PCSK1, PLXNA4, POMC, TAAR1, TCF7L2.

22. The method of claim 21, wherein the marker rs7997012 is located at position rs7997012 in the 5-HTR2A gene; the marker rs948854 is located at position rs948854 in the GAL gene; the marker rs1045642 is located at position rs1045642 in the ABCB1 gene; the marker rs4680 is located at position rs4680 in the COMT gene; the marker rs6347 is located at position rs6347 in the DAT1 gene; the marker rs1800497 is located at position rs1800497 in the DRD2 gene; the marker rs4532 is located at position rs4532 in the DRD1 gene; the marker rs1801133 is located at position rs1801133 in the MTHFR gene; the marker rs1611115 is located at position rs1611115 in the DBH gene; the marker rs2236861 is located at position rs2236861 in the DOR gene; the marker rs1799971 is located at position rs1799971 in the OPRM1 gene; the marker rs3758653 is located at position rs3758653 in the DRD4 gene; the marker rs211014 is located at position rs211014 in the GABA gene; the marker rs1051660 is located at position rs1051660 in the OPRK1 gene; and the marker rs9479757 is located at position rs9479757 in the MUOR gene, in the human genome.

23. The method of any one of claims 17 to 22, wherein the method further comprises: providing a status of the one or more metabolic markers in the subject at two or more time points to provide a series of metabolic marker status; and monitoring the risk for developing the GLP-1R associated condition, monitoring a response to the treatment regimen, and / or adjusting or changing the treatment regimen in the subject based on the risk score and the series of metabolic marker status.Docket No.107976.00039 24. The method of any one of claims 6 to 23, the method further comprising administering the treatment regimen to the subject.

25. The method of any one of claims 6 to 24, wherein the treatment regimen comprises a GLP- 1R agonist therapy, a DPP-4 inhibitor therapy, a SGLT2 inhibitor therapy, an anti-obesity agent therapy, a diet therapy, an exercise therapy, insulin therapy, oral diabetic medication, cognitive therapy, and / or behavioral therapy.

26. The method of claim 25, wherein the GLP-1R agonist therapy comprises semaglutide, liraglutide, dulaglutide, exenatide, or tirzepatide; wherein the DPP-4 inhibitor therapy comprises sitagliptin, saxagliptin, alogliptin, or linagliptin; wherein the SGLT2 inhibitor therapy comprises empagliflozin or canagliflozin; and / or wherein an anti-obesity agent therapy comprises naltrexone / bupropion, phentermine / topiramate, or orlistat.

27. The method of any one of claims 1 to 26, wherein the polymorphism comprises a single nucleotide polymorphism (SNP).

28. A system for predicting a risk for developing glucagon-like peptide-1 receptor (GLP-1R) associated condition or a response to a GLP-1R agonist therapy in a subject, the system comprising: a polymorphism analyzer to analyze a sample from the subject and to provide a first plurality of polymorphism profiles of each of a set of specified allelic variants in the sample; and a processor and a machine-readable storage medium storing (i) a trained ensemble model, the trained ensemble model being trained with inputs comprising a second plurality of polymorphism profiles associated with the specified set of allelic variants from a first plurality of test subjects and a third plurality of polymorphism profiles associated with the specified set of allelic variants from a second plurality of test subjects and outputs comprising a first plurality of outputs specifying that the first plurality of test subjects have been diagnosed with the GLP-1R associated condition and / or having good response to a GLP-1R agonist therapy, and a second plurality of outputs specifying that the second plurality of test subjects have not been diagnosedDocket No.107976.00039 with a GLP-1R associated condition and / or having poor response to a GLP-1R agonist therapy, and (ii) instructions, when executed by the processor, configured to: provide the trained ensemble model with the first plurality of polymorphism profiles to generate a score indicative of risk of developing a GLP-1R associated condition or responding poorly to a GLP-1R agonist therapy in the subject based on the first plurality of polymorphism profiles, in response to the score based on the first plurality of polymorphism profiles being greater than a pre-determined threshold, transmit to a user interface an output indicating a higher likelihood of the subject to develop the GLP-1R associated condition or a higher likelihood of the subject to respond to the GLP-1R agonist therapy, and in response to the score based on the first plurality of polymorphism profiles being less than or equal to the pre-determined threshold, transmit to the user interface an output indicating a lower likelihood of the subject to develop the GLP-1R associated condition or a lower likelihood of the subject to respond to the GLP-1R agonist therapy.

29. The system of claim 28, further comprising: a database storing a treatment regimen recommendation and an GLP-1R agonist-based treatment regimen recommendation.

30. The system of claim 28 or 29, further comprising: a device to collect and prepare a sample for the polymorphism analysis.

31. The system of any one of claims 28-30, wherein the machine-readable storage medium further has instructions to: in response to the score based on the first plurality of polymorphism profiles being greater than the pre-determined threshold, retrieve a treatment regimen recommendation from the database; and transmit to the user interface an output indicating the higher likelihood of the subject to develop the GLP-1R associated condition or the higher likelihood of the subject to respond to the GLP-1R agonist therapy, and the treatment regimen recommendation.

32. The system of any one of claims 28-31, wherein the trained ensemble model is generated by at least one of a random tree model and a support vector machine model.Docket No.107976.00039 33. The system of any one of claims 28-32, wherein the first plurality of polymorphism profiles comprise polymorphism in one or more of rs7997012, rs4680, rs4532, rs1800497, rs3758653, rs6347, rs1611115, rs1801133, rs1051660, rs211014, rs1799971, rs9479757, rs948854, rs2236861, rs1045642, rs10305492, rs10305420, rs2268639, rs2268640, rs57922, rs58428187, rs6923761, rs9299870, rs140226575, rs78979036, rs78052828, rs77501730, rs56100844, rs76895963, rs115545608, rs16856115, rs73174306, rs11708067, rs138917529, rs1514895, rs2168101, rs34222465, rs348330, rs9379084, rs10830963, rs183606969, rs12692596, rs8192556, rs118126621, rs78444298, rs35889227, rs11257655, rs6538804, rs2255805, rs17168486, rs115128825, rs34814128, rs10501320, rs11592309, rs3842753, rs231362, and rs3765467.

34. The system of claim 33, wherein the first plurality of polymorphism profiles further comprise polymorphism in one or more of GLP-1R, ANKK, FASN, GRIK1, HTR2C, INSR, LEPR, PCSK1, PLXNA4, POMC, TAAR1, and TCF7L2.

Images