Type 2 diabetes (T2D) is a heterogeneous disease shaped by genetic pathways related to insulin resistance and β-cell dysfunction, but how this heterogeneity is reflected molecularly remains unclear. We integrated partitioned polygenic scores (pPS) with proteomic and metabolomic profiling to define molecular signatures of T2D and their clinical relevance. We analyzed UK Biobank participants with genomic, proteomic, and metabolomic data. In a disease-free training subset, we used LASSO regression to identify multi-omic signatures associated with each pPS by jointly modeling proteins and metabolites. In an independent testing set, we constructed multi-omic scores and examined their associations with clinical traits and diabetes-related outcomes. Mediation analyses were used to investigate putative causal pathways. Key findings were evaluated in the Multi-Ethnic Study of Atherosclerosis (MESA). We identified distinct multi-omic signatures that capture the molecular architecture of T2D genetic risk across physiological subtypes. Compared with genetic scores alone, multi-omic pPS showed larger effect sizes and better disease discrimination. These scores recapitulated subtype-specific physiology and were associated with T2D risk. The Beta-Cell 2 multi-omic score showed marked stratification for insulin use, which was replicated in MESA, where it also predicted future insulin use. Mediation analyses implicated lipoprotein remodeling and fatty acid metabolism in the Lipodystrophy 1 cluster, accounting for 30-45% of the total effect of pPS on T2D risk. Integrating process-specific genetic risk with circulating multi-omic profiles reveals biologically distinct endotypes of T2D and supports a framework for improved patient stratification and risk assessment.

Introduction and Objective: Phenotypic clustering of T2D into aging-related (MARD), obesity-related (MOD), insulin-deficient (SIDD), and insulin-resistant (SIRD) subgroups may capture heterogeneity, but performance and stability in diverse populations are uncertain. Methods: We performed established k-means clustering models in 871 prevalent and 462 incident T2D cases with data for age at onset, BMI, HbA1c, HOMA2-IR, and HOMA2-β in the Multiethnic Study of Atherosclerosis across four exams (2000-2007). We assessed subgroup distribution, transitions, and mortality between exams. In 869 prevalent cases with long-term follow-up, we evaluated discrimination for mortality, CVD, and CKD using three models: discrete subgroup membership, continuous subgroup probabilities, or the original variables. Results: MARD was the most common subgroup (50% of prevalent; 64% of incident), while SIDD was rare among incident T2D (0.6%). Incident cases had higher BMI and HOMA2-β and lower HbA1c than prevalent T2D. Silhouette indices were modest (0.22-0.23). Across exams, MARD and SIRD increased in prevalence, and SIDD showed the greatest instability (only 21-42% remaining SIDD) and higher between exam mortality (4-10%). Over 18 years, cumulative incidence of mortality, CVD, and CKD were 38%, 32%, and 50%. Discrimination of these outcomes was similar across the three models (C-indices: mortality 0.696-0.700; CVD 0.642-0.647; CKD 0.669-0.677). Conclusion: In a multiethnic cohort, established T2D clustering shows limited separation, substantial subgroup transitions—especially for SIDD—and similar prognostic performance compared with clinical variables. Accounting for diabetes duration and progression may improve future clustering approaches. Disclosure L. Olson: None. F. Hsu: None. K. Smith: None. R. Dagostino Jr: Other - I am on a DSMB for this company; Current; Daiichi Sankyo. Consultant; Ended; Merck & Co., Inc. Other - I am on several DSMBs for this company; Current; AstraZeneca. M. Bancks: None. M. Udler: Advisory Panel; Ended; Novo Nordisk. Research Support; Current; Novo Nordisk. Funding National Institutes of Health (1U01DK140778 and 1U01DK140757)

Introduction and Objective: T2D is a heterogeneous disease involving defects in multiple metabolic processes, which could be assessed by partitioned pathway-specific polygenic scores (pPS) developed from clustering T2D GWAS lead variants based on associations with related cardiometabolic traits. Here, we demonstrate the use of T2D pPS to identify T2D patient subgroups with distinct clinical features and outcomes. Methods: We calculated standardized values of 7 T2D pPS (beta.cell.1, beta.cell.2, lipodystrophy.1, lipodystrophy.2, hyper.insulin, proinsulin, and obesity) in 29,887 US whites (All of Us, BioVU, MGBB) and 2,490 US Hispanics (SOL) with T2D. In each cohort, we assessed association of pPS with cardiometabolic traits adjusted for age, sex, BMI and principal components. We grouped individuals by k-means consensus clustering based on their pPS profiles, and meta-analyzed group-specific association with metabolic traits and outcomes including cardiovascular disease, diabetic kidney disease and diabetic retinopathy (DR). Results: The associations of pPSs with respective cardiometabolic traits were consistent with previous findings (P<0.0007). We identified 4 genetically anchored T2D subgroups with differential metabolic traits and outcomes, of which three subgroups replicate across all cohorts. Subgroup 1, characterized by lower beta.cell and lipodystrophy pPSs was associated with lower HbA1c, higher BMI, and reduced risk for DR. Subgroup 2, characterized by higher beta.cell and lipodystrophy pPSs was associated with higher HbA1c, lower BMI and HDL-C. Subgroup 3, characterized by higher beta.cell but lower lipodystrophy pPSs was associated with lower BMI, higher HDL-C, and higher risk for DR (β estimate=-0.21 to 0.21, P<0.0036). Conclusion: This study demonstrates the utility of pPS to subgroup individuals with T2D based on genetic profiles representing their underlying disease mechanisms. Disclosure J. Kim: None. K. Smith: None. P. Wu: None. X. Zhong: None. M.M. Shuey: None. F. Hsu: None. E. Gamazon: None. J.I. Rotter: None. M. Udler: Advisory Panel; Ended; Novo Nordisk. Research Support; Current; Novo Nordisk. Q. Qi: None. J. Mercader: None. M.C. Ng: None. Funding The National Institute of Diabetes and Digestive and Kidney Diseases (U01DK140757, U01DK140761, U01DK140778, U01DK140952)

Objective: Clinical heterogeneity in youth-onset type 2 diabetes is less understood than that of adult-onset type 2 diabetes. We performed phenotypic clustering of youth-onset type 2 diabetes to determine if clusters provided clinical utility. Research Design and Methods: We performed data-driven clustering in a diverse subset of autoantibody-negative, clinician-diagnosed type 2 diabetes before age 20 years in the TODAY [n=525] and SEARCH [n=333] studies. Participants were clustered using: 1) similar variables as previously described in adults and 2) novel routinely available clinical variables. We assessed the effectiveness of the clusters, as well as that of simple clinical measures, to predict treatment response in the TODAY clinical trial. Results: There were three youth-onset type 2 diabetes clusters: 1) Youth-onset insulin-deficient diabetes (YIDD-T2), 2) Youth-onset insulin-resistant diabetes (YIRD-T2), and 3) Intermediate youth-onset diabetes (IYOD-T2). These clusters had differential responses to therapies and risk of treatment failure in the TODAY study, with those in the YIDD-T2 cluster experiencing the highest rate of treatment failure, regardless of treatment arm. YIDD-T2 also had high rates of type 2 diabetes complications. We then generated three novel clusters, with also different rates of treatment failure, using variables available in routine clinical practice. Compared to both clustering methods, simple clinical measures performed comparably or better at predicting treatment response and complications. Conclusion: Youth-onset type 2 diabetes can be characterized into reproducible clusters that demonstrate differential response to treatments and risk of complications. Nevertheless, cluster membership did not add clinical utility beyond simple clinical measures for predicting outcomes.

Introduction and Objective: Type 2 diabetes (T2D) is polygenic and heterogeneous. Clustering of genome-wide association study (GWAS) signals has identified variant clusters underlying T2D trait heterogeneity but has not resolved them to genes and pathways. We tested whether clustering diabetes-associated SNPs by gene and functional annotation could resolve T2D genetic signals to pathways. Methods: We developed PIGEAN, a Bayesian method inferring SNP-to-gene/annotation associations from GWAS summary statistics, and EAGGL, a soft-clustering method using non-negative matrix factorization on PIGEAN outputs. We applied both to meta-analyzed T2D GWAS SNPs to define pathway clusters (factors). Factor polygenic scores (FPS) were made from each factor to test associations with T2D traits in 4 models, with or without T2D polygenic adjustment and using joint or individual factors, in AMP-T2D Knowledge Portal GWAS summary statistics and individual-level data from the United Kingdom Biobank. Results: PIGEAN identified 149 genes with >50% likelihood of T2D involvement by annotation; 67 were factor-specific. EAGGL clustered these into 11 annotation-derived factors. These factors overlapped T2D SNP-trait clusters, but aggregated liver clusters, resolved additional beta cell and lipid clusters, and introduced 2 novel insulin-response clusters. FPS trait associations were concordant, notably CRP with Adipocyte Hypertrophy p = 2.0 × 10-3, NAFLD with Hepatic Steatosis p = 7.0 × 10-3, LPA with Non-Leptin Adipokine Signaling p = 9.4 × 10-4, BMI with Leptin Signaling p = 3.3 × 10-3 and Obesity p < 10-300, and ALT with Adipose Failure p = 6.2 × 10-5. Of PIGEAN-identified genes, 58% (86) were novel findings, including PDE8B. Conclusion: Annotation-informed clustering identified biologically interpretable factors spanning pancreatic, adipose, hepatic, and mixed dysmetabolic pathways. This approach refines our understanding of T2D heterogeneity by mapping signals to genes and pathways. Disclosure S.D. Gage: None. K. Smith: None. A.J. Deutsch: None. M. Udler: Advisory Panel; Ended; Novo Nordisk. Research Support; Current; Novo Nordisk. J. Flannick: None. Funding NIDDK (U54 DK118612)

Elevated fasting insulin levels (FI), indicative of altered insulin secretion and sensitivity, may precede type 2 diabetes (T2D) and cardiovascular disease onset. In this study, we group FI-associated genetic variants based on their genetic and phenotypic similarities and identify seven clusters with distinct mechanisms contributing to elevated FI levels. Clusters fall into two types: "non-diabetogenic hyperinsulinemia," where clusters are not associated with increased T2D risk, and "diabetogenic hyperinsulinemia," where T2D associations are driven by body fat distribution, liver function, circulating lipids, or inflammation. In over 1.1 million multi-ancestry individuals, we demonstrated that diabetogenic hyperinsulinemia cluster-specific polygenic scores exhibit varying risks for cardiovascular conditions, including coronary artery disease, myocardial infarction (MI), and stroke. Notably, the visceral adiposity cluster shows sex-specific effects for MI risk in males without T2D. This study underscores processes that decouple elevated FI levels from T2D and cardiovascular risk, offering new avenues for investigating process-specific pathways of disease.

Introduction and Objective: Clustering type 2 diabetes (T2D) genetic loci based on phenotype associations is a powerful approach for uncovering disease mechanisms. Here, we compared two recent efforts from the T2D Global Genomics Initiative (TGGI) and Mass General Hospital (MGH) groups, which used distinct input matrices and clustering methods (k-means vs Bayesian non-negative matrix factorization, bNMF). Our goal was to identify robust and biologically meaningful clusters that transcend methodological differences. Methods: To compare the published clusters, we assessed: i) cluster-specific partitioned polygenic risk scores (pPS) in individuals from the Mass General Brigham Biobank (MGBB, N=64k), ii) ) variant overlap, iii) colocalization analyses, and iv) a cross-clustering analysis, where each clustering method was applied to the opposing study’s matrix. Results: In MGBB, we identified four cluster pairs with correlated pPS (MGH/TGGI): Beta Cell 2/Beta Cell PI+; Lipodystrophy 1/Lipodystrophy; Obesity/Obesity, Cholesterol/Liver Lipid Metabolism (R>0.4, p<1e-200). Phenotypic associations with cluster pPS reinforced these findings, such as both MGH/TGGI Lipodystrophy clusters with increased triglycerides and decreased BMI and HDL. These same cluster pairs had strong overlap of shared variants (Fisher’s p< 1e-10). In colocalization analyses, loci that overlapped between matrices and belonged to one of the above four cluster pairs were enriched for quantitative trait loci signals in metabolites, proteins, and GWAS traits (Chi-square p<0.002). When bNMF was applied to the TGGI matrix, >90% of variants from the four robust clusters still clustered together (excluding those that went unclustered). Conclusion: Our analysis demonstrates that while clustering results are influenced by inputs and methods, certain T2D genetic clusters emerge consistently, indicating robust biological signatures. These findings support cluster reproducibility and highlight key clusters for further investigation into T2D pathophysiology. K. Smith: None. A.J. Deutsch: None. J.E. Gervis: None. K. Lorenz: None. H.J. Taylor: None. R. Mandla: None. B.F. Voight: Other Relationship; Eli Lilly and Company. J.I. Rotter: None. C. Yap: None. C.N. Spracklen: None. E. Zeggini: None. J.M. Mercader: None. A. Morris: None. M. Udler: Research Support; Novo Nordisk. Doris Duke (241537)

To explore multiomic regulation of the metabolome, we used machine learning to predict metabolomic variation across ∼1000 different cancer cell lines with matched omics data from 8 biomolecular classes: genomics (copy-number and mutations), epigenomics (histone post-translational modifications (PTMs) and DNA-methylation), transcriptomics and RNA splice variants, non-coding transcriptomics (miRNA and lncRNA), proteomics, and phosphoproteomics. Overall, the metabolome is tightly associated with the transcriptome, with coding and non-coding RNAs emerging as top predictors. Peripheral metabolites are predictable via levels of corresponding enzymes, while those in central metabolism require combinatorial predictors in signaling and redox pathways, and may not reflect corresponding pathway expression. We reconstruct multiomic interaction subnetworks for highly predictable metabolites, and YAP1 signaling emerged as a top global predictor across 4 omic layers. We prioritize predictive multiomic features for single-cell and spatial metabolomics assays. Top predictors were enriched for synthetic-lethal interactions and synergistic combination therapies that target compensatory metabolic modulators.

To better characterize the potential biological mechanisms underlying insulin resistance (IR) and dementia, we derive cross-population and population specific polygenic scores [PSs] for fasting insulin and IR-related partitioned PSs [pPSs]. We conduct a cross-sectional study of the associations of these genetic scores with neurological outcomes in >17k participants (36% men, mean age 55 yrs) from the Trans-Omics for Precision Medicine (TOPMed) program (50% Non-Hispanic White, 23% Black/African American, 21% Hispanic/Latino American, and 4% Asian American). We report significant negative associations (P < 0.002) of the cross-population (P = 1.3 × 10-5) and European (PEA = 3.0 × 10-8) fasting insulin PSs with total cranial volume, and of a metabolic syndrome European PS with general cognitive function (BEA = -0.13, PEA = 0.0002) and lateral ventricular volume (BEA = 0.09, PEA = 0.002). We identify suggestive negative associations (P < 0.007) of metabolic syndrome and obesity pPSs with general cognitive function, and of lipodystrophy pPSs with total cranial volume. A higher genetic predisposition to IR is associated with lower brain size, and a genetic predisposition to specific IR-related type 2 diabetes subtypes, such as metabolic syndrome and mechanisms of IR mediated through obesity and lipodystrophy, is potentially involved in cognitive decline.

Genetic risk for type 2 diabetes (T2D), coronary artery disease (CAD), and obesity can be estimated with global extended polygenic scores (gePS), which capture risk across thousands of genetic variants. Nevertheless, knowledge of underlying disease pathways remains limited. Expression Quantitative Trait Scores (eQTS) connecting gePS to gene expression could help prioritize disease-causing genes and pathways in specific tissues. We generated gePS for T2D, CAD, and body mass index (BMI) in the Genotype-Tissue Expression (GTEx) database of 838 individuals and 49 tissues. eQTS were calculated using a linear model of association between gePS and tissue transcript expression level, adjusting for age, sex, 5 genetic PCs, and 20 gene expression PCs. Gene set enrichment was performed using Enrichr with Gene Ontology pathway sets and Fisher’s Exact Test (Q < 0.05). T2D, CAD, and BMI gePS were each associated with > 500 transcript expression levels (eQTS) across multiple tissues at p < 0.05. In multiple tissues, both T2D and BMI eQTS were significantly enriched for pathways related to mitochondrial function (Q<10-7), with the most significant T2D signal found in the heart atrial appendage and BMI signals in subcutaneous (SC) and visceral adipose. For T2D, pancreas eQTS were enriched for pancreatic beta cell development and mature onset diabetes of the young (MODY) (Q<10-3). For CAD, cholesterol metabolism pathways were enriched in the coronary artery (Q<10-7). In summary, T2D and BMI eQTS share mitochondrial and metabolic transcription mechanisms across several tissues, including SC adipose. Cholesterol metabolism was enriched in the coronary artery in CAD eQTS. This study provides novel insight into the shared genetic and transcriptomic architecture of T2D, CAD, and obesity. Disclosure C. Bryan: None. K. Smith: None. H. Dashti: None. M. Claussnitzer: None. A. Manning: None. J.M. Mercader: None. M. Udler: Other Relationship; Up-To-Date. Y. Huang: None. Funding Doris Duke Foundation

Introduction & Objective: Diabetes is a heterogeneous disease, with patients displaying varying degrees of adiposity, beta-cell dysfunction, and insulin resistance. Ahlqvist et al. established five distinct sub-types of adult-onset diabetes, with differing disease progression and risk for complications, based on clustering of six clinical variables. These clusters were developed and validated in adult populations, but their reproducibility in pediatric diabetes remains unknown. We sought to determine if pediatric patients can be sub-classified into clusters based on the same variables, and if these clusters had differing responses to treatment. Methods: We analyzed Progress in Diabetes Genetics in Youth (ProDiGY), a multi-ethnic resource that brings together youth with clinician-diagnosed type 2 diabetes (T2D) before age 20 years from SEARCH (n=428) and the Treatment Options for T2D in Adolescents and Youth (TODAY) study (n=455). In participants with negative autoantibodies, we used a k-means (k=3) clustering method, with variables BMI, HOMA2_B, HOMA2_IR, and HbA1c; we excluded age of diagnosis given its limited range in a pediatric population. Results: We identified three distinct pediatric clusters in each cohort-sex group. Cluster 1 had high HbA1c and low HOMA2_B, denoting low beta-cell function. Cluster 2 had the highest BMI, HOMA2_IR, and HOMA2_B, denoting insulin resistance. Cluster 3 had intermediate levels of HOMA2_IR and HOMA2_B but the lowest HbA1c, denoting milder diabetes. In TODAY, participants in cluster 1 had the highest rates of treatment failure (76% versus 46% overall, p<0.0001). Metformin alone was most likely to be successful in cluster 3 (failure rate 19%) and had the highest rates of failure in cluster 1 (failure rate 86%, p<0.0001). Conclusion: Clustering pediatric T2D by clinical measures reveals distinct sub-types with differing responses to treatment. Clustering might be used to inform treatment choices in pediatric T2D patients. Disclosure R.J. Kreienkamp: None. K. Smith: None. T.Y. Wangden: None. E.T. Jensen: Advisory Panel; TARGET PharmaSolutions, Inc. Research Support; TARGET PharmaSolutions, Inc. Consultant; Regeneron Pharmaceuticals Inc. A.S. Shah: None. C. Pihoker: None. L. Szczerbinski: None. J.C. Florez: Research Support; Novo Nordisk. Other Relationship; Novo Nordisk, AstraZeneca. M. Udler: Other Relationship; Up-To-Date. S. Srinivasan: None. Funding American Diabetes Association (11-22-PDFPM-03); RJK is supported by NIH (T32DK007699). JCF is supported by NIH/NHLBI (K24 HL157960). SS is supported by NIH (K23DK120932 and R03DK138213).

We previously used a soft clustering approach to group type 2 diabetes (T2D)-associated single nucleotide variants (SNVs) based on their associations with phenotypes. In previous work, we demonstrated associations between the resulting five clusters and various clinical phenotypes, outcomes, and tissues. To further understand the mechanistic pathways tied to these potential T2D subtypes, here we incorporate genome-scale metabolic models (GEMs). Cluster-specific partitioned polygenic risk scores (pPS) were generated for subjects in the GTEx database. Regression analyses of the pPS and gene expression levels identified up- and down-regulated genes for each cluster. These genes were used as constraints on a generic human GEM, thereby creating cluster-specific models. We also created individualized models for GTEx subjects, followed by statistical testing of fluxes between the top and bottom pPS decile groups of each cluster. Over 120 reactions were significantly altered in each cluster (Wilcoxon, p<0.05), with none significant in more than two clusters. In all clusters, the majority of hits were transport reactions. However, some subsystems were more heavily altered in specific clusters, including cholesterol biosynthesis and omega fatty acid metabolism in the beta-cell and lipodystrophy clusters, respectively. Of the 145 subsystems cataloged in our GEM, 39 contained a reaction that was associated with only one cluster, such as ubiquinone synthesis for the liver-lipid cluster (p=6e-4). By creating metabolic reconstructions of our T2D clusters, we have gained a systems-level understanding of the clusters’ unique characteristics. Our results pinpoint cluster-specific metabolic reactions with differential flux activity and may therefore potentially be useful for future precision medicine efforts. Disclosure K. Smith: None. A. Eames: None. M. Sevilla-Gonzalez: Research Support; Novo Nordisk Foundation. M. Udler: Other Relationship; Up-To-Date. Funding Doris Duke Foundation Award 2022063

Type 2 diabetes (T2D) is a multifactorial disease with substantial genetic risk, for which the underlying biological mechanisms are not fully understood. In this study, we identified multi-ancestry T2D genetic clusters by analyzing genetic data from diverse populations in 37 published T2D genome-wide association studies representing more than 1.4 million individuals. We implemented soft clustering with 650 T2D-associated genetic variants and 110 T2D-related traits, capturing known and novel T2D clusters with distinct cardiometabolic trait associations across two independent biobanks representing diverse genetic ancestral populations (African, n = 21,906; Admixed American, n = 14,410; East Asian, n =2,422; European, n = 90,093; and South Asian, n = 1,262). The 12 genetic clusters were enriched for specific single-cell regulatory regions. Several of the polygenic scores derived from the clusters differed in distribution among ancestry groups, including a significantly higher proportion of lipodystrophy-related polygenic risk in East Asian ancestry. T2D risk was equivalent at a body mass index (BMI) of 30 kg m-2 in the European subpopulation and 24.2 (22.9-25.5) kg m-2 in the East Asian subpopulation; after adjusting for cluster-specific genetic risk, the equivalent BMI threshold increased to 28.5 (27.1-30.0) kg m-2 in the East Asian group. Thus, these multi-ancestry T2D genetic clusters encompass a broader range of biological mechanisms and provide preliminary insights to explain ancestry-associated differences in T2D risk profiles.

CONTEXT:Polycystic ovary syndrome (PCOS) is a heterogeneous disorder, with disease loci identified from genome-wide association studies (GWAS) having largely unknown relationships to disease pathogenesis. OBJECTIVE:This work aimed to group PCOS GWAS loci into genetic clusters associated with disease pathophysiology. METHODS:Cluster analysis was performed for 60 PCOS-associated genetic variants and 49 traits using GWAS summary statistics. Cluster-specific PCOS partitioned polygenic scores (pPS) were generated and tested for association with clinical phenotypes in the Mass General Brigham Biobank (MGBB, N = 62 252). Associations with clinical outcomes (type 2 diabetes [T2D], coronary artery disease [CAD], and female reproductive traits) were assessed using both GWAS-based pPS (DIAMANTE, N = 898,130, CARDIOGRAM/UKBB, N = 547 261) and individual-level pPS in MGBB. RESULTS:Four PCOS genetic clusters were identified with top loci indicated as following: (i) cluster 1/obesity/insulin resistance (FTO); (ii) cluster 2/hormonal/menstrual cycle changes (FSHB); (iii) cluster 3/blood markers/inflammation (ATXN2/SH2B3); (iv) cluster 4/metabolic changes (MAF, SLC38A11). Cluster pPS were associated with distinct clinical traits: Cluster 1 with increased body mass index (P = 6.6 × 10-29); cluster 2 with increased age of menarche (P = 1.5 × 10-4); cluster 3 with multiple decreased blood markers, including mean platelet volume (P = 3.1 ×10-5); and cluster 4 with increased alkaline phosphatase (P = .007). PCOS genetic clusters GWAS-pPSs were also associated with disease outcomes: cluster 1 pPS with increased T2D (odds ratio [OR] 1.07; P = 7.3 × 10-50), with replication in MGBB all participants (OR 1.09, P = 2.7 × 10-7) and females only (OR 1.11, 4.8 × 10-5). CONCLUSION:Distinct genetic backgrounds in individuals with PCOS may underlie clinical heterogeneity and disease outcomes.

Previous studies have identified type 2 diabetes (T2D) subtypes using clinical phenotypes and/or genetic information. This study aimed to identify molecular signatures associated with genetically based T2D subtypes. We analyzed data from ~100,000 unrelated European UK Biobank participants without T2D and not taking lipid-lowering medications. T2D subtypes were characterized using five polygenic scores denoting three forms of insulin resistance (lipodystrophy, obesity, and impaired lipid/hepatic metabolism) and two subtypes of insulin secretion (beta-cell dysfunction and impaired proinsulin synthesis). Nuclear magnetic resonance profiles of 249 plasma metabolites on the Nightingale platform were tested for association with T2D subtypes using linear regression adjusted for age, sex, batch, and ancestry-derived principal components; P<0.0002 was considered significant. Independent validation was conducted in two independent cohorts, including 720 from the ANTORCHA cohort. In general, the T2D subtypes driven by insulin resistance were characterized by alterations in lipid and lipoproteins, while the impaired insulin secretion subtypes showed aberrant plasma amino acid levels. The lipodystrophy cluster showed the most distinct molecular signature, with higher concentrations of extremely large, very large, and large VLDL particles which were enriched by triglycerides and phospholipids, and lower concentrations of HDL. The beta-cell dysfunction cluster was characterized by higher concentrations of amino acids (alanine, valine, leucine, and tyrosine). These associations were broadly consistent after adjusting for BMI, and the pattern of alterations in lipid metabolites was partially replicated in the ANTORCHA cohort. We identified specific metabolic signatures related to T2D genetic clusters before T2D onset, offering further characterization of the putative underlying pathways. Disclosure M.Sevilla: None. I.Lamiquiz-moneo: None. K.Smith: None. M.Canyelles: None. J.C.Florez: Consultant; AstraZeneca, Novo Nordisk, Other Relationship; AstraZeneca, Merck & Co., Inc. A.Manning: None. F.Civeira: None. J.Merino: None. M.Udler: None. Funding American Diabetes Association (9-22-PDFPM-04 to M.S.)