Proteomics markers of human atherosclerosis

Abstract

Methods, kits, and computer-implemented systems are provided for diagnosing, predicting, or treating atherosclerosis or progression thereof in a subject.

Description

PROTEOMICS MARKERS OF HUMAN ATHEROSCLEROSISbyRavi ShahEric GamazonJ. Jeffrey CarrRavi Kai hanBassim El-SabawiPhillip LinAssignees: Vanderbilt University and Northwestern UniversityAttorney Docket No. : 11672N-25129WORELATED APPLICATIONS

[0001] This application claims priority from U.S. Provisional Application Serial No.63 / 755,879 filed February 7, 2025, the entire disclosure of which is incorporated herein by this reference.GOVERNMENT INTEREST

[0002] This invention was made with government support under HL122477 awarded by the National Institutes of Health. The government has certain rights in the invention.TECHNICAL FIELD

[0003] The presently disclosed subject matter generally relates the fields of medicine, proteomics, and atherosclerosis. More particularly, the disclosure relates to methods of diagnosing, predicting, and / or treating atherosclerosis or progression thereof in a subject.INTRODUCTION

[0004] Atherosclerotic cardiovascular disease (CVD) represents the leading cause of cardiovascular morbidity and mortality worldwide1and often develops silently over decades before the onset of major clinical events such as myocardial infarction or sudden cardiac death2. Despite extensive public health and pharmacologic efforts directed toward risk reduction, the prevalence of dysmetabolic states has continued to rise, contributing to increasing rates of early-onset cardiovascular disease, with some reports indicating that nearly half of cardiovascular deaths occur before age sixty-five3,4. This growing burden underscores the importance of delineating early, targetable mechanisms underlying atherosclerotic disease.

[0005] Coronary artery calcification (CAC) has emerged as a robust and quantifiable marker of subclinical coronary atherosclerosis5. CAC reflects a complex interplay of molecular activities occurring within the vascular endothelium and arterial wall, including local and systemic inflammation, immune activation, and metabolic and oxidative stress6. Although CAC is widely used as a clinical surrogate of early coronary disease, the biological processes that give rise to its initiation and progression remain incompletely defined.

[0006] Previous human studies have primarily relied on accessible systemic biomarkers, such as genetic variants, circulating metabolites, proteins, and transcriptomic signatures,7'9to investigate determinants of CAC and cardiovascular outcomes. However, these approaches necessarily rely on peripheral measurements rather than direct interrogation of disease-relevant coronary tissue6. Due to the inherent inaccessibility of human coronary arteries at scale, tissue-specific determinants of early atherosclerosis have remained insufficiently characterized, limiting efforts to identify molecular pathways most relevant to disease onset.

[0007] Although recent advancements in largescale genomic, transcriptomic, epigenomic, and proteomic resources have expanded opportunities to investigate cardiovascular biology, their application to coronary artery disease has often been limited by incomplete integration across data types, insufficient tissue resolution, and challenges in prioritizing candidate molecular targets. As a result, the field lacks fully resolved, multi-dimensional frameworks capable of discerning causal mechanisms in early human atherogenesis.

[0008] Accordingly, there remains a need in the art for improved, low- and non-invasive methods to for prediction, assessment, detection, risk stratification, and mechanistic understanding of atherosclerosis.SUMMARY

[0009] The presently disclosed subject matter meets some or all of the above-identified needs, as will become evident to those of ordinary skill in the art after a study of information provided in this document.

[0010] This Summary describes several embodiments of the presently disclosed subject matter, and in many cases lists variations and permutations of these embodiments. This Summary is merely exemplary of the numerous and varied embodiments. Mention of one or more representative features of a given embodiment is likewise exemplary. Such an embodiment can typically exist with or without the feature(s) mentioned; likewise, those features can be applied to other embodiments of the presently disclosed subject matter, whether listed in this Summary or not. To avoid excessive repetition, this Summary does not list or suggest all possible combinations of such features.

[0011] The presently disclosed subject matter relates to methods, kits, andcomputer-implemented methods for diagnosing or predicting atherosclerosis or progression thereof in a subject by assessing a plurality of biomarkers obtained from a biological sample. In various embodiments, the biomarkers comprise atherosclerosis-associated proteins or gene targets identified through integrative proteomic, transcriptomic, and genetic analyses, including those associated with prevalent or incident coronary artery calcification (CAC).

[0012] In certain embodiments, the invention provides methods in which a biological sample such as whole blood, plasma, or serum is obtained from the subject, and concentrations or expression levels of multiple biomarkers are quantified using one or more analytical platforms, including aptamer-based proteomic assays, immunoassays, mass-spectrometry-based proteomics, or transcriptomic workflows. The quantified biomarker measurements are processed using a multivariable statistical model or machine-learning model comprising biomarker-specific coefficients or parameters trained on atherosclerosis phenotypes, thereby generating anatherosclerosis risk score indicative of the subject’s likelihood of having or developing atherosclerosis. In some embodiments, the risk score is compared against a predetermined threshold, including thresholds derived from receiver operating characteristic (ROC) analyses or percentile-based reference distributions, to classify the subject as having increased risk.

[0013] The disclosed subject matter further provides kits comprising reagents for assessing selected biomarkers, including antibody pairs, aptamer-based binding agents, mass-spectrometry reagents, or nucleic-acid amplification or sequencing reagents, together with instructions for computing an atherosclerosis risk score using predetermined model coefficients. In some embodiments, the kits include a multiplexed assay cartridge, calibration standards, or an automated analytical device configured to quantify the biomarkers, normalize the resulting data, and transmit biomarker concentration data directly to a computing device executing the scoring algorithm.

[0014] Additional embodiments are directed to computer-implemented methods that receive biomarker concentrations or transcript levels as input, apply a model trained on atherosclerosis phenotypes, and output a continuous or categorical atherosclerosis risk score. These embodiments may include automated application of calibration curves, integration of clinical covariates such as age, sex, and race, generation of graphical outputs, and execution on local processors or cloud-based servers. In some embodiments, the computer-implemented method performs longitudinal analysis, computing risk scores at multiple time points to determine disease progression or therapeutic response.

[0015] In certain embodiments, the invention further comprises therapeutic decision-making, wherein subjects whose risk score exceeds a threshold receive a therapeutic agent or intervention selected from lipid-lowering medications, antihypertensive agents, antiplatelet therapies, or cardioprotective agents. The disclosed methods may additionally include ancillary coronary-risk testing, such as CAC imaging or cardiac stress testing, to corroborate and refine the biomarker-derived risk assessment.

[0016] The presently disclosed subject matter provides non-invasive, multi-omic, and computationally integrated approaches for early detection, risk stratification, and monitoring ofatherosclerosis, thereby improving accuracy relative to conventional single-analyte or non-tissue-specific biomarker assessments.BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The features of the invention are set forth with particularity in the appended claims. A better understanding of the features and advantages of the present invention will be obtained by reference to the following detailed description that sets forth illustrative embodiments, in which the principles of the invention are used, and the accompanying drawings of which:

[0018] FIG. 1. Study scheme illustrating multi-level approach for discovery in human coronary disease.

[0019] FIG. 2A-2C. Circulating proteomic architecture of human coronary calcification. (FIG. 2A) Volcano plot displaying the relationships of aptamers with the extent of CAC in the validation set. (FIG. 2B) Plot showing beta coefficients for prevalent CAC>0 in the derivation sample and incident CAC>0 for the aptamers with the top 20 and bottom 20 beta coefficients for incident CAC>0. (FIG. 2C) Heatmap illustrating the log2 -transformed, standardized, and winsorized levels of aptamers across all participants. The heatmap features 59 aptamers associated with the presence and extent of CAC at Year 25 in CARDIA, which were also associated with incident CAC at Year 35. The heatbars on the left represent the beta coefficients from models assessing prevalent CAC>0 at Year 25, extent of CAC at Year 25, and incident CAC at Year 35. The heatbar on the top shows ln(CAC+l) at Year 25 per subject.

[0020] FIG. 3A-3B. Replication of proteomic associations with CAC in the Framingham Heart Study. (FIG. 3A) Scatterplot showing aptamer effect sizes from the derivation sample in CARDIA versus the Framingham Heart Study (FHS) for ln(CAC+l). (FIG. 3B) Scatterplot showing aptamer effect sizes from the derivation sample in CARDIA versus FHS for CAC>0.

[0021] FIG. 4A-4D. Proteo-genomics of human coronary calcification. (FIG. 4A) Circular heatmap depicting 100 proteins associated with both the presence and extent of CAC and with available protein quantitative trait loci (pQTLs) as genetic instruments for Mendelian randomization analysis of CAC, coronary artery disease (CAD), and myocardial infarction (MI) from genome-wide association studies (GWAS). The heatmap illustrates the relationshipsbetween beta coefficients for proteomic associations with prevalent CAOO and incident CAOO, as well as inverse variance weighted (IVW) effect sizes for CAC, CAD, and MI, demonstrating overall concordance. The inner most tracks highlight genes with IVW nominal p-values < 0.05 with any outcome (CAC, CAD, or MI). (FIG. 4B) PWAS for coronary atherosclerosis in UK Biobank (UK Biobank code: I9 CORATHER). (FIG.4C) PWAS for myocardial infarction in UK Biobank (UK Biobank code: I9 MI). For (FIG.4B) and (FIG.4C), labeled proteins were associated with presence and extent of CAC in CARDIA (derivation P<0.05) and the PWAS (P<5xl0-3). Of 753 proteins with derivation P<0.05 for the proteomic association in CARDIA, 316 had available genetic models of protein abundance and could be tested in PWAS.(FIG. 4D) Quantile-quantile plot showing enrichment for (PWAS) associations of genetically determined protein abundance with coronary atherosclerosis in the UK Biobank (UK Biobank code: I9 CORATHER) among the proteins in circulation associated with presence and extent of CAC in CARDIA (derivation P<0.05).

[0022] FIG. 5A-5E. Human coronary transcriptome-wide association study of CAC and single cell transcription. (FIG. 5A) Transcriptome-wide association study of CAC (N=35,776) using genetic models of gene expression in human coronary samples. (FIG. 5B) Quantilequantile plot showing enrichment for TWAS associations among the proteins associated with presence and extent of CAC in CARDIA (derivation P<0.05). (FIG. 5C) Causal inference of CAC TWAS associations using Mendelian randomization. Causal inference was run using five different Mendelian randomization (MR) approaches: MR-JTI, MR-Egger, median-based, maximum likelihood, and inverse-variance weighted. Direction of effect is indicated by the direction of each triangle (right side up for positive and upside down for negative), and MR analyses that were statistically significant for causality are indicated with solid triangles. (FIG. 5D) Heatmap showing associations of the genetically determined expression of the 8 genes across heritable phenome captured in the UK Biobank. All TWAS associations have P-value < 0.05. (FIG. 5E) Gene expression profile from single-cell transcriptomics. Using single-cell RNA sequencing data in human atherosclerosis, the expression profile of the 8 genes across the cell types was identified.

[0023] FIG. 6A-6E. Human coronary functional genomics studies. (FIG. 6A) CAC GWAS SNPs linked to one of eight associated genes from convergent proteome and TWAS associations.A total of 11,795 CAC GWAS SNPs were identified that overlapped a coronary artery regulatory element in 3D contact (Hi-C) with one of the eight CAC-associated genes (HS6ST3, GPC6, S100A12, SPINK2, OAF, TNFSF12, RPP25, and NOTCH3). Of these SNPs, two (rs9515203 linked to GPC6 and rs28610385 linked ioRPP25)' were genome-wide significant (p < 5 x 10'8). (FIG.6B) LocusZoom plot of rs9515203. Top: Depicted are all functional SNPs (i.e., overlapping regulatory elements in coronary artery) linked to CAC-associated genes in coronary artery that neighbor rs9515203. Bottom: Zoomed out plot of rs9515203 contact locus. SNP rs9515203 overlaps with an insulator in coronary artery (designatedchrl3:l 10397000:110398000). This element falls within an intron of COL4A2 and makes distal 3D contact (Hi-C) with GPC6, a gene over 10 Mb upstream on chromosome 13. (FIG. 6C) Chromatin signature around rs9515203 implicates an insulator in coronary artery. Regulatory elements in this class are characterized by enrichment of CTCF and cohesin complex subunits RAD21 and SMC3. (FIG. 6D) Top: LocusZoom plot of rs28610385. Depicted are all functional SNPs (i.e., overlapping regulatory elements in coronary artery) linked to CAC-associated genes in coronary artery that neighbor rs28610385. Bottom: Zoomed in plot of rs28610385 contact locus. SNP rs28610385 overlaps with a transcribed gene body element in coronary artery (designated chrl 5:78798000:78799000). This element partially overlaps an intron o ADAMTS7 and makes distal 3D contact (Hi-C) with RPP25. (FIG.6E) Chromatin signature around rs28610385 shows a transcribed gene body element in coronary artery. Regulatory elements in this class are characterized by chromatin accessibility (in this case, measured by ATAC-seq), as well as the histone marks H3K36me3, H3K79me2, and H4K20mel. Enrichment of these 3 chromatin marks were observed throughout the locus of rs28610385 in the coronary artery epigenome.

[0024] FIG. 7A-7C. Single-cell-based differential expression analysis of GPC6 and RPP25. Feature maps show (FIG. 7A) GPC6 and (FIG. 7B) RPP25 expression profiles in human coronary artery single-cell transcriptomics. (FIG. 7C) Both genes show differential expression effects in atherosclerosis in specific cell types derived from human coronary artery, adding functional support to the GWAS variant-target gene relationship identified by epigenomic, chromatin, and transcription factor binding experimental data in the same tissue.DESCRIPTION OF EXEMPLARY EMBODIMENTS

[0025] The details of one or more embodiments of the presently disclosed subject matter are set forth in this document. Modifications to embodiments described in this document, and other embodiments, will be evident to those of ordinary skill in the art after a study of the information provided in this document. The information provided in this document, and particularly the specific details of the described exemplary embodiments, is provided primarily for clearness of understanding and no unnecessary limitations are to be understood therefrom. In case of conflict, the specification of this document, including definitions, will control.

[0026] The presently disclosed subject matter includes methods for diagnosing or predicting atherosclerosis or progression thereof in a subject, kits for diagnosing or predicting atherosclerosis or progression thereof in a subject, and computer-implemented methods of diagnosing or predicting risk of atherosclerosis or progression thereof in a subject. Embodiments of the presently disclosed subject matter make use of assessment of a plurality of biomarkers disclosed herein, obtained from a biological sample.

[0027] The plurality of biomarkers can be selected from a defined group of atherosclerosis-associated proteins or gene targets, as disclosed herein. In certain embodiments, biomarkers are identified through the integrative proteomic, transcriptomic, and genetic association analyses described herein, including those associated with prevalent and incident coronary artery calcification. In certain embodiments, the plurality of biomarkers comprises a prioritized subset of markers that exhibit multi-level support across circulating proteomic association analyses, coronary-artery transcriptome-wide association studies, Mendelian randomization, and functional genomic assays.

[0028] In some embodiment, the biomarkers are selected from the group consisting of STAR, NUCB1, ANTXR1, MXRA8, SLITRK3, B4GALT6, ART3, SVEP1, DNAJB9, PCOLCE, APOA5, WFDC2, TREM2, TRA2B, RPL30, EFS, GABARAPL1, SRSF7, EWSR1, WFIKKN2, IL22, KIF3A, TPT1, IGFBP6, CNTFR, TRAPPC3, VOPP1, CDH1, RBBP4, CD248, NPTXR, GLIPR2, MSR1, SLITRK1, ANTXR2, NCAN, HTRA1, EPHA4, GIP, GSN, CDCP1, OMG, HS6ST3, ATOX1, RBP5, CSTB, MYL4, NCAM1, ITGAV|ITGB3, DSG2, MAG, GCHFR, DCC, C1QL3, FBLN7, MED11, PPP1R10, EPHA6, MAPK6, CCL22, TIMP1,PYCARD, RIC8A, VEGFA, ANGPT2, NTRK3, EGFR, SIGLEC7, MMP7, IGFBP4, CNTN1, LGALS4, CCL3, RARRES2, LYVE1, PIGR, SERPIND1, RGMB, THBS2, SERPINC1, BCAN, GDF15, MMP12, SLITRK5, HBA1|HBB, CAPG, N0TCH1, NOTCH3, S100A9, SEMA3E, FGFR1, RNASE6, SIRPB2, CLEC3B, LEAP2, ABHD14A, EHMT2, PTPRS, ARHGAP36, CLSTN3, BAGE2, C1QTNF1, CCDC126, IGLON5, ENPP5, ALPG, LRRTM2, ADAM23, VWA2, ATP1B2, ZP4, ALPP, SCO1, SLAMF1, SCG3, TMPO, VTN, CMPK1, CILP2, CACNA2D3, TMEM132B, TFF2, LGALS9, UBE2G2, TREM1, REG3A, PTPRD, KITLG, B4GALT2, YBX1, and IGDCC4. In some embodiments, the biomarkers are selected from those set forth in Tables 4A and 4B. In some embodiment, the biomarkers are selected from the group consisting ofNOTCH3, SPINK2, S100A12, RPP25, OAF, HS6ST3, TNFSF12, and GPC6.

[0029] The proteins and / or gene targets in the biological sample can be assessed using any suitable analytical method routine and known to one of ordinary skill in the art, with consideration to the type of biological sample being employed. Such methods include but not limited to aptamer-based proteomic assays, immunoassays (e.g., ELISA, multiplex immunoassays, bead-based assays), mass-spectrometry-based proteomics (e.g., targeted or untargeted LC-MS / MS), or nucleic-acid-based assays for quantifying gene expression (e.g., quantitative PCR, digital PCR, microarray analysis, or RNA-sequencing). In some embodiments, circulating proteins are measured using high-throughput proteomic platforms capable of simultaneously quantifying hundreds to thousands of proteins from a blood-derived sample. In other embodiments, assessment of gene targets comprises measurement of transcript abundance using any suitable transcriptomic workflow, including sequencing-based or hybridization-based modalities. These and other routine analytical techniques may be employed alone or in combination to determine the level of one or more biomarkers in the biological sample.

[0030] In certain embodiments, quantifying concentrations comprises measuring a panel of markers selected from those disclosed herein. The panel can include as few as two proteins or gene targets, or up to all proteins or gene targets listed in Tables 4A and 4B. Representative panels may include any number of biomarkers within these ranges, for example 2 to 130 proteins or gene targets. In some embodiments the panel comprises 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 of the proteins or gene targets. In some embodiments the panelcomprises 2, 3, 4, 5, 6, 7, or 8 of the proteins or gene targets in the group consisting of NOTCH3, SPINK2, S100A12, RPP25, OAF, HS6ST3, TNFSF12, and GPC6.

[0031] In certain embodiments, the selection of proteins or gene targets for a given panel is based on ranking by absolute value of the coefficient, biological plausibility, and / or technical feasibility for targeted proteomic analysis.

[0032] In certain embodiments, the biomarkers identified herein are proteins. These proteins are present in plasma or serum and can be quantified using proteomic techniques known in the art, such as mass spectrometry, immunoassays, or automated proteomic analysis systems. In certain embodiments, as an alternative to or in addition to quantifying concentrations of the identified proteins, assessment of atherosclerosis or risk can comprise quantifying gene expression products (e.g., mRNA transcripts) that encode the identified proteins. Such quantification can be performed using nucleic acid amplification or sequencing techniques known in the art, including but not limited to quantitative PCR, digital PCR, or next-generation sequencing. These embodiments provide flexibility for implementing biomarker-based assessment without limiting the approach to proteomic analysis.

[0033] In certain embodiments, proteomic and transcriptomic data are integrated to enhance the accuracy of predicting atherosclerosis or progression thereof. For example, concentrations of circulating proteins may be combined with transcript abundance of corresponding gene targets in a multivariable model to generate a composite atherosclerosis risk score. Integrating proteomic and transcriptomic features provides complementary biological information and supports construction of improved risk-stratification algorithms relative to the use of either data type alone.

[0034] The presently disclosed subject matter includes methods for diagnosing or predicting atherosclerosis or progression thereof in a subject by assessing, in a biological sample obtained from the subject, a plurality of biomarkers selected from the biomarkers disclosed herein whose molecular associations with coronary artery calcification and related atherosclerotic phenotypes were identified through largescale proteomic, transcriptomic, and integrative genetic analyses. In certain embodiments, the method comprises obtaining a biological sample, such as whole blood, plasma, or serum, from the subject and quantifying a plurality of circulating proteins orgene-expression markers associated with subclinical or progressive coronary artery disease. The quantified biomarker measurements may then be processed in a quantitative analytical workflow to generate an atherosclerosis risk score that reflects the subject’s likelihood of harboring or developing coronary atherosclerosis, including phenotypes such as coronary artery calcification. In some embodiments, the method further comprises identifying the subject as having atherosclerosis or an increased risk thereof when the atherosclerosis risk score exceeds a predetermined threshold selected to optimize discrimination performance.

[0035] In certain embodiments, the biological sample comprises a blood-derived sample, including whole blood, plasma, or serum, from which the plurality of biomarkers is measured. In certain embodiments, the biological sample is subjected to one or more pre-analytical preparation steps, including centrifugation, component separation, buffer exchange, protein stabilization, denaturation, normalization, or other processing steps suitable for quantitative measurement of the biomarkers, and may further include routine sample-handling steps used in clinical laboratory workflows (e.g., storage, thawing, pre-analytical quality control).

[0036] In certain embodiments, assessing the plurality of biomarkers comprises quantifying circulating protein levels using an aptamer-based proteomic platform configured to measure hundreds to thousands of proteins in parallel. Such platforms employ nucleic-acid-based affinity reagents designed to bind selectively to target proteins, facilitating high-throughput, multiplexed detection suitable for generating quantitative inputs for the atherosclerosis risk score.

[0037] In certain embodiments, assessing the plurality of biomarkers comprises performing an immunoassay-based quantification, including formats such as enzyme-linked immunosorbent assays or multiplex bead-based immunoassays, which allow simultaneous or sequential detection of multiple atherosclerosis-associated biomarkers within the biological sample. Such immunoassay approaches provide analyte-specific detection through antibody-antigen binding.

[0038] In certain embodiments, assessing the plurality of biomarkers comprises performing a mass-spectrometry-based proteomic assay, including targeted or untargetedliquid-chromatography tandem mass spectrometry workflows. Such assays can be used to quantify peptides derived from the biomarkers following optional sample digestion, providinghigh-resolution identification and quantification of circulating proteins associated with atherosclerosis.

[0039] In certain embodiments, assessing the plurality of biomarkers comprises obtaining transcriptomic data, including single-cell or bulk transcriptomic measurements, for genes corresponding to the biomarkers. Such transcriptomic data may be generated using sequencing-based or hybridization-based methods and can be incorporated as inputs into the atherosclerosis risk-scoring workflow alongside, or in place of, proteomic measurements.

[0040] In certain embodiments, assessing the plurality of biomarkers is performed using an automated analytical device configured to process the biological sample and quantify the biomarkers using a calibration curve tailored to atherosclerosis-associated biomarker concentration ranges. The device may incorporate automated signal acquisition, data normalization, and output of quantitative biomarker values to ensure consistent measurement performance across analytical runs.

[0041] In certain embodiments, generating the atherosclerosis risk score comprises applying a multivariable regression model to the quantified concentrations or expression levels of the plurality of biomarkers. In such embodiments, the model may include biomarker-specific coefficients that weight each marker according to its contribution to predicting atherosclerosis or progression thereof. The regression model may be linear, logistic, or otherwise configured to output a continuous value that reflects the subject’s relative risk.

[0042] In certain embodiments, generating the atherosclerosis risk score comprises applying a machine-learning model trained on atherosclerosis phenotypes. The machine-learning model may include biomarker-specific parameters or learned feature weights and may be developed using labeled training data from populations for which biomarker values and atherosclerosis outcomes are known. Representative machine-learning approaches include supervised algorithms configured to model nonlinear relationships among biomarkers and to improve predictive performance relative to conventional statistical models.

[0043] In certain embodiments, the multivariable model is validated using receiver operating characteristic (ROC) analysis, and the threshold value is selected to achieve a desired balance ofsensitivity and specificity. The ROC-derived threshold may correspond to the point on the curve that maximizes discriminative performance, such as the Youden index, or may be selected based on clinical or operational criteria relevant to detection of atherosclerosis or progression thereof.

[0044] In certain embodiments, the threshold value against which the atherosclerosis risk score is compared corresponds to a predetermined percentile of risk scores within a reference population lacking clinical manifestations of atherosclerosis. Such percentile-based thresholds provide for normalization of risk scores across populations and facilitate classification of subjects whose biomarker-based profiles exceed expected distributions for clinically healthy individuals.

[0045] In certain embodiments, generating the atherosclerosis risk score further comprises integrating one or more clinical covariates, including age, sex, and race, into the multivariable model. Incorporating such covariates facilitates refinement of the risk score by accounting for subject- specific factors known to influence the prevalence or severity of atherosclerosis, thereby improving the accuracy and clinical interpretability of the resulting risk classification.

[0046] In certain embodiments, generating the atherosclerosis risk score further comprises adjusting the quantified biomarker concentrations or expression levels for one or more covariates selected from age, sex, and race prior to, or as part of, the scoring algorithm. Such adjustments may reduce confounding, harmonize measurements across heterogeneous populations, and provide for more robust comparison of biomarker-based risk scores across subjects.

[0047] In certain embodiments, generating the atherosclerosis risk score comprises applying a multivariable model that includes biomarker-specific coefficients derived from a regression analysis or a machine-learning algorithm trained on atherosclerosis phenotypes. The model may incorporate the quantified concentrations or expression levels of the biomarkers to produce a continuous numerical output representative of the subject’s risk.

[0048] In certain embodiments, generating the atherosclerosis risk score is performed by a computing device executing instructions stored on a non-transitory computer-readable medium, where the instructions cause the computing device to: (i) receive the quantified biomarker concentrations or expression levels; (ii) apply the multivariable model comprisingbiomarker-specific coefficients; and (iii) output the resulting atherosclerosis risk score together with an indication of whether the score exceeds a threshold value associated with increased risk.

[0049] In certain embodiments, the multivariable model may be periodically retrained or updated using newly acquired biomarker datasets, provided that the model structure continues to apply biomarker-specific coefficients or parameters learned from atherosclerosis-related phenotypes.

[0050] In certain embodiments, the method further comprises administering to the subject a therapeutic agent or intervention for treating or preventing atherosclerosis when the atherosclerosis risk score exceeds the threshold value. In such embodiments, treatment initiation is guided by the biomarker-based risk classification, facilitating targeted application of preventive or therapeutic measures in subjects identified as being at elevated risk.

[0051] Examples of therapeutic agents include those that lower Low-Density Lipoprotein Cholesterol (LDL-C) and / or Apolipoprotein B (ApoB), the primary structural protein found on all atherogenic, cholesterol-containing lipoproteins. These treatments effectively reduce the total volume of circulating cholesterol and / or the actual number of particles that drive plaque formation. Representative therapeutic agents in this category include Statins, Ezetimibe, Bempedoic Acid, Bile Acid Sequestrants, and Proprotein Convertase Subtilisin / Kexin type 9 (PCSK9) Inhibitors. Additionally, agents such as Fibrates and Niacin (Nicotinic Acid) may be utilized to lower triglyceride levels and reduce the burden of triglyceride-rich, ApoB-containing lipoproteins, thereby improving the overall lipid profde.

[0052] Additional examples of therapeutic agents include agents that regulate blood pressure to reduce arterial strain, such as Angiotensin-Converting Enzyme (ACE) Inhibitors, Angiotensin II Receptor Blockers (ARBs), and Beta-Blockers.

[0053] Additional examples of therapeutic agents include antiplatelet medications, such as Aspirin or Clopidogrel, to prevent the formation of blood clots. Further examples of therapeutic agents include Sodium-Glucose Cotransporter-2 (SGLT2) Inhibitors and Glucagon-Like Peptide-1 (GLP-1) Receptor Agonists, which have been identified for their significant cardioprotective benefits.

[0054] In certain embodiments, the therapeutic agents alter the abundance of one or more atherosclerosis-associated biomarkers. In certain embodiments, administration of the therapeutic agent results in measurable changes in one or more biomarkers disclosed herein, providing for longitudinal monitoring under the disclosed method.

[0055] Beyond pharmacologic therapy, clinical management includes structured lifestyle interventions, such as diet modification and exercise programs. Additionally, the method may comprise further diagnostic testing for coronary risk, including Coronary Artery Calcium (CAC) Scoring, Echocardiography, Cardiac Catheterization, or Stress Testing.

[0056] In certain embodiments, the therapeutic agent administered to the subject is selected from a statin, a bile sequestrant, niacin, a PCSK-9 inhibitor, or a fibrate, or any combination thereof. These agents may be used to reduce lipid burden, modulate inflammation, or otherwise attenuate the molecular processes underlying atherosclerosis progression in subjects identified as high-risk based on the computed atherosclerosis risk score.

[0057] In certain embodiments, the method further comprises performing one or more additional tests for coronary risk, such as coronary artery calcification imaging, echocardiography, cardiac catheterization, or stress testing. These ancillary evaluations may be used to corroborate, refine, or contextualize the biomarker-derived atherosclerosis risk score, providing a complementary assessment of coronary structure or function in subjects identified as being at elevated risk.

[0058] The presently disclosed subject matter includes kits for diagnosing or predicting atherosclerosis or progression thereof in a subject. In such embodiments, the kit comprises a plurality of reagents configured for assessing, in a biological sample obtained from the subject, a plurality of biomarkers selected from the biomarkers disclosed herein. The kit further comprises instructions, provided in printed form or stored on a non-transitory computer-readable medium, that direct a user or computing device to calculate an atherosclerosis risk score based on quantified concentrations or expression levels of the biomarkers. In certain embodiments, the instructions specify that the risk score is calculated using predetermined coefficients or other model parameters derived from statistical or computational modeling trained on atherosclerosis-related phenotypes.

[0059] In certain embodiments, the plurality of reagents included in the kit is provided within a multiplexed assay cartridge configured for simultaneous detection of two or more biomarkers in a single analytical run. Such cartridges may include physically discrete reaction regions, wells, or microfluidic channels pre-loaded with biomarker-specific reagents to facilitate parallel assessment of multiple proteins or gene-expression targets using a common sample input volume.

[0060] In certain embodiments, the multiplexed assay cartridge comprises a non-standard calibration curve optimized for concentration ranges characteristic of atherosclerosis-associated biomarkers. Such calibration curves may be constructed using reference standards selected to reflect low-abundance or high-dynamic-range analytes identified through the multi-omic analyses described herein, facilitating quantification within clinically relevant ranges.

[0061] In certain embodiments, the reagents included in the kit comprise antibody pairs configured for use in a sandwich immunoassay for detecting one or more of the disclosed atherosclerosis biomarkers. The antibody pairs may include capture and detection antibodies that selectively bind to distinct epitopes of the target biomarker, facilitating analyte-specific signal generation in enzyme-linked or bead-based immunoassay formats.

[0062] In certain embodiments, the plurality of reagents included in the kit comprises aptamer-based binding agents configured to selectively bind to one or more of the disclosed atherosclerosis-associated biomarkers. Aptamer-based reagents may include chemically modified nucleic-acid sequences engineered for high-affinity and high-specificity binding to target proteins, thereby allowing quantitative detection of circulating biomarkers using aptamer-based proteomic platforms. Such reagents are suitable for integration into multiplexed assay cartridges and may facilitate simultaneous measurement of multiple biomarkers disclosed herein, including those identified through proteomic and transcriptomic analyses indicative of coronary artery calcification and related phenotypes.

[0063] In certain embodiments, the kit further comprises mass-spectrometry compatible reagents suitable for preparing the biological sample for targeted or untargeted proteomic analysis. Such reagents may include, without limitation, protein denaturation buffers, enzymatic digestion enzymes, ionization-enhancing reagents, or solvent systems compatible withliquid-chromatography tandem mass-spectrometry workflows. These reagents allow efficient generation of peptide fragments corresponding to the disclosed biomarkers, including but not limited to those present in plasma or serum, thereby supporting quantitative assessment of biomarker abundance using mass-spectrometry-based detection modalities.

[0064] In certain embodiments, the kit further comprises nucleic-acid amplification or sequencing reagents configured to measure transcript levels corresponding to one or more of the disclosed biomarkers. Such reagents may support amplification-based workflows (e g., quantitative PCR, digital PCR) or sequencing-based workflows (e.g., RNA sequencing, single-cell RNA sequencing) suitable for quantifying transcriptomic biomarkers that correlate with atherosclerosis phenotypes. Inclusion of these reagents allows for assessment of gene-expression markers that may complement or substitute for proteomic measurements when generating the atherosclerosis risk score and facilitates multi-omic integration to improve diagnostic or predictive performance.

[0065] In certain embodiments, the kit further comprises an automated immunoassay or analytical device configured to process the multiplexed assay cartridge and quantify one or more of the disclosed atherosclerosis-associated biomarkers. The automated device may include hardware and embedded software configured to perform sample handling, reagent mixing, incubation, signal acquisition, and data preprocessing with minimal user intervention. The device may additionally be calibrated using biomarker-specific calibration curves tailored to atherosclerosis-associated concentration ranges, thereby ensuring consistent and reproducible measurement performance across analytical runs.

[0066] In certain embodiments, the automated device is configured to transmit biomarker concentration data to a computing device executing the atherosclerosis risk-scoring algorithm. Transmission may occur via wired or wireless communication protocols and may include automated formatting of quantitative biomarker values in a structure suitable for model input. This integration allows for seamless workflow connectivity, permitting the computing device to receive quantified biomarker concentrations or expression levels directly from the analytical device and apply predetermined coefficients or model parameters to generate the atherosclerosis risk score in real time.

[0067] In certain embodiments, the kit further comprises calibration standards containing predetermined concentrations of at least one of the disclosed biomarkers. Such calibration standards may be used to calibrate the automated device, validate assay performance, and ensure quantitative accuracy across analytical runs. The standards may include purified proteins, synthetic peptides, or other reference materials selected to reflect clinically relevant ranges of atherosclerosis-associated biomarker abundance. Use of such calibration standards enhances reproducibility and supports rigorous quantification necessary for accurate computation of the atherosclerosis risk score.

[0068] In certain embodiments, the instructions stored on the non-transitorycomputer-readable medium comprise executable code for applying a multivariable model configured to generate the atherosclerosis risk score from quantified concentrations or expression levels of the biomarkers. The multivariable model may comprise biomarker-specific coefficients derived from regression analysis or machine-learning training on atherosclerosis phenotypes, including coronary artery calcification and related molecular signatures. Execution of the code allows for automated application of the trained model to newly acquired biomarker data to produce a quantitative risk estimate.

[0069] In certain embodiments, the instructions further comprise executable code for comparing the computed atherosclerosis risk score to a threshold value. The threshold value may be predetermined based on performance characteristics derived from receiver operating characteristic (ROC) curve analysis used to classify subjects according to the likelihood or presence of atherosclerosis. The executable code facilitates automated determination of whether the computed risk score exceeds this ROC-based threshold, thereby facilitating reproducible and objective subject classification.

[0070] In certain embodiments, the instructions further comprise executable code for outputting a risk classification that indicates whether the subject is at increased risk of atherosclerosis or progression thereof. Such code may generate an output that includes categorical indicators, numerical scores, alerts, or interpretive risk tiers derived from the biomarker-based model. The output may be configured for presentation on a graphical userinterface or for storage in an electronic medical record system, thereby facilitating clinical interpretation and downstream decision-making.

[0071] In certain embodiments, the instructions further comprise executable code for integrating clinical covariates, including age, sex, and race, into the multivariable model used to compute the atherosclerosis risk score. Incorporation of these covariates allows for adjustment of biomarker-derived risk estimates to account for subject-specific demographic factors known to influence coronary artery disease susceptibility. The covariate-integration functionality may be embedded within the scoring algorithm or applied during preprocessing of biomarker measurements.

[0072] In certain embodiments, the computing device is configured to automatically trigger the risk-score calculation upon receiving biomarker concentration data transmitted from the analytical device. Automatic initiation of the scoring workflow reduces manual intervention, enhances laboratory throughput, and ensures consistent application of the scoring algorithm across analytical runs. This automation supports seamless integration of biomarker quantification and computational risk assessment within a unified diagnostic workflow.

[0073] In certain embodiments, the instructions further comprise executable code for storing and comparing biomarker profiles across multiple time points to support longitudinal risk assessment. Such functionality allows the computing device to maintain historical biomarker datasets for a given subject, compute successive atherosclerosis risk scores, and identify trends or changes in biomarker-based risk over time. Longitudinal tracking facilitates monitoring of disease progression, therapeutic response, or changes in subclinical atherosclerotic activity.

[0074] The presently disclosed subject matter further includes computer-implemented methods for diagnosing or predicting atherosclerosis or progression thereof in a subject. In such embodiments, one or more computing devices execute machine-readable instructions stored on a non-transitory computer-readable medium to process biomarker data and generate an atherosclerosis risk score. The input data may comprise quantified concentrations or expression levels of a plurality of biomarkers selected from the biomarkers disclosed herein, obtained from biological samples such as plasma or serum and optionally generated by the kits and analytical workflows described herein.

[0075] In certain embodiments, the computer-implemented method comprises receiving, as input, quantified concentrations or expression levels of a plurality of biomarkers as disclosed herein, and applying a multivariable model trained on atherosclerosis phenotypes to the input data. The multivariable model comprises biomarker-specific coefficients derived from regression analysis or machine-learning training using datasets that include atherosclerosis-related phenotypes, such as coronary artery calcification and related molecular signatures. Execution of the model on the input biomarker data yields a computed atherosclerosis risk score that reflects the subject’s atherosclerotic status or risk.

[0076] In certain embodiments, the computer-implemented method further comprises generating and outputting an atherosclerosis risk score indicative of the subject’s atherosclerosis status. The generated risk score may be represented as a continuous numerical value or transformed into categorical risk strata and may be output in a format suitable for clinical interpretation or further computational processing. In some embodiments, the output includes both the computed atherosclerosis risk score and an associated indication of the subject’s relative risk or disease status, which may be displayed on a graphical user interface or stored in an electronic record for subsequent review and longitudinal comparison.

[0077] In certain embodiments, the quantified concentrations or expression levels used as inputs to the computer-implemented method comprise circulating proteomic measurements or single-cell or bulk transcriptomic data corresponding to the disclosed biomarkers. Such input data may include high-throughput proteomic measurements obtained from aptamer-based assays, immunoassays, or mass-spectrometry-based workflows, as well as transcript abundance values generated through sequencing- or hybridization-based transcriptomic profiling. These proteomic or transcriptomic inputs provide quantitative molecular features facilitating computational prediction of atherosclerosis or progression thereof.

[0078] In certain embodiments, applying the multivariable model further comprises normalizing biomarker concentrations using a calibration curve tailored toatherosclerosis-associated biomarker ranges. Calibration curves may be generated using known analyte concentrations to convert raw assay signal into standardized, quantitative biomarkervalues. Incorporation of such normalization steps enhances measurement consistency across analytical runs and ensures that biomarker inputs fall within the dynamic ranges expected by the computational model.

[0079] In certain embodiments, the computing device comprises a processor configured to execute instructions stored on a non-transitory computer-readable medium to apply the multivariable model. The processor may receive the quantified biomarker concentrations or expression levels, apply the biomarker-specific coefficients or machine-learning parameters, and generate the resulting atherosclerosis risk score. Such embodiments encompass local execution environments including desktop computers, laboratory workstations, and dedicated analytics hardware.

[0080] In certain embodiments, the computing device comprises a cloud-based server configured to receive biomarker concentration or expression data transmitted from a remote assay device. The cloud-based system may execute the multivariable model using biomarker-specific coefficients or machine-learning parameters and return the computed atherosclerosis risk score to the originating device or a clinical interface, facilitating distributed or high-throughput computational analysis across multiple sites.

[0081] In certain embodiments, the multivariable model comprises a regression model that outputs a continuous atherosclerosis risk score based on the quantified biomarker concentrations or expression levels. Such regression models may employ linear, logistic, or other regression-based modeling frameworks to generate a numerical risk estimate reflecting the subject’s likelihood of having or developing atherosclerosis.

[0082] In certain embodiments, applying the multivariable model comprises executing a trained machine-learning algorithm, including gradient-boosting regression, random forest regression, or neural network regression. These machine-learning models may capture nonlinear or higher-order interactions among biomarkers and improve predictive performance relative to conventional statistical approaches by leveraging large, labeled datasets comprising biomarker measurements and atherosclerosis phenotypes.

[0083] In certain embodiments, the multivariable model is trained on atherosclerosis phenotypes, including coronary artery calcification (CAC) status, CAC extent, or incident CAC development within a cohort of subjects. Incorporating such training data allows the model to learn biomarker-specific patterns associated with early, subclinical, or progressive atherosclerotic disease.

[0084] In certain embodiments, applying the multivariable model further comprises integrating clinical covariates, including age, sex, and race, into the atherosclerosis risk score calculation. Integration or adjustment for such covariates enhances predictive accuracy by accounting for demographic factors known to influence atherosclerotic disease development and biomarker distributions.

[0085] In certain embodiments, generating the atherosclerosis risk score further comprises comparing the computed score to a threshold value determined using receiver operating characteristic (ROC) analysis. ROC-derived thresholds may be selected to optimize discriminative performance, including sensitivity, specificity, or the Youden index, facilitating consistent classification of subjects based on biomarker-driven disease likelihood.

[0086] In certain embodiments, outputting the atherosclerosis risk score comprises generating a graphical user interface that displays the computed risk score along with a classification status indicating whether the score exceeds the threshold for increased atherosclerosis risk. The GUI may present numerical values, categorical indicators, visual alerts, interpretive text, or longitudinal trends derived from stored biomarker data.

[0087] In certain embodiments, the computer-implemented method further comprises receiving biomarker concentration data transmitted from an analytical or immunoassay device and automatically applying the multivariable model to the received data. This automated data transfer and processing workflow reduces user intervention and supports seamless integration between laboratory instrumentation and computational risk-scoring infrastructure.

[0088] In certain embodiments, the computer-implemented method further comprises longitudinal analysis, including receiving biomarker concentrations or expression levels for the subject at two or more time points, computing a corresponding atherosclerosis risk score for each1time point, and determining disease progression based on changes in the computed scores. Such longitudinal tracking allows for assessment of therapeutic response, disease trajectory, or emerging subclinical risk.

[0089] While the terms used herein are believed to be well understood by those of ordinary skill in the art, certain definitions are set forth to facilitate explanation of the presently disclosed subject matter.

[0090] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as is commonly understood by one of skill in the art to which the invention(s) belong.

[0091] All patents, patent applications, published applications and publications, GenBank sequences, databases, websites and other published materials referred to throughout the entire disclosure herein, unless noted otherwise, are incorporated by reference in their entirety.

[0092] Where reference is made to a URL or other such identifier or address, it understood that such identifiers can change and particular information on the internet can come and go, but equivalent information can be found by searching the internet. Reference thereto evidences the availability and public dissemination of such information.

[0093] As used herein, the abbreviations for any protective groups, amino acids and other compounds, are, unless indicated otherwise, in accord with their common usage, recognized abbreviations, or the IUPAC-IUBMB Joint Commission on Biochemical Nomenclature (See, iubmb . qmul . ac.uk / ) .

[0094] Although any methods, devices, and materials similar or equivalent to those described herein can be used in the practice or testing of the presently disclosed subject matter, representative methods, devices, and materials are described herein.

[0095] In certain instances, nucleotides and polypeptides disclosed herein are included in publicly available databases, such as NCBI® Gene (also known as Entrez Gene), GENBANK® and UNIPROT®. Information including sequences and other information related to such nucleotides and polypeptides included in such publicly available databases are expressly incorporated by reference. Unless otherwise indicated or apparent the references to such publiclyavailable databases are references to the most recent version of the database as of the filing date of this Application.

[0096] Following long-standing patent law convention, the terms “a”, “an”, and “the” refer to “one or more” when used in this application, including the claims. Thus, for example, reference to “a protein” includes a plurality of such proteins, and so forth.

[0097] Unless otherwise indicated, all numbers expressing quantities of ingredients, properties such as reaction conditions, and so forth used in the specification and claims are to be understood as being modified in all instances by the term “about”. Accordingly, unless indicated to the contrary, the numerical parameters set forth in this specification and claims are approximations that can vary depending upon the desired properties sought to be obtained by the presently disclosed subject matter.

[0098] As used herein, the term “about,” when referring to a value or to an amount of mass, weight, time, volume, concentration or percentage is meant to encompass variations of in some embodiments ±20%, in some embodiments ±10%, in some embodiments ±5%, in some embodiments ±1%, in some embodiments ±0.5%, in some embodiments ±0.1%, in some embodiments ±0.01%, and in some embodiments ±0.001% from the specified amount, as such variations are appropriate to perform the disclosed method.

[0099] As used herein, ranges can be expressed as from “about” one particular value, and / or to “about” another particular value. It is also understood that there are a number of values disclosed herein, and that each value is also herein disclosed as “about” that particular value in addition to the value itself. For example, if the value “10” is disclosed, then “about 10” is also disclosed. It is also understood that each unit between two particular units are also disclosed. For example, if 10 and 15 are disclosed, then 11, 12, 13, and 14 are also disclosed.

[0100] As used herein, an “aptamer-based proteomic platform” refers to ahigh-throughput analytical system that uses nucleic-acid aptamers engineered to bind specific protein targets for quantitative measurement of circulating proteins. An example includes the SomaScan platform, which measures thousands of proteins using panels of aptamer reagents to generate standardized quantitative protein abundance data.

[0101] As used herein, “atherosclerosis” refers to a condition of the arterial vasculature characterized by molecular, cellular, and structural changes within the vascular wall. These changes can include endothelial dysfunction, local and systemic inflammation, immune activation, and metabolic-oxidative stress, which collectively contribute to the development of coronary artery disease. Atherosclerosis, as described in the present disclosure, encompasses phenotypes detectable at early or subclinical stages, exemplified by, but not limited to, coronary artery calcification (CAC), and reflects the underlying molecular architecture.

[0102] As used herein, the term “automated analytical device” refers to an instrument configured to perform biomarker quantification with minimal manual intervention, including but not limited to immunoassay platforms, automated liquid handling systems, or mass spectrometry-based proteomic analyzers.

[0103] As used herein, “assessing” refers to detecting, measuring, quantifying, or otherwise determining the presence, absence, or amount of a biomarker, including but not limited to obtaining circulating proteomic measurements, obtaining single-cell or bulk transcriptomic data, or deriving processed quantitative values from such data, including computationally derived quantitative values based on such measurements.

[0104] As used herein, “atherosclerosis risk score” refers to a numerical or categorical value indicative of atherosclerotic status or risk, calculated based on assessed concentrations or expression levels of a plurality of biomarkers and optionally weighted using predetermined coefficients or other model parameters derived from statistical or computational modeling. The score reflects the likelihood, presence, or progression of atherosclerosis in a subject and may be produced by linear or nonlinear combinations of biomarker values, regression-based estimators, or machine-learning models trained on datasets containing atherosclerosis-related phenotypes such as CAC, and may optionally incorporate clinical covariates (e.g., age, sex, race). The score may be validated using statistical performance metrics.”

[0105] “Biological sample,” as used herein, refers to any material obtained from a subject that contains cells, cellular components, nucleic acids, proteins, or other biomolecules suitable for assessing one or more of the disclosed protein or gene targets. The term includes, without limitation, whole blood, serum, plasma, circulating cell-free nucleic acids, isolated cells,tissue biopsies (including coronary artery tissue), and processed derivatives thereof, provided that the sample retains detectable levels of the analyte(s) of interest. As used herein, the term “plasma” refers to the liquid component of blood obtained after centrifugation of whole blood treated with an anticoagulant, and the term “serum” refers to the liquid component obtained after clotting and centrifugation of whole blood without anticoagulant.

[0106] As used herein, “biomarkers” refer to measurable molecular entities, such as proteins, gene transcripts, or processed quantitative derivatives thereof, that exhibit an association with atherosclerosis phenotypes, including coronary artery calcification, coronary tissue transcriptomic signatures, or circulating proteomic signatures identified through the multi-omic analyses described herein. Biomarkers include, but are not limited to, the proteomic and transcriptomic targets disclosed in the application, as well as any feature quantitatively derived from such targets (e.g., normalized protein abundance values, inferred gene expression levels, or model-generated composite variables). Biomarkers may further include genetic variants associated with atherosclerotic phenotypes, including variants identified through GWAS, TWAS, pQTL, or integrative multi-omic analyses.

[0107] As used herein, the term “biomarker-specific coefficients” refers to numerical weights assigned to individual biomarkers within a predictive model, derived from statistical or machine learning training on atherosclerosis phenotypes.

[0108] As used herein, the term “calibration curve” refers to a mathematical relationship generated by measuring known concentrations of an analyte to establish the correspondence between instrument signal and biomarker abundance. Calibration curves may be linear or nonlinear and are used to convert raw instrument output into quantitative values. In certain embodiments, calibration curves may be tailored to the expected concentration ranges of atherosclerosis-associated biomarkers.

[0109] As used herein, the term “clinical covariates” refers to subject-specific variables that provide additional context for disease risk assessment, including but not limited to age, sex, and race.

[0110] As used here, the term “clinically healthy individuals” refers to a group of subjects who lack clinical evidence of atherosclerosis and / or overt cardiovascular disease, as determined by standard clinical evaluation, such as the absence of coronary artery calcification (CAC) or other clinically manifest coronary disease. Such individuals serve as a reference population for establishing biomarker distributions and threshold values (e.g., a predetermined percentile of the risk-score distribution in a reference population without clinical atherosclerosis).

[0111] The present application can “comprise” (open ended) or “consist essentially of’ the components of the present invention as well as other ingredients or elements described herein. As used herein, “comprising” is open ended and means the elements recited, or their equivalent in structure or function, plus any other element or elements which are not recited. The terms “having” and “including” are also to be construed as open ended unless the context suggests otherwise.

[0112] As used herein, “computer-implemented method” refers to a process executed by one or more processors configured to perform the disclosed steps using machine-readable instructions stored on a non-transitory computer-readable medium.

[0113] As used herein, the term “computing device” refers to any hardware system capable of executing instructions stored on a computer-readable medium, including but not limited to servers, desktop computers, laptops, tablets, or cloud-based systems.

[0114] As used herein, the term “enzymatic digestion” refers to the cleavage of proteins into peptides using proteolytic enzymes.

[0115] As used herein, the term “executable code” refers to computer instructions stored on a non-transitory medium that can be executed by a processor to perform specific functions, such as atherosclerosis risk score calculation.

[0116] As used herein, the term “generating,” when used in connection with an atherosclerosis risk score, refers to calculating or producing the atherosclerosis risk score by applying a mathematical or algorithmic model to biomarker concentration data, optionally integrating clinical covariates.

[0117] As used herein, “graphical user interface” refers to a visual display environment that presents calculated results (e.g., atherosclerosis risk score) in a human-readable format and optionally provides interpretive categories, alerts, and recommendations.

[0118] As used herein, the term “identifying,” when used in connection with a subject, refers to classifying a subject as having atherosclerosis or an increased risk thereof by comparing the atherosclerosis risk score to a predetermined threshold value.

[0119] As used herein, the term “intervention” refers to any non-pharmacologic treatment strategy focusing on improving vascular health, including but not limited to diet modification (such as adopting the Mediterranean or DASH diets, which emphasize high fiber, lean proteins, and healthy fats while minimizing saturated fats, sugars, and sodium) and exercise programs (such as regular physical activity, including at least 150 minutes of moderate-intensity aerobic exercise per week). In relevant subject, intervention can include smoking cessation, weight management, reducing systemic inflammation, stress-reduction techniques, and blood pressure management.

[0120] As used herein, the term “machine learning algorithm” refers to a computational method that learns patterns from data to make predictions, including but not limited to gradient boosting, random forest, or neural network models.

[0121] As used herein, “multivariable model” refers to a mathematical or computational model that uses two or more independent variables (e.g., biomarker concentrations, clinical covariates) to generate a predictive output, such as a prediction, classification, or risk score.

[0122] As used herein, “non-transitory computer-readable medium” refers to a physical storage medium that retains data and executable instructions, such as hard drives, solid-state drives, optical discs, or flash memory, and explicitly excludes transitory signals. The physical storage medium can be configured to store instructions for generating an atherosclerosis risk score. As used herein, “optional” or “optionally” means that the subsequently described event or circumstance does or does not occur and that the description includes instances where said event or circumstance occurs and instances where it does not.

[0123] As used herein, “predetermined coefficients” refer to numerical weights, parameters, regression coefficients, machine-learning model parameters, or other quantitative values that have been derived by training a statistical or computational model on datasets comprising atherosclerosis-related phenotypes (e.g., coronary artery calcification, coronary transcriptomic profiles, circulating proteomic signatures). These coefficients may originate from linear models, generalized linear models, regularized models, machine-learning algorithms, or multi-omic integration frameworks described in the disclosure.

[0124] As used herein, the term “processing whole blood” refers to any procedure that separates plasma or serum from cellular components, including but not limited to centrifugation, filtration, or other fractionation techniques commonly used in clinical and research settings.

[0125] As used herein, the term “protein denaturation” refers to the disruption of secondary and tertiary protein structures to expose peptide bonds for enzymatic digestion, typically achieved using heat, chemical agents (e.g., urea, guanidine hydrochloride), or detergents.

[0126] As used herein, the term “quantifying concentrations” refers to measuring the amount of each biomarker in a biological sample using analytical techniques known in the art, including but not limited to mass spectrometry-based proteomics, immunoassays, or automated proteomic analysis systems. Such quantification may include sample preparation steps such as protein digestion, denaturation, and calibration.

[0127] As used herein, the term “receiver operating characteristic (ROC) analysis” refers to a performance evaluation method for a predictive model that plots the true positive rate (sensitivity) on the Y-axis against the false positive rate, which is calculated as 1 - specificity, on the X-axis across a range of decision thresholds. This analysis illustrates the trade-off between correctly identifying positive cases and incorrectly classifying negative cases and is commonly used to assess the discriminatory ability of diagnostic and predictive algorithms. The area under the ROC curve (AUC) provides a quantitative measure of overall model performance, with values closer to 1.0 indicating superior accuracy.

[0128] As used herein, the term “reference population” refers to a group of subjects used to establish baseline or comparative distributions of biomarker concentrations, atherosclerosis risk scores, or other quantitative measures. In some embodiments, the reference population comprises clinically healthy individuals who lack clinical evidence of atherosclerosis; in other embodiments, the reference population may be defined by demographic, clinical, orstudy-specific criteria relevant to deriving threshold values or percentiles.

[0129] As used herein, the term “regression analysis” refers to a statistical modeling technique that estimates relationships between dependent and independent variables, including linear regression, logistic regression, or other regression-based approaches commonly used in biomedical data analysis.

[0130] As used herein, “risk” refers to the probability or likelihood that a subject will develop atherosclerosis or experience progression of atherosclerosis within a defined time horizon, as estimated by comparing the subject’s atherosclerosis risk score to a reference distribution or by applying a predictive model trained on empirical outcome data. As will be appreciated by one of ordinary skill in the art, risk prediction does not imply certainty or guarantee of future outcomes; rather, it provides a probabilistic estimate based on populationlevel associations and statistical modeling. Such estimates inherently involve variability and uncertainty and are intended to inform clinical decision-making rather than serve as an absolute determinant of disease occurrence.

[0131] As used herein, “statistical or computational modeling analysis” refers to an analytical method or algorithm that processes biomarker data to identify patterns associated with atherosclerosis. Such analyses include, but are not limited to, regression modeling, Mendelian randomization, proteome-wide association studies (PWAS), transcriptome-wide association studies (TWAS), pQTL-integration methods, multi-omic data integration, machine-leaming-based prediction algorithms, or other approaches that relate biomarker variation to coronary artery disease phenotypes as described in the disclosure.

[0132] As used herein, the term “subject” refers to any mammalian individual for whom assessment of atherosclerosis or risk thereof is desired. In certain embodiments, the subject is a human. In other embodiments, the subject is a non-human mammal, including but not limited tocompanion animals (e.g., dogs, cats), livestock (e.g., horses, cattle), or research animals (e.g., rodents, primates). The term encompasses healthy individuals as well as those with existing or suspected atherosclerosis, including coronary artery calcification (CAC), or related conditions.

[0133] As used herein, the term “therapeutic agent” refers to any pharmacologic compound, biologic, or intervention administered to treat, manage, or reduce risk or progression of atherosclerosis.

[0134] As used herein, the term “threshold value” refers to a predetermined cutoff used to classify a subject as positive for atherosclerosis or at increased risk thereof, wherein the cutoff may be static or dynamically updated and may be optimized using statistical methods to achieve desired sensitivity and specificity.”

[0135] As used herein, the term “transcriptomic data” refers to measurements that quantify the abundance of RNA transcripts in a biological sample, including but not limited to mRNA, non-coding RNA, or processed transcript-level features. Transcriptomic data may be obtained using sequencing-based methods (e.g., RNA-sequencing, single-cell RNA-sequencing) or hybridization-based platforms (e.g., microarrays), and may include bulk transcriptomic profiles, single-cell profiles, or computationally derived gene-expression estimates.

[0136] The presently disclosed subject matter is further illustrated by the following specific but non-limiting examples. The following examples may include compilations of data that are representative of data gathered at various times during the course of development and experimentation related to the present invention.EXAMPLES

[0137] Example 1 - Clinical cohorts

[0138] CARDIA: A total of 2,971 participants with circulating proteomics and CAC scores by computed tomography at the Year 25 visit were included74. Clinical-demographic data was collected by standardized assessment as described6575, and computed tomography for CAC was performed as previously noted via standardized protocols76,77. In addition, 793 participants had additional computed tomographic CAC data collected at Year 35 in CARDIA for analysis ofincident CAC development by Year 35 (in those individuals without CAC at Year 25). Study participants provided written informed consent, and the study was approved by the respective clinical site Institutional Review Board.

[0139] FHS: 573 individuals were studied with extant proteomics and computed tomographic CAC measurements from the FHS Generation 2 (“Offspring”) cohort to assess for replication of the findings in CARDIA78. Of note, CAC measures and proteomics were separated by median of 17.4 years in FHS. Methods for clinical and calcification traits have been reported79,80. All FHS participants gave written informed consent, and all study protocols received approval from the Institutional Review Board at Boston University Medical Center.

[0140] Example 2 - Quantification of the circulating proteome

[0141] In CARDIA, the SomaScan platform (aptamer-based technology) was used to quantify 7,228 aptamers in CARDIA (Somalogic, Boulder, CO), as reported in previous work74.71 aptamers with a coefficient of variation exceeding 20% were excluded. After log2 transformation and standardization of aptamer levels, aptamer values over 5 standard deviations from the mean value were winsorized. In FHS, a total of 1,128 Somascan aptamers was utilized, and methods for handling proteomic data (batch pooling, transformations, and rank normalization) have been reported81, and they have been reproduced with minimal change to maximize reproducibility. In brief, due to variations in collection batch in FHS, protein levels were standardized within each of the two batches, pooled, and subjected to rank normalization across all samples. Subsequent residualization against the assay plate was performed before analysis to eliminate plate-based effects81. Matching of CARDIA to FHS aptamers was performed by SomaScan seqID (1035 overlapping aptamers).

[0142] Example 3 - Human genetic studies

[0143] pQTL and GWAS specification: For the target proteins, protein quantitative trait loci (pQTLs, defined here as genetic instruments associated with circulating protein levels at genome-wide significance) were identified that were measured with 4,907 aptamers in 35,559 Icelanders (ecode.com / summarydata / ; quality control reported previously82). pQTLs exceeding a genomic threshold (P<5xl0-8) for the target proteins were selected, leveraging both c / .s-pQTLs(+2 Mbp on both directions of the encoding locus) and / ra / .s-pQTL (when available). This subset was subject to linkage disequilibrium (LD) pruning (T?2<0.001, lOkbp window, 1000 Genome European reference). pQTLs were restricted to those exceeding an F statistic over 10 to minimize downstream weak-instrument bias. Of 131 unique proteins that were associated with presence and extent of CAC in CARDIA, at least one pQTL for 100 proteins was identified.

[0144] For disease states, three of the largest GWASs capturing CAC-relevant clinical disease states was selected for assessment in Mendelian randomization: (1) coronary calcification (N=36,720 Europeans)36; (2) coronary artery disease (CAD; N=l, 165,690 Europeans)83; (3) myocardial infarction (N=639,221 Europeans)84. A summary of studies is presented in Table 1.

[0145] Example 4 - Two-sample Mendelian randomization (MR)

[0146] The two-sample MR framework tested whether proteins implicated in the initial epidemiologic and tissue evaluation were causal by human genetic approaches to CAC and relevant outcomes. Protein quantitative trait loci (pQTLs), serving as instrumental variables (IVs) for genetically determined protein levels, were utilized as exposures, and the three CAC-relevant clinical disease states were considered as outcomes.

[0147] Quality control was performed using TwoSampleMR R package2. Potential causal associations were primarily tested with inverse variance weighted robust-penalized method, using MendelianRandomization R package3. ‘Default effect’ model was used for association testing, or ‘random effect’ model when IVs for the protein demonstrated significant heterogeneity in effect size (Q-het P<0.05).

[0148] Given the multiple lines of evidence preceding these approaches, the protein was considered as potentially causally associated with the CAC-relevant outcome at a nominal P <0.05. In addition, median penalized (to limit IV-related bias) and Egger-penalized robust methods (for pleiotropy) were used as sensitivity analyses to enhance the confidence in causal interpretations, consistent with STROBE-MR guideline for causal inferencing using observational data85.

[0149] Example 5 - Single-cell RNA sequencing of human coronary arteries

[0150] Single-cell RNA sequencing (scRNA-seq) data derived from 13 human coronary arteries (8 with atherosclerotic lesions and the remainder lesion-free controls) involving 56,183 cells32were leveraged. The data were analyzed using the Seurat R package86. Doublet removal was performed using scDblFinder87. As ambient RNA contamination can negatively affect gene expression profiling, correction was performed via DecontX using default parameters88. Raw count normalization was applied through a regularized negative binomial regression approach as implemented in SCTransform89. To avoid confounding by cell cycle state, adjustments were made for cell cycle variance. For integration of processed sequencing libraries across potential technical differences and dataset-specific conditions, Harmony90was used, k = 20 dimensionality reduction was applied using principal component analysis (PCA) on the normalized counts, leveraging the first 30 principal components (PCs) for clustering. The functions RunUMAP (using 30 neighboring points for local approximation of manifold structure) and FindNeighbors (using for the ^-nearest neighbor algorithm) were invoked in Seurat, setting reduction to ‘harmony’ to use the Harmony embeddings. Cell type annotation leveraged transfer learning from the Tabula Sapiens (specifically vasculature data in 42,650 cells available at cellxgene.cziscience.com / e / a2d4d33e-4c62-4361-b80a-9be53d2e50e8.cxg / )91. For differential expression analysis between the samples with atherosclerotic lesions and the lesion-free samples, pseudobulking-based DESeq292was used.

[0151] Example 6 - Proteome-wide association studies

[0152] Proteome-wide association studies (PWAS) of CAC were performed using genetic models of circulating proteome developed in a large-scale proteo-genomic study (N = 7,213 European ancestry)93. The models had been trained using the PrediXcan methodologyapplied to 4,657 aptamers (Somalogic) representing 4,435 unique genes. Only those proteins that attain nominal significance during model training (P<0.05, cross-validation r=0.10, 1,340 unique proteins) were included. This resource does not overlap with the UK Biobank (N = 361,194), allowing us to estimate the association between the genetically determined circulating protein expression and related traits in an independent phenomic resource. GWAS summary statistics from a linear model of a phenotype as a function of the first 20 genotype-based PCs, sex, age, age2, sex*age, sex*age2, and the protein under test were used to estimate the effect size of the protein on the two phenotypes in the UK Biobank: coronary atherosclerosis (code I9 CORATHER) and myocardial infarction (code I9_MI). Of 753 proteins associated with presence and extent of CAC in CARDIA (derivation P<0.05), 316 were tested in PWAS.

[0153] Example 7 - Transcriptome-wide association studies

[0154] Gene expression models in human coronary artery samples from 268 unrelated individuals were developed. JTI TWAS methodology was used, which resulted in a substantial increase in the number of imputable genes («iGenes= 9,918) relative to the PrediXcan methodology («iGenes= 5,050). Prediction models of normalized gene expression were trained using sex, platform, and 5 genotype-based principal components as covariates. The features consisted of common SNPs (minor allele frequency > 5%) in the gene’s cA-region whose extent, a model hyperparameter, was determined using cross validation. The tissue-specific JTI association test controlled the type I error rate: at the significance level of 5%, the type I error rate was 4.96% (compared to 5.04% for PrediXcan and 5.17% for UTMOST). Phenome-wide association studies (PheWAS) were performed for a set of prioritized genes by identifying the associations of the genes with the heritable phenome (heritability A2P < 0.05) as captured in the UK Biobank.

[0155] Example 8 - Causal inference on TWAS associations

[0156] To provide additional support for the role of the implicated genes in CAC, causal inference was performed via MR with gene expression as “exposure” and CAC as “outcome”. For inference on causality, five MR approaches were used, each relying on strong assumptions on the underlying ground truth: MR-Egger94, median-based95, maximum likelihood96, inversevariance weighted97, and MR-JTI98.

[0157] Example 9 - Functional genomics

[0158] A broad collection of 18 assays was used for the identification of human coronary artery regulatory elements. These assays measure chromatin marks: H3K27ac, H3K27me3, H3K4mel, H3K4me2, H3K4me3, H3K36me3, H3K79me2, H3K9ac, H3K9me3, and H4K20mel; chromatin accessibility, specifically DNase-Seq and ATAC-Seq; and transcription factor binding (including cohesin complex subunits RAD21 and SMC3 and RNA polymerase II subunit POL2RA). Leveraging these epigenomic, chromatin accessibility, and transcription factor assays within a regulatory annotation framework" allowed us to classify the regulatory elements (e.g., active enhancers, distal or proximal insulators, active promoters) in human coronary artery. Tissue-specific Hi-C data were leveraged to identify the target gene of a regulatory element in coronary artery.

[0159] Example 10 - Associations of the circulating proteome to CAC

[0160] In CARDIA, protein aptamers were log transformed, centered, standardized to unit variance, and winsorized to 5 standard deviations for regression. Given absence of known datasets with harmonized proteomic platform and concurrent measures of CAC, the CARDIA sample was randomly split into a derivation (70%) and validation subsample (30%). Models were estimated in the discovery subsample for presence / absence of CAC (as a binary variable in logistic models) and its extent (continuous in linear models; modeled as ln[CAC+l] to account for zero CAC values). While ln(CAC+l) transformation does not fully normalize the distribution, it is a commonly used approach in prior studies to accommodate the skewed nature of CAC data for statistical modeling36,37’100. Each model contained one aptamer at a time, adjusted for age, sex, and race, with a 5% false discovery rate (Benj ami ni -Hochberg FDR) applied across models to control for multiplicity. Aptamers that were significant at a 5% FDR threshold were passed to identical models in the validation subsample. Aptamers that were significant at 5% FDR for both presence of CAC and its extent in both derivation and validation subsamples, were analyzed in similar logistic regression models (adjusted for age, sex, and race) for incident CAC among subjects who had CAC scores of 0 at Year 25 and available computed tomography data at Year 35. Aptamers in the incident CAC models were considered significant using a threshold of 5% FDR.

[0161] In order to explore generalizability of the findings across geographies and populations, population-level proteomics in FHS were studied. In FHS, protein profiling was completed in two batches. In batch one, 1129 aptamers were profiled in 821 individuals, and batch 2 included an expanded panel of 1372 aptamers, which was assayed in 1092 participants. Aptamer levels of 1128 aptamers common to both batches were log-transformed, standardized within batch, then pooled and normalized to a mean value of 0 and SD unit of 1 using the Blom rank-based inverse-normal method. These values were then regressed on assay plate ID to account for plate effects, and the standardized residuals were used for regression. Replication studies in FHS were performed in a similar fashion (logistic for presence or absence of CAC; continuous linear for extent as standardized ln[CAC+l]), with each protein in separate models adjusted for age, sex, and race (white vs. non-white) (with 5% FDR used to control multiplicity). Concordance of CARDIA and FHS effect size were assessed via Pearson correlation.

[0162] Example 11 - Data and materials availability

[0163] Data from CARDIA used in these analyses are available through the Coronary Artery Risk Development in Young Adults study (CARDIA; cardia.dopm.uab.edu) or at dbGaP (for proteomics, dbGaP identifier phs003491.vl.pl). Data from the Framingham Heart Study (FHS) is available at dbGaP (dbGaP identifier pht006013) or via contact with the FHS coordinating center (www.firaminghamheartstudy.org). UK Biobank data (accessed under application number 94960) is available at UK Bio bank Research Access Portal. Code used for analysis in this study are available on Zenodo (doi.org / 10.5281 / zenodo.3842289) and on the Github repository for this project (github.com / gamazonlab / CACOmics).

[0164] Example 12 - Characteristics of proteomic discovery cohorts

[0165] The overall study scheme is shown in FIG. 1. Clinical characteristics of the study populations (Coronary Artery Risk Development in Young Adults; CARDIA; N=2,971;Framingham Heart Study; FHS; N=573) are in Table 2A-2B. CARDIA participants at Year 25 had a mean age of 50.2±3.6 years, with approximately even distribution by sex and selfidentified race (56% women, 46% Black) and prevalent cardiometabolic risk (mean systolic blood pressure 119±15 mmHg, 8% diabetes, mean BMI 30.3±7.1 kg / m2) and CAC (-29%). A subgroup of participants with baseline proteomics and without prevalent CAC at Year 25 hadcomputed tomography performed ~10 years later (Year 35 study visit in CARDIA; N=793), with an incident CAC rate of -31%. FHS participants were similar in age and sex, with a low prevalence of clinical cardiovascular disease (~4%) at the time of proteomics, with a high rate of incident CAC (assessed ~20 years after proteomics, -69%).>>>>>

[0166] Example 13 - Identifying and characterizing the circulating proteomic architecture of coronary artery calcification

[0167] 136 aptamers (representing 131 unique proteins) associated with presence and extent of CAC were identified across discovery and validation sets within CARDIA (FIG.2A; presence of CAC, 186 aptamers; extent of CAC, 201 aptamers; regression summaries not shown). Proteins with highest effect sizes in regression exhibited strong biological plausibility across known mechanisms of vascular remodeling (top 20 aptamers for CAC extent in validation set shown in Table 3), including fibrosis and inflammatory mechanisms (GDF-1510, CDCP111, GSN12, TSP-213, chemokines), oxidative lipid metabolism (CILP214), extracellular matrix remodeling and signaling (MMP-7, MMP-1215, TIMP-1, integrins16), calcification (Notch I17, ARHGAP3618), and general metabolism (GIP19). Importantly, a majority of identified protein associations (though biologically plausible) had not been previously widely reported in human coronary disease or calcification, including mechanisms of protein catabolism (UBE2G220), signal transduction (RHG36), macrophage efferocytosis and lipid metabolism (TREM227), endothelial cell states (EGFR22), and extracellular matrix metabolism (PCOC1).

[0168] To address epidemiologic limitations on cross-sectional associations (reverse causation), the relation of the prevalent CAC proteome was quantified with incident development of CAC 10 years later in a subsample of CARDIA without CAC at time of proteomics (N = 793; FIG. 2B-2C, Table 4A-4B). Among aptamers related to prevalence and extent of CAC, strong consistency in directionality of effect (129 / 136 aptamers) with incident CAC was observed, including 59 aptamers that were associated with incident CAC at 10 years. In addition, proteomic associations with CAC measured 20 years after proteomics in FHS were examined, with overall consistency in effect directionality and size (for extent of CAC: Pearson r = 0.47, p<0.001) being observed, with 649 of 1,035 aptamers exhibiting concordant directionality of effect (FIG. 3A-3B; full regression results not shown).>>>>>>

[0169] Example 14 - Human genetic prioritization of mediators of cardiovascular disease in the human CAC proteome

[0170] To increase confidence in the targetability of select mediators and their causal relevance to human CVD, a potential causal role for proteins implicated by CAC was assessed through two complementary approaches: (1) Mendelian randomization (MR) and (2) proteome-wide association studies (PWAS). For MR, established protein quantitative trait loci (pQTL) were used for proteins associated with presence and extent of CAC in CARDIA (100 includedthat satisfied quality control thresholds; see Example 4). Broad consistency was observed between MR-based genetic estimates of effect on three major disease conditions (CAC, coronary artery disease, and myocardial infarction; spanning ~1.8 million individuals; study characteristics shown in Table 1) and proteomic-phenotype associations from CARDIA (FIG.4A, full results including detailed results of 2-sample Mendelian Randomization analyses not shown), several of which have not been widely reported in coronary vascular calcification in humans. Proteins with both proteomic and genetic evidence in favor of increased calcification and clinical risk included mediators involved in central mechanisms of metabolism and inflammation, including GABARAPL123(autophagy), SI00A9 calgranulin B (inflammatory signaling / atherosclerosis in murine models24), VEGFA25226(angiogenesis and lipid handling), C1QTNF122(macrophage activation). Importantly, several genes with evidence of protection against endpoints at a proteomic and genomic level had not been broadly implicated in calcification in humans, including proteins implicated in central nervous system development and physiology (SLITRK12^,2<>: noradrenergic system development; IGLONF'2, NTRK3 extracellular matrix metabolism (HS6STG3, DAM23) and fibroblast development (FGFRT).

[0171] In a parallel approach, using genetic data from 7,213 individuals with proteomics and genomic data, a pipeline was developed for PWAS of circulating protein abundance across 4,657 aptamers (Somalogic; see Example 6). The association of genetically determined protein abundance with atherosclerotic disease phenotypes in a cohort of 361,194 UK Biobank participants was estimated. At a 5% Bonferroni threshold, four proteins were observed with significant evidence for involvement of genetically determined protein abundance in atherosclerosis (PCSK9, APOCI, HTRA1, and FGFR1, FIG. 4B) and two for myocardial infarction (PCSK9, INHBC, FIG. 4C; full results not shown) among those proteins that had suitable PWAS models (i.e., P<0.05, cross-validation r=0.10, 1,340 unique proteins) and evidence for association with CAC in CARDIA proteomics. Notably, the directly measured, CAC-associated proteins were substantially enriched for genetically determined CAC associations (FIG.4D). Circulating PCSK9 expression is linked to greater severity and vulnerability of coronary plaque31, with successful PCSK9 interruption linked to lower vascular event rates32. Both CXCL12 (an inflammatory chemokine) an APOCI (lipoprotein metabolism) have been implicated in human GWAS or functional studies of atherosclerosis33,34. INHBC(implicated in adipocyte metabolism) appears to link cardiometabolic health to coronary risk, with higher levels related to higher coronary risk35.

[0172] Example 15 - Coronary artery-specific transcriptome-wide association studies (TWAS) and single cell transcription implicates both known and novel targets in human atherosclerosis

[0173] Circulating proteome-wide association — either via genetic instruments (PWAS) or directly measured proteins (CARDIA) — cannot resolve target tissue-specific causal mechanisms of disease pathogenesis. TWAS in a mechanistically relevant tissue may address this limitation, estimating genetically determined transcript expression within a trait-related tissue (in this case, human coronary artery) to use in conjunction with cognate phenotype GWAS data (in this case, in vivo human measured calcified coronary plaque or CAC). TWAS in such mechanistically relevant tissues enhances specificity of identified GWAS associations.Furthermore, GWAS variant associations are frequently non-coding and can be within genomically “dense” loci or gene “deserts,” limiting precise inference on which SNP-linked gene is functional.

[0174] Therefore, genetic models were constructed for gene expression in 268 human coronary arteries (see Example 7), mapping these models to the largest extant GWAS of CAC36to identify gene-level associations with CAC. Coronary artery expression of genes associated with CAC in TWAS (FIG.5A; full results not shown) included known (PHACTR131, MRAS^, MORF4L1-ADAMTS738and not previously widely implicated targets GIGYF1, DMPK, RPL9). PHACTR1 (phosphatase and actin regulator-1) and MRAS demonstrated strong association with CAC, consistent with pathogenic roles in endothelial dysfunction, CAC, or inflammatory cell function37,39’40. Interestingly, M0RF4LI and ADAMTS7 are adjacent at 15q25.1 with the lead SNP in an intronic region of MORF4L13?', leading to difficulty in resolving the actual causal disease gene. While only ADAMTS7 has wide support in models of atherosclerosis38, the TWAS results (based on genetically determined RNA expression in coronary artery) along with MR to account for LD contamination implicated both genes in CAC pathogenesis. Other genes identified in coronary TWAS of CAC displayed phenotypic association with anatomically or functionally adjacent processes, though not directly with atherosclerosis itself GIGYFF. type 2diabetes41; DMPK'. cardiac conduction42; RPL9'. ribosomal metabolism43). Importantly, despite differences in origin, among the genes encoding proteins significantly associated with CAC in CARDIA, an enrichment was observed for coronary artery TWAS associations with CAC (FIG.5B), including both known mediators of vascular dysfunction or plaque homeostasis (NOTCH344, TNFSF1245, S100A1246') as well as several not previously widely related to vascular disease.

[0175] Overall, the approach prioritized 8 genes with multi-level support across proteomic association with CAC and human coronary artery TWAS of CAC. In comprehensive causal analyses of CAC using 5 different MR approaches (each with its own set of strong assumptions; Example 8), all 8 genes (N0TCH3, SPINK2, S100A12, RPP25, OAF, HS6ST3, TNFSF12, GPC6) had supportive evidence for causal associations (FIG. 5C). For 5 of these genes (SPINK2, RPP25, OAF, HS6ST3, GPC6), the direction of effect observed in MR analyses was consistent with the direction of the circulating proteomic association with CAC. Of note, none of the 8 genes have been therapeutically targeted in CAD, and less than half have been implicated in CVD directly (as noted above for N0TCH3, TNFSF12, S100A12')44'46. Genetic variants in HS6ST3 — a gene whose product has been implicated in protein-extracellular matrix interactions — have been linked to CAC in some settings47, though implications in human atherosclerosis have not been investigated. In addition, while not previously directly implicated in CVD, both GPC6 and RPP25 have support: GPC6 blocks canonical Wnt signaling48, a pathway central to vascular calcification49-50, and RPP25 (a part of ribonuclease P) has been implicated in genetic studies of cardiometabolic disease51and in scleroderma (a disease marked by multi-organ calcification and fibrosis). In phenome-wide association studies (PheWAS) across the heritable phenome (heritability A2P < 0.05) in UK Biobank using genetically determined expression in coronary artery, many of the 8 prioritized genes implicated vascular, inflammatory, and obesity-related phenotypes (FIG. 5D). The 8 genes were expressed across broad cell types in human coronary arteries52implicated in mechanisms of progressive calcification, including vascular inflammation, fibrosis, and smooth muscle cell remodeling (FIG. 5E). Collectively, these results demonstrate genetic support for proteomic targets from circulation and implicate a target gene expression within human coronary arteries in CAC pathogenesis.

[0176] Example 16 - Coronary artery-specific functional genomics identifies genome-wide significant trans regulatory elements for targets with multi-level evidence in human CAC

[0177] Extensive characterization of the nature of the genetic regulation of these 8 genes was conducted to gain crucial insights into disease mechanisms, with critical methodological relevance for the functional interpretation of GWAS findings, using a comprehensive array of epigenomic, chromatin accessibility and transcription factor binding assays in human coronary artery. A total of 11,795 SNPs interrogated in the CAC GWAS were identified to overlap a coronary artery regulatory element in contact (Hi-C) with one of the 8 genes.

[0178] Two loci (rs9515203 and rs28610385) attained genome-wide significance (P<5xl0‘8; FIG.6A). The SNP rs9515203 lies in an intron of COL4A2 and falls in an insulator — characterized by enrichment of CTCF and cohesin complex subunits RAD21 and SMC3 — in 3D contact with the distal target GPC6, a gene over 10 MB upstream on chromosome 13 (FIG.6B-6C). Similarly, the SNP rs28610385 lies in an intron of ADAMTS7 and falls in a transcribed gene body element — characterized by accessible chromatin (as measured by ATAC-seq) as well as enrichment of histone marks H3K36me3, H3K79me2, and H4K20mel — in 3D contact with the distal target gene RPP25 (FIG. 6D-6E).

[0179] Using the single-cell transcriptomics data for additional support of the gene regulatory relationships for the GWAS-implicated CAC -associated variants, GPC6 (glypican-6) was significantly differentially expressed in atherosclerosis in fibroblasts (P=4.87xl0‘25), pericytes (P=7.63xl0‘7), and smooth muscle cells (P=6.29xl0‘5). RPP25 was differentially expressed in atherosclerosis in fibroblasts (P=0.004) and endothelial cells (P=0.01; FIG. 7A-7C). Collectively, these results highlight the role of trans (distal) genetic regulation, by GWAS-implicated variants, of tissue-specific gene expression in disease pathogenesis.

[0180] Example 17 - Discussion of Examples 1-16

[0181] Coronary atherosclerosis has a rich history of epidemiologic, mechanistic, histopathologic and therapeutic study, spanning population-based risk factor studies, models of atherogenesis53, coronary structure-functional studies54, imaging55,56, and population-levelgenetic risk loci and phenotype characterization57'61. These studies have informed the current treatment and prevention of coronary disease with cardiometabolic-inflammatory contributors, leading to metabolic (e.g., lipids, glucose, select lipoproteins) and inflammatory (e.g., statins, IL-1b antagonists) targets to intercept pathogenesis and prevent clinical events and premature cardiovascular death. Nevertheless, off-target effects (e.g., sepsis with IL-lb antagonists) and high residual clinical risk despite therapy suggest precise identification of cellular mechanisms responsible for early coronary pathogenesis may further refine novel therapeutic development. Unfortunately, translational studies in humans at a molecular level have largely been limited to quantifying circulating molecules or genetic liability for atherosclerotic CVD, missing tissuespecific contributions secondary to the inability to obtain in vivo samples. Approaches to phenotyping molecular states in human coronaries are rare52’62'64, likely missing opportunities to prioritize important, potentially functional mechanisms of coronary atherosclerosis for intervention and tracking.

[0182] In these studies, population-wide circulating proteomics, non-invasive coronary imaging, and coronary artery tissue-specific functional genomics and transcription were integrated to prioritize targets for CVD susceptibility.

[0183] The studies within a large population sample (CARDIA) indexed a broad human proteome to prevalent and incident CAC — an independent marker of subclinical disease and incident clinical CAD and death prior to age 60 years65— implicating pathogenic mechanisms across oxidative stress and lipid metabolism, extracellular matrix remodeling, immune cell function, coagulation, and inflammation7,66’67. Several proteins displayed evidence of causal relevance to CVD outcomes and phenotypes across multiple genetic approaches (MR and PWAS).

[0184] Based on the rationale that coronary artery-specific contextualization may focus discovery for CVD, the largest human coronary artery TWAS of CAC was subsequently performed. In this study, bulk transcriptomics from 268 human coronary arteries were mapped to the largest CAC GWAS, and enrichment was demonstrated for genes implicated by proteomic associations across similar pathways of metabolic-inflammatory activation expressed in key cell types relevant to coronary disease progression.

[0185] Finally, coronary artery- specific regulation was used in MR for targets with shared evidence across proteomics and coronary TWAS, yielding a final set of 8 genes. Of note, several genes had prior biological annotation in CVD-relevant pathways, but none had been previously targeted in CVD. Phenome-wide association studies and single cell transcriptomics in human coronary arteries with and without atherosclerosis provided supportive tissue and phenotypic context. Collectively, these results not only provide the largest scientific resource to date of population-level multi-omics, coronary TWAS and coronary functional genomics-based discovery: more generally, they furnish a framework to decipher disease-relevant targets through integration of human genetic approaches with multi-omics.

[0186] Over the last decade, two parallel innovations in human genetic research have accelerated progress beyond traditional GWAS approaches: (1) broad molecular characterization of disease states (“multi-omics”) and (2) functional and tissue-based genetic approaches (TWAS, functional genomics). These approaches have leveraged the power of human genetics to identify novel disease targets through surrogate phenotype-proximal molecular traits (e.g., protein or metabolite QTLs) and to colocalize disease GWAS findings with tissue expression QTLs.Despite successes in metabolic and infectious disease, application of these approaches to human CVD has been limited. The current study comes in the context of an emergence of approaches to integrate tissue and biofluid domains to foster discovery in precision cardiovascular medicine68. Modem multi-omics linking genetics, transcriptomics, and proteomics have led to discovery in cardiovascular science36,68,69, including potential for druggability.

[0187] The approach was predicated on recent studies linking the human proteome to tissue-specific transcriptional states to implicate clinical-functional biomarkers of coronary disease68,69. In addition to well-known mechanisms of inflammation and fibrosis (Table 3), proteomics identified directionally consistent targets with strong evidence in model systems (less in humans), including TREM2 (attenuates macrophage uptake of oxidized lipids, limiting progression of atherosclerosis in mice21), Notchl (loss potentiates WA2-mediated calcification17), and ARHGAP36 (overexpression associated with connective tissue to bone transformation18), among others (Table 3). Importantly, a strong concordance was observed between effect estimates for prevalent and 10-year incident CAC in CARDIA with strongest agreement for both known (e.g., MMP-12, GDF-15) and emerging mediators of vascular risk.These results were further resolved with two complementary genomic approaches, standard MR and PWAS. Standard MR approaches rely on identification of cv.s-pQTLs for causal inference, while PWAS generates optimally predictive genetic models of protein abundance not necessarily reliant on specific pQTL identification. Standard MR approaches suggested directional consistency in genomic-proteomic effects for many targets (FIG. 4A), many of which have not been extensively implicated in human coronary disease by GWAS studies but have relevant metabolic-vascular mechanisms (e.g., HS6ST3™, PTPRD1, CTQT1, DNJB912'). PWAS in UK Biobank pinpointed four additional proteins with strong evidence for mechanistic involvement in atherosclerosis (PCSK931,32, FGFR1, HTRA1, APOC13,34) and myocardial infarction (PCSK9, INHBC5).

[0188] A key innovation in the approach is tissue-specific discovery via transcriptomewide association studies in human coronary artery tissue. The TWAS approach has recently been used to identify targets essential to cardiometabolic therapeutics73, though its application to coronary arteries has not previously been performed. The current report represents the largest TWAS in human coronary arteries to date. Human coronary TWAS associations with CAC included known (PHACTR11, MRAS^, MORF4L1-ADAMTS7:’8) and not widely implicated targets (GIGYF1, DMPK, RPL9). Enrichment was observed, among CAC-associated circulating proteins, for coronary artery TWAS, with targets implicated in vascular homeostasis (NOTCH 344, TNFSF1245, S100A 124&) as well as several targets not previously linked to coronary disease. Through extensive MR analyses for robust causal inference, all 8 overlapping targets implicated by both TWAS and circulating proteomics had supporting causal relation to CAC or myocardial infarction. PheWAS of these genes implicated vascular, inflammatory, and obesity-related phenotypes. Importantly, none of these genes had been therapeutically targeted in CVD, with a handful implicated in CVD-adjacent mechanisms (as described above44’46).

[0189] Using an epigenomic atlas of regulatory elements in coronary artery and chromatin conformation capture (Hi-C) experimental data that characterize the spatial organization of chromatin, substantial support was found for the role of trans (distal) genetic regulation, by GWAS-implicated CAC variants, of 2 of the 8 genes. Notably, confirmation of the gene regulatory relationship was found by showing that the gene targets of the GWAS-implicated CAC variants, GPC6 and RPP25, were significantly differentially expressed insingle-cell transcriptomic analysis of human atherosclerosis. Neither of these have been previously implicated in CAC, though both are implicated in CVD-relevant mechanisms: GPC6 blocks canonical Wnt signaling48(critical to vascular calcification49-50), and RPP25 is implicated in cardiometabolic disease51and in scleroderma (a disease marked by multi-organ calcification). These results establish a comprehensive paradigm beyond traditional GWAS, leveraging the power of functional genomics, tissue-specific transcription, and population-level proteomic studies to hone disease targets.

[0190] There are several considerations associated with the studies. While CAC is a marker of early CVD, studies across even earlier disease states reflecting the earliest changes in endothelial function (e.g., measures coronary flow and structure55) may uncover additional mechanisms of coronary physiology. In addition, coronary artery utilized for functional genomic approaches in the work may necessarily not reflect the range of coronary phenotypes relevant to disease discovery. Nevertheless, concordance with signals from the circulating proteome and CAC GWAS support external validity, generalizability, and broad biological significance of the findings. Moreover, availability of coronary arteries early in the development of CVD (with appropriate preservation for molecular studies) remains challenging. Genetic instruments of the circulating proteome were leveraged through PWAS, and human coronary specific PWAS is likely to yield additional mechanistic insights. Finally, discovery across heterogeneous populations may afford further generalizability. Nevertheless, multi-level genetic concordance in tissues and populations provide a focus for calcification mechanisms and biology for wider studies and therapeutic approaches.

[0191] In conclusion, an integrated approach was presented to address coronary artery disease susceptibility by examining the circulating proteome, coronary artery-specific transcription (including data from 268 human coronary arteries and single-cell analyses), and coronary artery-specific functional genomics. This strategy implicated genes not previously reported in causal CVD pathogenesis. This tissue-specific genetic approach has not previously been leveraged for human genetic discovery. Targets identified by the approach (N0TCH3, SPINK2, S100A12, RPP25, OAF, HS6ST3, TNFSF12, GPC6) enjoyed multi-level support, including via single cell transcription from atherosclerotic human coronary arteries, phenome-wide association, and implication in biologically relevant, CVD-adjacent mechanisms.Strikingly, the approach allowed “de-orphanization” of two loci significant in CAC GWAS without linkage to a cis target (GPC6, RPP25), demonstrating the power of the multi-omic approach in discovery. These results support a broader utilization of this approach as an adaptable framework applicable to all organ systems to parse precision targets for prevention, surveillance, and therapy of cardiovascular disease.

[0192] All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference, including the references set forth in the following list:REFERENCES1 Martin, S. S. et al. 2024 Heart Disease and Stroke Statistics: A Report of US and Global Data From the American Heart Association. Circulation 149, e347-e913 (2024). doi .org / 10.1161 / CIR.00000000000012092 Bos, D. et al. Comparison of Atherosclerotic Calcification in Major Vessel Beds on the Risk of All-Cause and Cause-Specific Mortality: The Rotterdam Study. Circ Cardiovasc Imaging 8 (2015). doi.org / 10.1161 / circimaging.115.0038433 Ritchey, M. D., Wall, H. K., George, M. G. & Wright, J. S. US trends in premature heart disease mortality over the past 50 years: Where do we go from here? Trends Cardiovasc Med 30, 364-374 (2020). doi.org / 10.1016 / j.tcm.2019.09.0054 Sniderman, A. D., Thanassoulis, G., Williams, K. & Pencina, M. Risk of Premature Cardiovascular Disease vs the Number of Premature Cardiovascular Events. JAMA cardiology 1, 492-494 (2016). doi.org / 10.1001 / jamacardio.2016.09915 Arnett, D. K. et al. 2019 ACC / AHA Guideline on the Primary Prevention of Cardiovascular Disease: A Report of the American College of Cardiology / American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 74, el77-e232 (2019). doi.org / 10.1016 / j j acc.2019.03.0106 Hashmi, S., Shah, P. W ., Aherrahrou, Z., Aikawa, E. & Aherrahrou, R. Beyond the Basics: Unraveling the Complexity of Coronary Artery Calcification. Cells 12 (2023). doi.org / 10.3390 / cellsl2242822Royer, P. et al. Plasma proteomics for prediction of subclinical coronary artery calcifications in primary prevention. Am Heart .7271, 55-67 (2024). doi.org / 10.1016 / j.ahj.2024.01.011Cruz, D. E. et al. Metabolomic Analysis of Coronary Heart Disease in an African American Cohort From the Jackson Heart Study. JAMA Cardiol 7, 184-194 (2022). doi .org / 10.1001 / j amacardio.2021.4925Peyser, P. A. et al. Heritability of coronary artery calcium quantity measured by electron beam computed tomography in asymptomatic adults. Circulation 106, 304-308 (2002). doi.org / 10.1161 / 01. cir.0000022664.21832.5dWollert, K. C., Kempf, T. & Wallentin, L. Growth Differentiation Factor 15 as a Biomarker in Cardiovascular Disease. Clin Chem 63, 140-151 (2017).doi .org / 10.1373 / clinchem.2016.255174Liu, D. et al. Myocardial Recovery in Recent Onset Dilated Cardiomyopathy: Role of CDCP1 and Cardiac Fibrosis. Circ Res 133, 810-825 (2023).doi.org / 10.1161 / circresaha.123.323200Zhang, Q., Wen, X. H, Tang, S. L., Zhao, Z. W. & Tang, C. K. Role and therapeutic potential of gelsolin in atherosclerosis. J Mol Cell Cardiol 178, 59-67 (2023). doi.org / 10.1016 / j.yjmcc.2023.03.012Zhang, K., Li, M., Yin, L., Fu, G. & Liu, Z. Role of thrombospondin 1 and thrombospondin 2 in cardiovascular diseases (Review). Int J Mol Med 45, 1275-1293 (2020). doi.org / 10.3892 / ijmm.2020.4507Hu, W. et al. Circulating Levels of CILP2 Are Elevated in Coronary Heart Disease and Associated with Atherosclerosis. OxidMed Cell Longev 2020, 1871984 (2020). doi .org / 10.1155 / 2020 / 1871984Amor, M. et al. Genetic deletion of MMP12 ameliorates cardiometabolic disease by improving insulin sensitivity, systemic inflammation, and atherosclerotic features in mice. Cardiovasc Diabetol 22, 327 (2023). doi.org / 10.1186 / sl 2933-023-02064-3 Jenkins, W. S. et al. In vivo alpha-V beta-3 integrin expression in human aortic atherosclerosis. Heart 105, 1868-1875 (2019). doi.org / 10.1136 / heartjnl-2019-315103 Rusanescu, G., Weissleder, R. & Aikawa, E. Notch signaling in cardiovascular disease and calcification. Curr Cardiol Rev 4, 148-156 (2008).doi. org / 10.2174 / 157340308785160552Melo, U. S. et al. Enhancer hijacking at the ARHGAP36 locus is associated with connective tissue to bone transformation. Nat Commun 14, 2034 (2023). doi.org / 10.1038 / s41467-023-37585-8Kahles, F. etal. The incretin hormone GIP is upregulated in patients with atherosclerosis and stabilizes plaques in ApoE(- / -) mice by blocking monocyte / macrophage activation. Mol Metab 14, 150-157 (2018). doi. org / 10.1016 / j.molmet.2018.05.014Liu, Y. et al. The Protective Effect of UBE2G2 Knockdown Against Atherosclerosis in Apolipoprotein E-Deficient Mice and Its Association with miR-204-5p. Mol Biotechnol 64, 1045-1054 (2022). doi.org / 10.1007 / sl2033-022-00482-4Patterson, M. T. et al. Trem2 promotes foamy macrophage lipid uptake and survival in atherosclerosis. Nat Cardiovasc Res 2, 1015-1031 (2023). doi.org / 10.1038 / s44161-023-00354-3Dubourg, V. et al. EGFR activation differentially affects the inflammatory profiles of female human aortic and coronary artery endothelial cells. Set Rep 13, 22827 (2023). doi.org / 10.1038 / s41598-023-50148-7Boyer-Guittaut, M. etal. The role of GABARAPLl / GECl in autophagic flux and mitochondrial quality control in MDA-MB-436 breast cancer cells. Autophagy 10, 986- 1003 (2014). doi.org / 10.4161 / auto.28390Croce, K. etal. Myeloid-related protein-8 / 14 is critical for the biological response to vascular injury. Circulation 120, 427-436 (2009).doi .org / 10.1161 / CIRCULATION AHA.108.814582Chen, H. et al. Variation in VEGFA and risk of cardiovascular disease in the UK Biobank. Front Cardiovasc Med 10, 1240288 (2023). doi.org / 10.3389 / fcvm.2023.1240288 Zhou, Y. et al. The Role of the VEGF Family in Coronary Heart Disease. Front Cardiovasc Med 8, 738325 (2021). doi.org / 10.3389 / fcvm.2021.738325Gu, Y. etal. CTRP1 Aggravates Cardiac Dysfunction Post Myocardial Infarction by Modulating TLR4 in Macrophages. Front Immunol 12, 635267 (2021). doi.org / 10.3389 / fimmu.2021.635267Hatayama, M. etal. SLITRK1 -mediated noradrenergic projection suppression in the neonatal prefrontal cortex. Commun Biol 5, 935 (2022). doi. org / 10.1038 / s42003 -022-03891-yHatayama, M. & Aruga, J. Developmental control of noradrenergic system by SLITRK1 and its implications in the pathophysiology of neuropsychiatric disorders. Front Mol Neurosci 15, 1080739 (2022). doi.org / 10.3389 / fnmol.2022.1080739Xiong, W. et al. Identification of candidate genes and pathways in limonin-mediated cardiac repair after myocardial infarction. Biomed Pharmacother 142, 112088 (2021). doi .org / 10.1016 / j .biopha.2021.112088Cheng, J. M. et al. PCSK9 in relation to coronary plaque inflammation: Results of the ATHEROREMO-IVUS study. Atherosclerosis 248, 117-122 (2016).doi .org / 10.1016 / j .atherosclerosis.2016.03.010Sabatine, M. S. etal. Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. NEnglJMed 376, 1713-1722 (2017). doi.org / 10.1056 / NEJMoal615664 Farouk, S. S., Rader, D. J., Reilly, M. P. & Mehta, N. N. CXCL12: a new player in coronary disease identified through human genetics. Trends Cardiovasc Med 20, 204-209 (2010). doi. org / 10.1016 / j.tcm.2011.08.002Natarajan, P. etal. Association of an HDL Apolipoproteomic Score With Coronary Atherosclerosis and Cardiovascular Death. J Am Coll Cardiol 73, 2135-2145 (2019). doi.org / 10.1016 / j.jacc.2019.01.073Loh, N. Y. et al. Bidirectional Mendelian Randomization Highlights Causal Relationships Between Circulating 1NHBC and Multiple Cardiometabolic Diseases and Traits. Diabetes 13, 2084-2094 (2024). doi.org / 10.2337 / db24-0168Kavousi, M. et al. Multi -ancestry genome-wide study identifies effector genes and druggable pathways for coronary artery calcification. Nature genetics 55, 1651-1664 (2023). doi.org / 10.1038 / s41588-023-01518-4O'Donnell, C. J. et al. Genome-wide association study for coronary artery calcification with follow-up in myocardial infarction. Circulation 124, 2855-2864 (2011). doi.org / 10.1161 / CIRCULATIONAHA.110.974899Chung, A., Reilly, M. P. & Bauer, R. C. ADAMTS7: a Novel Therapeutic Target in Atherosclerosis. Current atherosclerosis reports 25, 447-455 (2023). doi.org / 10.1007 / sl 1883-023-01115-0Ma, X. et al. PHACTR1, a coronary artery disease risk gene, mediates endothelial dysfunction. Front Immunol 13, 958677 (2022). doi.org / 10.3389 / fimmu.2022.958677Shah, P. W., Reinberger, T., Hashmi, S., Aherrahrou, Z. & Erdmann, J. MRAS in coronary artery disease-Unchartered territory. IUBMB Life 76, 300-312 (2024). doi.org / 10.1002 / iub.2805Zhao, Y. et al. GIGYF1 loss of function is associated with clonal mosaicism and adverse metabolic health. Nature communications 12, 4178 (2021). doi.org / 10.1038 / s41467-021-24504-yBerul, C. I. et al. DMPK dosage alterations result in atrioventricular conduction abnormalities in a mouse myotonic dystrophy model. J Clin Invest 103, Rl-7 (1999). doi.org / 10.1172 / JCI5346Lezzerini, M. etal. Ribosomal protein gene RPL9 variants can differentially impair ribosome function and cellular metabolism. Nucleic acids research 48, 770-787 (2020). doi.org / 10.1093 / nar / gkzl042Romay, M. C. etal. Age-related loss of Notch3 underlies brain vascular contractility deficiencies, glymphatic dysfunction, and neurodegeneration in mice. J Clin Invest 134 (2024). doi.org / 10.1172 / JC1166134Fernandez-Laso, V. etal. TWEAK blockade decreases atherosclerotic lesion size and progression through suppression of STAT1 signaling in diabetic mice. Scientific reports 7, 46679 (2017). doi.org / 10.1038 / srep46679Goyette, J. etal. Pleiotropic roles of S100A12 in coronary atherosclerotic plaque formation and rupture. Journal of immunology 183, 593-603 (2009). doi.org / 10.4049 / jimmunol.0900373Divers, J. etal. Genome-wide association study of coronary artery calcified atherosclerotic plaque in African Americans with type 2 diabetes. BMC Genet 18, 105 (2017). doi.org / 10.1186 / sl2863-017-0572-9Yiu, G. K., Kaunisto, A., Chin, Y. R. & Toker, A. NF AT promotes carcinoma invasive migration through glypican-6. The Biochemical journal 440, 157-166 (2011). doi.org / 10.1042 / BJ20110530Bundy, K., Boone, J. & Simpson, C. L. Wnt Signaling in Vascular Calcification.Front Car di ovasc Med 8, 708470 (2021). doi.org / 10.3389 / fcvm.2021.708470 Reinhold, S. & Blankesteijn, W. M. Wnt / beta-Catenin Inhibitor Di ckkopf 1.Arterioscler Thromb Vase Bio! NT 121-123 (2019).doi.org / 10.1161 / ATVBAHA.118.312144Common Metabolic Disease Knowledge Portal.Mosquera, J. V. etal. Integrative single-cell meta-analysis reveals disease-relevant vascular cell states and markers in human atherosclerosis. Cell Rep 42, 113380 (2023). doi.org / 10.1016 / j . celrep.2023.113380Meir, K. S. & Leitersdorf, E. Atherosclerosis in the apolipoprotein-E-deficient mouse: a decade of progress. Arterioscler Thromb Vase Biol 24, 1006-1014 (2004). doi.org / 10.1161 / 01. ATV.0000128849.12617.f4Lin, A. etal. Deep learning-enabled coronary CT angiography for plaque and stenosis quantification and cardiac risk prediction: an international multicentre study. Lancet Digit Health 4, e256-e265 (2022). doi.org / 10.1016 / S2589-7500(22)00022-X Murthy, V. L. et al. Improved cardiac risk assessment with noninvasive measures of coronary flow reserve. Circulation 124, 2215-2224 (2011).doi.org / 10.1161 / CIRCULATIONAHA.l 11.050427Han, D. et al. Association of Plaque Location and Vessel Geometry Determined by Coronary Computed Tomographic Angiography With Future Acute Coronary Syndrome-Causing Culprit Lesions. JAMA Cardiol 7, 309-319 (2022).doi. org / 10.1001 / jamacardio.2021.5705Tcheandjieu, C. etal. Large-scale genome-wide association study of coronary artery disease in genetically diverse populations. Nat Med 28, 1679-1692 (2022).doi . org / 10.1038 / s41591 -022-01891 -3de Vries, P. S. et al. Whole-genome sequencing uncovers two loci for coronary artery calcification and identifies ARSE as a regulator of vascular calcification. Nat Cardiovasc Res 2, 1159-1172 (2023). doi.org / 10.1038 / s44161-023-00375-yShah, S. H. et al. Association of a peripheral blood metabolic profile with coronary artery disease and risk of subsequent cardiovascular events. Circulation. Cardiovascular genetics 3, 207-214 (2010). doi.org / 10.1161 / CIRCGENETICS.109.852814Basak, T. et al. Plasma proteomic analysis of stable coronary artery disease indicates impairment of reverse cholesterol pathway. Sci Rep 6, 28042 (2016). doi.org / 10.1038 / srep28042Helgason, H. et al. Evaluation of Large-Scale Proteomics for Prediction of Cardiovascular Events. JAMIA 330, 725-735 (2023). doi. org / 10.100 l / jama.2023.13258Hu, Z. et al. Single-Cell Transcriptomic Atlas of Different Human Cardiac Arteries Identifies Cell Types Associated With Vascular Physiology. Arterioscler Thromb Vase Biol l, 1408-1427 (2021). doi.org / 10.1161 / ATVBAHA.120.315373Wu, M., Wu, Y., Tang, S., Huang, J. & Wu, Y. Single-cell RNA-seq uncovers distinct pathways and genes in endothelial cells during atherosclerosis progression. Front Mol Biosci 10, 1176267 (2023). doi.org / 10.3389 / fmolb.2023.1176267Wirka, R. C. et al. Atheroprotective roles of smooth muscle cell phenotypic modulation and the TCF21 disease gene as revealed by single-cell analysis. Nature medicine 25, 1280-1289 (2019). doi.org / 10.1038 / s41591-019-0512-5Carr, J. J. etal. Association of Coronary Artery Calcium in Adults Aged 32 to 46 Years With Incident Coronary Heart Disease and Death. JAMA cardiology 2, 391-399 (2017). doi.org / 10.1001 / jamacardio.2016.5493Kolossvary, M. etal. Proteomic Signature of Subclinical Coronary Artery Disease in People With HIV: Analysis of the REPRIEVE Mechanistic Substudy. J Infect Dis 226, 1809-1822 (2022). doi.org / 10.1093 / infdis / jiacl96Herrington, D. M. et cd. Proteomic Architecture of Human Coronary and Aortic Atherosclerosis. Circulation 137, 2741-2756 (2018).doi . org / 10.1161 / CIRCULATION AHA.118.034365Oh, H. S. et al. Organ aging signatures in the plasma proteome track health and disease. Nature 624, 164-172 (2023). doi.org / 10.1038 / s41586-023-06802-lChan, M. Y. et al. Prioritizing Candidates of Post-Myocardial Infarction Heart Failure Using Plasma Proteomics and Single-Cell Transcriptomics. Circulation 142, 1408-1421 (2020). doi. org / 10.1161 / CIRCULATIONAHA.119.045158Deo, R. et al. Proteomic cardiovascular risk assessment in chronic kidney disease. Eur Heart J 44, 2095-2110 (2023). doi.org / 10.1093 / eurheartj / ehadll5Choucair, N. et al. Evidence that homozygous PTPRD gene microdeletion causes trigonocephaly, hearing loss, and intellectual disability. Mol Cytogenet 8, 39 (2015). doi.org / 10.1186 / sl3039-015-0149-0Bestepe, F. et al. Deficiency of miR-409-3p improves myocardial neovascularization and function through modulation of DNAJB9 / p38 MAPK signaling. Mol Ther Nucleic Acids 32, 995-1009 (2023). doi.org / 10.1016 / j.omtn.2023.05.021Sass, F. et al. NK2R control of energy expenditure and feeding to treat metabolic diseases. Nature 635, 987-1000 (2024). doi.org / 10.1038 / s41586-024-08207-0Choi, B. etal. Proteomic Biomarkers of Quantitative Interstitial Abnormalities in COPD Gene and CARDIA Lung Study. Am J Re spir Crit Care Med 209, 1091-1100 (2024). doi.org / 10.1164 / rccm.202307-1129OCFriedman, G. D. et al. CARDIA: study design, recruitment, and some characteristics of the examined subjects. J Clin Epidemiol 41, 1105-1116 (1988). doi.org / 10.1016 / 0895-4356(88)90080-7Carr, J. J. et al. Calcified coronary artery plaque measurement with cardiac CT in population-based studies: standardized protocol of Multi-Ethnic Study of Atherosclerosis (MESA) and Coronary Artery Risk Development in Young Adults (CARDIA) study. Radiology 234, 35-43 (2005). doi.org / 10.1148 / radiol.2341040439Gidding, S. S. etal. Pathobiological Determinants of Atherosclerosis in Youth (PDAY) Risk Score in Young Adults Predicts Coronary Artery and Abdominal Aorta Calcium in Middle Age: The CARDIA Study. Circulation 133, 139-146 (2016). doi . org / 10.1161 / CIRCUL ATIONAHA.115.018042Kannel, W. B., Feinleib, M„ McNamara, P. M., Garrison, R. J. & Castelli, W. P. An investigation of coronary heart disease in families. The Framingham offspring study. Am J Epidemiol 110, 281-290 (1979). doi.org / 10.1093 / oxfordjournals.aje.all2813 Levy, D., Garrison, R. J., Savage, D. D., Kannel, W. B. & Castelli, W. P. Left ventricular mass and incidence of coronary heart disease in an elderly cohort. The Framingham Heart Study. Ann Intern Med 110, 101-107 (1989).Shah, R. V. et al. Association of Multiorgan Computed Tomographic Phenomap With Adverse Cardiovascular Health Outcomes: The Framingham Heart Study. JAMA cardiology 2, 1236-1246 (2017). doi.org / 10.1001 / jamacardio.2017.3145Nayor, M. et al. Aptamer-Based Proteomic Platform Identifies Novel Protein Predictors of Incident Heart Failure and Echocardiographic Traits. Circ Heart Fail 13, e006749 (2020). doi.org / 10.1161 / circheartfailure.l 19.006749Ferkingstad, E. et al. Large-scale integration of the plasma proteome with genetics and disease. Nat Genet 53, 1712-1721 (2021). doi.org / 10.1038 / s41588-021-00978-w Aragam, K. G. etal. Discovery and systematic characterization of risk variants and genes for coronary artery disease in over a million participants. Nature genetics 54, 1803-1815 (2022). doi.org / 10.1038 / s41588-022-01233-6Hartiala, J. A. et al. Genome-wide analysis identifies novel susceptibility loci for myocardial infarction. Eur Heart J 42, 919-933 (2021).doi.org / 10.1093 / eurheartj / ehaal 040Skrivankova, V. W. et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomisation (STROBE-MR): explanation and elaboration. BmjYIS, n2233 (2021). doi.org / 10.1136 / bmj.n2233Stuart, T. et al. Comprehensive Integration of Single-Cell Data. Cell 177, 1888-1902 el821 (2019). doi.org / 10.1016 / j.cell.2019.05.031Germain, P. L., Lun, A., Garcia Meixide, C., Macnair, W. & Robinson, M. D. Doublet identification in single-cell sequencing data using scDblFinder. FlOOORes 10, 979 (2021). doi.org / 10.12688 / fl 000research.73600.2Yang, S. et al. Decontamination of ambient RNA in single-cell RNA-seq with DecontX. Genome biology 21, 57 (2020). doi.org / 10.1186 / sl3059-020-1950-6 Hafemeister, C. & Satija, R. Normalization and variance stabilization of single-cell RNA-seq data using regularized negative binomial regression. Genome biology 20, 296 (2019). doi.org / 10.1186 / sl3059-019-l 874-1Korsunsky, I. etal. Fast, sensitive and accurate integration of single-cell data with Harmony. Nat Methods 16, 1289-1296 (2019). doi. org / 10.1038 / s41 92-019-0619-0 Tabula Sapiens, C. et al. The Tabula Sapiens: A multiple-organ, single-cell transcriptomic atlas of humans. Science 376, eabl4896 (2022). doi.org / 10.1126 / science.abl4896 Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome biology 15, 550 (2014). doi.org / 10.1186 / S13059-014-0550-8Zhang, J. etal. Plasma proteome analyses in individuals of European and African ancestry identify cis-pQTLs and models for proteome-wide association studies. Nature genetics 54, 593-602 (2022). doi.org / 10.1038 / s41588-022-01051-wBurgess, S. & Thompson, S. G. Interpreting findings from Mendelian randomization using the MR-Egger method. European journal of epidemiology 32, 377-389 (2017). doi.org / 10.1007 / sl0654-017-0255-xBowden, J., Davey Smith, G., Haycock, P. C. & Burgess, S. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet Epidemiol 40, 304-314 (2016). doi. org / 10.1002 / gepi.21965Burgess, S., Dudbridge, F. & Thompson, S. G. Combining information on multiple instrumental variables in Mendelian randomization: comparison of allele score and summarized data methods. StatMed35, 1880-1906 (2016). doi.org / 10.1002 / sim.6835Davey Smith, G. & Hemani, G. Mendelian randomization: genetic anchors for causal inference in epidemiological studies. Human molecular genetics 23, R89-98 (2014). doi . org / 10.1093 / hmg / ddu 328Zhou, D. et al. A unified framework for joint-tissue transcriptome-wide association and Mendelian randomization analysis. Nature genetics 52, 1239-1246 (2020).doi .org / 10.1038 / s41588-020-0706-2Betti, M. J., Aldrich, M. C. & Gamazon, E. R. Minimum entropy framework identifies a novel class of genomic functional elements and reveals regulatory mechanisms at human disease loci. bioRxiv (2023). doi. org / 10.1101 / 2023.06.11.544507 Wojczynski, M. K. et al. Genetics of coronary artery calcification among African Americans, a meta-analysis. BMC Med Genet 14, 75 (2013).doi.org / 10.1186 / 1471-2350-14-75Wang, J., Wei, L., Yang, X. & Zhong, J. Roles of Growth Differentiation Factor 15 in Atherosclerosis and Coronary Artery Disease. J Am Heart Assoc 8, e012826 (2019). doi.org / 10.1161 / j aha.119.012826Deneka, A., Korobeynikov, V. & Golemis, E. A. Embryonal Fyn-associated substrate (EFS) and CASS4: The lesser-known CAS protein family members. Gene 570, 25-35 (2015). doi.org / 10.1016 / j.gene.2015.06.062Zhao, C. F. & Herrington, D. M. The function of cathepsins B, D, and Xin atherosclerosis. Am J Cardiovasc Dis 6, 163-170 (2016).Hebbel, R. P. et al. Abnormal Endothelial Gene Expression Associated With Early Coronary Atherosclerosis. J Am Heart Assoc 9, e016134 (2020). doi.org / 10.1161 / jaha.120.016134Baxter, R. M., Crowell, T. P., George, J. A., Getman, M. E. & Gardner, H. The plant pathogenesis related protein GLIPR-2 is highly expressed in fibrotic kidney and promotes epithelial to mesenchymal transition in vitro. Matrix Biol 26, 20-29 (2007). doi.org / 10.1016 / j.matbio.2006.09.005106 Rypdal, K. B. et al. The extracellular matrix glycoprotein ADAMTSL2 is increased in heart failure and inhibits TGF signalling in cardiac fibroblasts. Sci Rep 11, 19757 (2021). doi .org / 10.1038 / s41598-021 -99032-2107 Chen, Y. et al. Tetranectin as a Potential Biomarker for Stable Coronary Artery Disease. Sci Rep 5, 17632 (2015). doi.org / 10.1038 / srepl7632108 Goncalves, I. et al. Elevated Plasma Levels of MMP-12 Are Associated With Atherosclerotic Burden and Symptomatic Cardiovascular Disease in Subjects With Type 2 Diabetes. Arterioscler Thromb Vase Biol 35, 1723-1731 (2015).doi. org / 10.1161 / atvbaha.l 15.305631109 Brauninger, H. et al. Matrix metalloproteinases in coronary artery disease and myocardial infarction. Basic Res Cardiol 118, 18 (2023). doi.org / 10.1007 / s00395-023-00987-2 110 Reed, M. J. et al. Expression of thrombospondins by endothelial cells. Injury is correlated with TSP-1. Am J Pathol 147, 1068-1080 (1995).111 Heimbiirger, S. M. et al. Glucose-dependent insulinotropic polypeptide (GIP) and cardiovascular disease. Peptides 125, 170174 (2020).doi. org / 10.1016 / j. peptides.2019.170174112 Ekim Kocabey, A., Kost, L., Gehlhar, M., Rbdel, G. & Gey, U. Mitochondrial Sco proteins are involved in oxidative stress defense. Redox Biol 21, 101079 (2019). doi.org / 10.1016 / j. redox.2018.101079113 Gerhards, R. et al. Oligodendrocyte myelin glycoprotein as a novel target for pathogenic autoimmunity in the CNS. Acta Neuropathol Commun 8, 207 (2020).doi .org / 10.1186 / s40478-020-01086-2114 Papaioannou, P., Wallace, M. J., Malhotra, N., Mohler, P. J. & El Refaey, M. Biochemical Structure and Function of TRAPP Complexes in the Cardiac System. JACC Basic Transl Sci 8, 1599-1612 (2023). doi.org / 10.1016 / j.jacbts.2023.03.011115 Saijo, S. et al. Serine / arginine-rich splicing factor 7 regulates p21-dependent growth arrest in colon cancer cells. J Med Invest 63, 219-226 (2016). doi.org / 10.2152 / jmi.63.219

[0193] It will be understood that various details of the presently disclosed subject matter can be changed without departing from the scope of the subject matter disclosed herein.Furthermore, the foregoing description is for the purpose of illustration only, and not for the purpose of limitation.

Claims

CLAIMSWhat is claimed is:

1. A method of diagnosing or predicting atherosclerosis or progression thereof in a subject, comprising:(a) obtaining a biological sample from the subject;(b) assessing, in the biological sample, a plurality of biomarkers selected from the group consisting of those set forth in Tables 4A and 4B;(c) generating an atherosclerosis risk score based on the quantified concentrations; and(d) identifying the subject as having atherosclerosis or an increased risk thereof when the atherosclerosis risk score exceeds a threshold value.

2. The method of claim 1, wherein the plurality of biomarkers comprises one or more biomarkers selected from the group consisting of N0TCH3, SPINK2, S100A12, RPP25, OAF, HS6ST3, TNFSF12, and GPC6.

3. The method of claim 1, wherein the biological sample comprises whole blood, plasma, or serum.

4. The method of claim 1, wherein the biological sample is a plasma or serum sample, and obtaining the plasma or serum sample comprises (i) processing whole blood to isolate plasma or serum; and (ii) performing protein denaturation and / or enzymatic digestion of the plasma or serum prior to assessing the plurality of biomarkers.

5. The method of claim 1, wherein assessing the plurality of biomarkers comprises quantifying circulating protein levels using an aptamer-based proteomic platform.

6. The method of claim 1, wherein assessing the plurality of biomarkers comprises performing an immunoassay-based quantification, selected from the group consisting of enzyme-linked immunosorbent assay (ELISA) and multiplex bead-based immunoassay.

7. The method of claim 1, wherein assessing the plurality of biomarkers comprises performing a mass spectrometry-based proteomic assay.

8. The method of claim 1, wherein assessing the plurality of biomarkers comprises obtaining single-cell or bulk transcriptomic data for genes corresponding to the biomarkers.

9. The method of claim 1, wherein assessing the plurality of biomarkers is performed using an automated analytical device configured with a calibration curve tailored to atherosclerosis-associated biomarker concentration ranges.

10. The method of claim 1, wherein generating the atherosclerosis risk score comprises applying a multivariable regression model that assigns biomarker-specific coefficients to the quantified concentrations or expression levels.

11. The method of claim 1, wherein generating the atherosclerosis risk score comprises applying a machine-learning model trained on atherosclerosis phenotypes, the machine-learning model comprising biomarker-specific parameters learned from labeled training data.

12. The method of claim 1, wherein the multivariable model is validated using receiver operating characteristic (ROC) analysis, and the threshold value is selected based on an area under the ROC curve (AUC) performance characteristic.

13. The method of claim 1, wherein the threshold value corresponds to a predetermined percentile of the risk score distribution in a reference population without clinical atherosclerosis.

14. The method of claim 1, wherein generating the atherosclerosis risk score further comprises integrating clinical covariates selected from age, sex, and race into the multivariable model.

15. The method of claim 1, wherein generating the atherosclerosis risk score further comprises adjusting for one or more covariates selected from age, sex, and race.

16. The method of claim 1, wherein generating the atherosclerosis risk score comprises applying a multivariable model comprising biomarker-specific coefficients derived from a regression analysis or machine-learning algorithm trained on atherosclerosis phenotypes.

17. The method of claim 1, wherein generating the atherosclerosis risk score is performed by a computing device executing instructions stored on a non-transitory computer-readable medium, the instructions causing the computing device to:(i) receive, as input, the quantified concentrations or expression levels of the plurality of biomarkers;(ii) apply the multivariable model comprising biomarker-specific coefficients; and (iii) output the atherosclerosis risk score and an indication of whether the atherosclerosis risk score exceeds the threshold value.

18. The method of claim 1, further comprising administering to the subject a therapeutic agent or intervention for treating or preventing atherosclerosis when the atherosclerosis risk score exceeds the threshold value.

19. The method of claim 18, wherein the therapeutic agent is selected from the group consisting of a statin, a bile sequestrant, niacin, a PCSK-9 inhibitor, and a fibrate.

20. The method of claim 1, further comprising performing additional testing for coronary risk, selected from the group consisting of coronary artery calcification imaging, echocardiography, cardiac catheterization, and stress testing.

21. A kit for diagnosing or predicting atherosclerosis or progression thereof in a subject, comprising:(a) a plurality of reagents configured for assessing, in a biological sample obtained from the subject, a plurality of biomarkers selected from the group consisting of those set forth in Tables 4A and 4B; and(b) instructions, stored on a non-transitory computer-readable medium or provided in printed form, for calculating an atherosclerosis risk score based on the assessment of the plurality of biomarkers using predetermined coefficients derived from a statistical or computational modeling analysis trained on atherosclerosis phenotypes.

22. The kit of claim 21, wherein the plurality of biomarkers comprises one or more biomarkers selected from the group consisting of N0TCH3, SPINK2, S100A12, RPP25, OAF, HS6ST3, TNFSF12, and GPC6.

23. The kit of claim 21, wherein the plurality of reagents is provided in a multiplexed assay cartridge configured for simultaneous detection of at least two biomarkers in a single analytical run.

24. The kit of claim 21, wherein the multiplexed assay cartridge comprises a non-standard calibration curve optimized for atherosclerosis biomarker concentration ranges.

25. The kit of claim 21, wherein the reagents comprise antibody pairs configured for use in a sandwich immunoassay for detecting one or more of the atherosclerosis biomarkers.

26. The kit of claim 21, wherein the plurality of reagents comprises aptamer-based binding agents configured to bind to one or more of the biomarkers.

27. The kit of claim 21, further comprising mass-spectrometry compatible reagents selected from denaturation reagents, digestion enzymes, or ionization-enhancing buffers.

28. The kit of claim 21, further comprising nucleic-acid amplification or sequencing reagents for detecting transcript levels corresponding to one or more biomarkers.

29. The kit of claim 21, further comprising an automated immunoassay or analytical device configured to process the multiplexed assay cartridge.

30. The kit of claim 21, wherein the automated device is configured to transmit biomarker concentration data to a computing device executing the scoring algorithm.

31. The kit of claim 21, further comprising calibration standards containing predetermined concentrations of at least one biomarker for calibrating the automated device.

32. The kit of claim 21, wherein the instructions stored on the non-transitory computer-readable medium comprise executable code for applying a multivariable model comprising biomarker-specific coefficients derived from regression or machine-learning analysis trained on atherosclerosis phenotypes.

33. The kit of claim 21, wherein the instructions further comprise executable code for comparing the computed atherosclerosis risk score to a threshold value, the threshold value being determined from a receiver operating characteristic (ROC) curve for atherosclerosis classification.

34. The kit of claim 21, wherein the instructions further comprise executable code for outputting a risk classification indicating whether the subject has an increased risk of atherosclerosis or progression thereof.

35. The kit of claim 21, wherein the instructions further comprise executable code for integrating clinical covariates selected from age, sex, and race into the multivariable model used to compute the atherosclerosis risk score.

36. The kit of claim 21, wherein the computing device is configured to automatically trigger the risk-score calculation upon receiving biomarker concentration data from the analytical device.

37. The kit of claim 21, wherein the instructions further comprise executable code for storing and comparing biomarker profdes across multiple time points for longitudinal risk assessment.

38. A computer-implemented method of diagnosing or predicting atherosclerosis or progression thereof in a subject, comprising:(a) receiving as input quantified concentrations or expression levels of a plurality of biomarkers selected from the group consisting of those set forth in Tables 4A and 4B;(b) applying a multivariable model trained on atherosclerosis phenotypes to said concentrations or expression levels, wherein the model comprises biomarker-specific coefficients derived from regression analysis or machine learning; and(c) generating and outputting an atherosclerosis risk score indicative of the subject’s atherosclerosis status.

39. The method of claim 38, wherein the plurality of biomarkers comprises one or more biomarkers selected from the group consisting of N0TCH3, SPINK2, S100A12, RPP25, OAF, HS6ST3, TNFSF12, and GPC6.

40. The method of claim 38, wherein the quantified concentrations or expression levels are obtained from data comprising circulating proteomic measurements or single-cell or bulk transcriptomic data corresponding to the biomarkers.

41. The method of claim 38, wherein applying the multivariable model further comprises normalizing biomarker concentrations using a calibration curve tailored to atherosclerosis biomarker ranges.

42. The method of claim 38, wherein the computing device comprises a processor configured to execute instructions stored on a non-transitory computer-readable medium to apply the multivariable model.

43. The method of claim 38, wherein the computing device comprises a cloud-based server configured to receive biomarker concentration or expression data from a remote assay device and to execute the multivariable model.

44. The method of claim 38, wherein the multivariable model comprises a regression model that outputs a continuous atherosclerosis risk score based on the biomarker concentrations or expression levels.

45. The method of claim 38, wherein applying the multivariable model comprises executing a trained machine-learning algorithm selected from gradient boosting regression, random forest regression, or neural network regression.

46. The method of claim 38, wherein the multivariable model is trained on atherosclerosis phenotypes comprising coronary artery calcification (CAC) status, CAC extent, or incident CAC development in a cohort of subjects.

47. The method of claim 38, wherein applying the multivariable model further comprises integrating clinical covariates selected from age, sex, and race into the atherosclerosis risk score calculation.

48. The method of claim 38, wherein generating the atherosclerosis risk score further comprises comparing the atherosclerosis risk score to a threshold value determined using receiver operating characteristic (ROC) analysis for atherosclerosis classification.

49. The method of claim 38, wherein outputting the atherosclerosis risk score comprises generating a graphical user interface that displays (i) the atherosclerosis risk score and (ii) a classification status indicating whether the risk score exceeds the threshold value.

50. The method of claim 38, further comprising receiving, at the computing device, biomarker concentration data transmitted from an analytical or immunoassay device, and automatically applying the multivariable model to the received data.

51. The method of claim 38, further comprising:(i) receiving biomarker concentration or expression data for the subject at two or more time points;(ii) computing an atherosclerosis risk score for each time point using the multivariable model; and(iii) determining progression of atherosclerosis risk based on changes in the computed atherosclerosis risk scores over time.

Images