Publications
A combined polygenic score of 21,293 rare and 22 common variants improves diabetes diagnosis based on hemoglobin A1C levels.
Dornbos P, et al.
Nat Genet. 2022 Oct 24. doi: 10.1038/s41588-022-01200-1.
Exome sequencing of 20,791 cases of type 2 diabetes and 24,440 controls.
Flannick J, et al.
Nature. 2019 May 22. doi: 10.1038/s41586-019-1231-2
Sequence data and association statistics from 12,940 type 2 diabetes cases and controls.
Flannick J, Fuchsberger C, Mahajan A, et al.
Sci Data. 2017 Dec 19;4:170179. doi: 10.1038/sdata.2017.179
The genetic architecture of type 2 diabetes.
Fuchsberger C, Flannick J, Teslovich TM, Mahajan A, Agarwala V, Gaulton KJ, et al.
Nature 2016 Aug 4;536(7614):41-7. doi: 10.1038/nature18642
Association of a low-frequency variant in HNF1A with type 2 diabetes in a Latino population.
SIGMA Type 2 Diabetes Consortium, et al.
JAMA. 2014 Jun 11;311(22):2305-14. doi: 10.1001/jama.2014.6511
Whole-exome sequencing of 2,000 Danish individuals and the role of rare coding variants in type 2 diabetes.
Lohmueller KE, et al.
Am J Hum Genet. 2013 Dec 5;93(6):1072-86. doi: 10.1016/j.ajhg.2013.11.005
Phenotypes
- Adiponectin
- BMI
- Diastolic blood pressure
- Fasting C-peptide
- Fasting glucose
- Fasting insulin
- HbA1c
- HDL cholesterol
- Height
- Hip circumference
- HOMA-B
- HOMA-IR
- LDL cholesterol
- Leptin
- Serum creatinine
- Systolic blood pressure
- Two-hour glucose
- Two-hour insulin
- Total cholesterol
- Triglycerides
- Triglyceride-to-HDL ratio
- Waist circumference
- Waist-hip ratio
Subjects
| Project |
Cases |
Controls |
Cohort (Click to view selection criteria for cases and controls) |
Ancestry |
| ESP |
389 |
2,842 |
Exome Sequencing Project (ESP)
| Case selection criteria |
Control selection criteria |
- ARIC: Self report of Type 2 Diabetes
- CARDIA: Self report of Type 2 Diabetes
- CHS: ADA definition
- Framingham: Self report of Type 2 Diabetes
- JHS: Type 2 Diabetes Status defined using ADA 2004
- MESA: Exam 1 Diabetes mellitus using 2003 ADA
- WHI: Self report of Type 2 Diabetes
|
|
|
European |
| GoT2D |
1,289 |
1,251 |
Multiple cohorts
Finland-United States Investigation of NIDDM Genetics (FUSION) Study
| Case selection criteria |
Control selection criteria |
- Unrelated cases selected from FUSION families and stage 2 replication
- Samples met 1999 World Health Organization (WHO) criteria of fasting plasma glucose ≥ 7.0 mmol/l or postload glucose during an OGTT ≥ 11.1 mmol/l, by report of diabetes medication use, or based on medical record review
- Prioritized FUSION families with ≥ 2 first-degree relatives with T2D; BMI ≥ 18.5kg/m2; case with GWAS data or earliest age at onset, if no GWAS data available
- Prioritized FUSION stage 2 replication set with Metabochip data; BMI ≥ 18.5kg/m2; earliest age of onset; age of onset ≥ 35
|
- Unrelated controls with normal glucose tolerance (NGT) based on WHO (1999) definitions: fasting plasma glucose <6.1 mM and 2 hour postload glucose during an OGTT > 7.8 mM
- Frequency matched to cases by birth province; BMI ≥ 18.5kg/m2; age ≤ 80
- Within each birth province, prioritized samples from stage 2 replication with highest values for age + 2*BMI
|
Malmö-Botnia Study
| Case selection criteria |
Control selection criteria |
- A liability score was generated (Guey LT et al. 2011) which measures risk to T2D in the context of three known risk factors (age at onset, BMI, and gender) in 27,500 individuals drawn from three prospective cohorts: the Malmö Preventive Project, the Scania Diabetes Registry, and the Botnia Study; only BMI and gender used to construct scores for Scania and Botnia studies
- Eligible cases limited to individuals between 35 and 60 years of age and with a BMI between 20 and 35
- To match for ethnicity, 250 Botnia cases with the most extreme liability scores were selected, while 125 cases were selected from each of the Scania and Malmö studies
|
- Controls selected from the extreme of a liability score distribution, based upon gender, age and BMI at last follow-up visit; only BMI and gender used to construct scores for Malmö study
- Eligible controls limited to individuals above 35 years of age at follow-up and with a BMI between 20 and 40
- To match for ethnicity, equal numbers of controls were selected from the Botnia and Malmö studies
|
UK Type 2 Diabetes Genetics Consortium (UKT2D)
| Case selection criteria |
Control selection criteria |
- Cases drawn from the Wellcome Trust Case Control Consortium (WTCCC)
- Female samples with age of diagnosis ≥ 66 years or BMI ≥ 32kg/m2 excluded; male samples with age of diagnosis ≥ 62 years or BMI ≥ 31kg/m2 excluded
- Remaining samples were ranked by age and BMI, and the two ranks multiplied. 356 samples with the lowest values for this rank multiplier were selected for initial inclusion in the study
|
- Unrelated samples selected as controls from the Twins UK study
- A twin pair was considered for selection if there was no recorded family history of diabetes, neither twin was ever recorded as impaired glucose tolerant (defined as fasting glucose >6.1mmol/L in any reading), there were available quantitative trait and genetic (GWAs) data, and no evidence of admixture in MDS analysis of GWAs data
- From set of qualifying twin pairs, the best control twin was selected from each pair with the lowest ratio of fasting glucose level to BMI across all readings, and further prioritization of the qualifying unrelated samples involved selecting samples that had the lowest fasting glucose to (BMI * age) ratios
- Top two principal components were used to perform pairwise sample matching between cases and possible controls, and the best control for each case was selected
|
KORAgen Study Helmholtz zentrum München (KORA)
| Case selection criteria |
Control selection criteria |
- Samples drawn from KORA F3 and F4
- Diabetic status validated by doctor or by medication use
- Cases have ≥ 1 first degree relative with type 2 diabetes (self-reported)
- Cases have either BMI ≤ 30 and age of onset < 65, or BMI ≤ 33 and age of onset ≤ 60
|
- Controls selected from KORA F4
- All controls are normal glucose tolerant: fasting glucose level < 6.1 mmol/l and two hour glucose level after oral glucose tolerance test < 7.8 mmol/l
- Controls are either > 60 years of age with BMI > 32 or over 65 years of age with BMI > 31
|
|
European |
| T2D-GENES |
478 |
498 |
Metabolic Syndrome in Men Study (METSIM)
| Case selection criteria |
Control selection criteria |
- Previous diagnosis of T2D, or both fasting and 2-hr criteria met for new T2D diagnosis
- C-peptide > 0.10 nmol/L
- Anti-GAD antibody < 50 U/mL to rule out T1D
- Family history of diabetes (parents, sibs, children, grandparents, avuncular, cousins)
- Unrelated individuals based on family ID and IBS analyses
- Preferentially select individuals with with genotype data (N=494), as well as non-genotyped individuals with earlier possible age of diagnosis (N=26)
|
- Normal glucose tolerance at baseline and follow-up visits
- Prioritized samples with no family history of diabetes and meeting strict NGT criteria: fasting glucose < 5.6 mmol/l and 2 hour post-challenge glucose < 7.8 mmol/l
- Additional samples selected with fasting glucose < 6.1 mmol/l and 2 hour post-challenge glucose < 7.8 mmol/l
- Unrelated samples
- Older controls preferentially selected
|
|
European |
| T2D-GENES |
506 |
360 |
Ashkenazi
| Case selection criteria |
Control selection criteria |
- Ashkenazi Jewish origin, defined as having all four grandparents born in Northern or Eastern Europe; subjects with known or suspected Sephardic Jewish or non-Jewish ancestry excluded
- T2D defined according to the World Health Organisation criteria (fasting glucose > 140 mg/dl on two or more occasions or random glucose > 200 mg/dl)
- To avoid late-onset T1D, patients who became insulin-dependent within 2 years of diagnosis excluded; anti-GAD or anti-islet cell antibody titers not routinely measured
- T2D cases were selected from two separate DNA collections:
- Genome-wide, affected-sibling-pair linkage study (Permutt et al. Diabetes 2001). Families in which both parents were known to have diabetes were excluded. One affected individual selected from each family and, wherever possible, sibling with youngest age of diagnosis selected.
- Study to determine genetic risk for diabetic complications (Blech et al. PLoS One 2011). Patients ascertained by the Israel Diabetes Research Group between 2002 and 2004 from 15 diabetes clinics throughout Israel. Primary selection criteria: (1) known T1D or T2D for more than 10 years, (2) 4 grandparents being either Ashkenazi or Sephardic-North African Jewish. For this study, only T2D patients with all 4 grandparents of Ashkenazi Jewish origin and age of diagnosis between 35 and 60 were selected.
|
- Fasting blood glucose < 7 mmol/l
- No personal history of diabetes
- No anti-diabetic medications
|
|
European |
| LuCAMP |
992 |
985 |
Lundbeck Foundation Centre for Applied Medical Genomics in Personalised Disease Prediction, Prevention, and Care
| Case selection criteria |
Control selection criteria |
- diagnosis of type 2 diabetes
- BMI > 27.5 kg/m2
- hypertension (systolic/diastolic BP > 140/90 mmHg or use of antihypertensive medication)
|
- fasting plasma glucose < 5.6 mmol/l
- 2 h post-OGTT plasma glucose < 7.8 mmol/l
- BMI < 27.5 kg/m2
- blood pressure < 140/90 mmHg
|
|
European |
| T2D-GENES |
949 |
943 |
Genetics of Diabetes and Audit Research Tayside Study (GoDARTS)
| Case selection criteria |
Control selection criteria |
|
|
|
|
European |
| T2D-GENES |
390 |
589 |
Framingham Heart Study (FHS)
| Case selection criteria |
Control selection criteria |
- For the Original offspring, used the diabetes file to define T2D (on medication or (non-fasting) blood glucose > 200)
|
|
|
European |
Projects
Exome Sequencing project (ESP) Learn more >
The Exome Sequencing Project (ESP) aims to discover novel genes and mechanisms contributing to heart, lung and blood disorders by pioneering the application of next-generation sequencing of the protein coding regions of the human genome across diverse, richly-phenotyped populations and to share these datasets and findings with the scientific community to extend and enrich the diagnosis, management and treatment of heart, lung and blood disorders.
Genetics of Type 2 Diabetes (GoT2D) Learn more >
The GoT2D consortium aims to understand the allelic architecture of type 2 diabetes through whole-genome sequencing, high-density SNP genotyping, and imputation. The reference panel based on this work is intended as a comprehensive inventory of low-frequency variants in Europeans, including SNPs, small insertions and deletions, and structural variants.
Lubeck Foundation Centre for Applied Medical Genomics in Personalised Disease Prediction, Prevention and Care (LuCamp) Learn more >
The LuCamp research consortium aims to discover and characterize novel variation in the human genome conferring an increased risk of visceral obesity, type 2 diabetes and hypertension and eventually premature cardiovascular disorders and death; discover and characterize novel variation in the human gut microbiome influencing metabolic and cardiovascular health; investigate how the novel molecular signatures interact mutually and with other risk markers, particularly in health behavior, influencing the risk of developing widespread metabolic disorders; and investigate how these discoveries may help improve metabolic and cardiovascular health in the at-risk population.
Type 2 Diabetes Genetic Exploration by Next-Generation Sequencing in Ethnic Samples (T2D-GENES) Learn more >
T2D-GENES (Type 2 Diabetes Genetic Exploration by Next-generation sequencing in multi-Ethnic Samples) is a large collaborative effort to find genetic variants that influence risk of type 2 diabetes. With funding from NIDDK, the group is pursuing three projects: (1) deep whole-exome sequencing in 10,000 people from five ethnicities (African-American, East Asian, South Asian, European, and Hispanic); (2) deep whole-genome sequencing of 600 individuals selected from extended Mexican American pedigrees; and (3) a trans-ethnic fine-mapping "mega-meta-analysis."
Overview of analysis and results
The T2D analysis reported in Flannick et al. 2019 was extended to 24 quantitative traits. Briefly, single-variant association analyses were conducted for each trait (n=24) at the sample sub-group level (n=25). We analyzed samples using the EMMAX test via the EPACTS software package. A genomic relationship matrix (GRM) was included to account for relatedness across samples. In all cases, multiallelic sites were categorized as a single "non-reference" allele.
For each of the 25 sub-studies × 24 traits (n=600) single-variant analyses, quantile-quantile (QQ)-plots of all and common (MAC>20) alleles were used to assess whether the regression models were well-calibrated (i.e., comparison of regression results to expected results under a null-model). Stringencies of the variant filters were increased based on digression from the null model. Variants were filtered based on sequencing quality or excessive heterozygosity. Covariates to adjust for subject ascertainment and/or differing sequencing technologies across sub-studies were added where appropriate. Phenotype-specific outliers were removed where appropriate. Following all sub-study level quality control, a 25-group fixed-effect inverse-variance weighted meta-analysis was then conducted using METAL.
