Genome-wide analyses using UK Biobank data provide insights into the genetic architecture of osteoarthritis

Abstract

Osteoarthritis is a common complex disease imposing a large public-health burden. Here, we performed a genome-wide association study for osteoarthritis, using data across 16.5 million variants from the UK Biobank resource. After performing replication and meta-analysis in up to 30,727 cases and 297,191 controls, we identified nine new osteoarthritis loci, in all of which the most likely causal variant was noncoding. For three loci, we detected association with biologically relevant radiographic endophenotypes, and in five signals we identified genes that were differentially expressed in degraded compared with intact articular cartilage from patients with osteoarthritis. We established causal effects on osteoarthritis for higher body mass index but not for triglyceride levels or genetic predisposition to type 2 diabetes.

Access options

Rent or Buy article

Get time limited or full article access on ReadCube.

from$8.99

All prices are NET prices.

Fig. 1: Power to detect association in the discovery stage.
Fig. 2: Regional association plots for the nine novel osteoarthritis loci.
Fig. 3: Heat map of genetic correlations between osteoarthritis phenotypes in UK Biobank and 35 traits grouped in ten categories from GWAS consortia.
Fig. 4: Two-sample MR estimates and 95% CI values of the effects of obesity-related measures, triglyceride levels, years of schooling and type 2 diabetes liability on different definitions of osteoarthritis.

References

  1. 1.

    Loeser, R. F., Goldring, S. R., Scanzello, C. R. & Goldring, M. B. Osteoarthritis: a disease of the joint as an organ. Arthritis Rheum. 64, 1697–1707 (2012).

    PubMed  PubMed Central  Google Scholar 

  2. 2.

    Felson, D. T. et al. Osteoarthritis: new insights. Part 1: the disease and its risk factors. Ann. Intern. Med. 133, 635–646 (2000).

    CAS  PubMed  Google Scholar 

  3. 3.

    Cibrián Uhalte, E., Wilkinson, J. M., Southam, L. & Zeggini, E. Pathways to understanding the genomic aetiology of osteoarthritis. Hum. Mol. Genet. 26, R193–R201 (2017).

    Google Scholar 

  4. 4.

    Ma, C., Blackwell, T., Boehnke, M. & Scott, L. J., GoT2D investigators. Recommended joint and meta-analysis strategies for case-control association testing of single low-count variants. Genet. Epidemiol. 37, 539–550 (2013).

    PubMed  PubMed Central  Google Scholar 

  5. 5.

    Broome, D. T. & Datta, N. S. Mitogen-activated protein kinase phosphatase-1: function and regulation in bone and related tissues. Connect. Tissue Res. 57, 175–189 (2016).

    CAS  PubMed  Google Scholar 

  6. 6.

    Rodríguez-Carballo, E., Gámez, B. & Ventura, F. p38 MAPK signaling in osteoblast differentiation. Front. Cell Dev. Biol. 4, 40 (2016).

    PubMed  PubMed Central  Google Scholar 

  7. 7.

    Wei, L., Sun, X. J., Wang, Z. & Chen, Q. CD95-induced osteoarthritic chondrocyte apoptosis and necrosis: dependency on p38 mitogen-activated protein kinase. Arthritis Res. Ther. 8, R37 (2006).

    PubMed  PubMed Central  Google Scholar 

  8. 8.

    Wang, Q. et al. Bioinformatics analysis of gene expression profiles of osteoarthritis. Acta Histochem. 117, 40–46 (2015).

    CAS  PubMed  Google Scholar 

  9. 9.

    Prasadam, I. et al. Osteoarthritic cartilage chondrocytes alter subchondral bone osteoblast differentiation via MAPK signalling pathway involving ERK1/2. Bone 46, 226–235 (2010).

    CAS  PubMed  Google Scholar 

  10. 10.

    Prasadam, I. et al. Inhibition of p38 pathway leads to OA-like changes in a rat animal model. Rheumatology (Oxford) 51, 813–823 (2012).

    CAS  Google Scholar 

  11. 11.

    Prasadam, I. et al. ERK-1/2 and p38 in the regulation of hypertrophic changes of normal articular cartilage chondrocytes induced by osteoarthritic subchondral osteoblasts. Arthritis Rheum. 62, 1349–1360 (2010).

    CAS  PubMed  Google Scholar 

  12. 12.

    Zhang, Y., Pizzute, T. & Pei, M. A review of crosstalk between MAPK and Wnt signals and its impact on cartilage regeneration. Cell Tissue Res. 358, 633–649 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  13. 13.

    Namdari, S., Wei, L., Moore, D. & Chen, Q. Reduced limb length and worsened osteoarthritis in adult mice after genetic inhibition of p38 MAP kinase activity in cartilage. Arthritis Rheum. 58, 3520–3529 (2008).

    CAS  PubMed  PubMed Central  Google Scholar 

  14. 14.

    Zhang, R., Murakami, S., Coustry, F., Wang, Y. & de Crombrugghe, B. Constitutive activation of MKK6 in chondrocytes of transgenic mice inhibits proliferation and delays endochondral bone formation. Proc. Natl Acad. Sci. USA 103, 365–370 (2006).

    CAS  PubMed  Google Scholar 

  15. 15.

    Castañón, M. J., Walko, G., Winter, L. & Wiche, G. Plectin-intermediate filament partnership in skin, skeletal muscle, and peripheral nerve. Histochem. Cell Biol. 140, 33–53 (2013).

    PubMed  PubMed Central  Google Scholar 

  16. 16.

    Tseng, K. Y., Chen, Y. H. & Lin, S. Zinc finger protein ZFP36L1 promotes osteoblastic differentiation but represses adipogenic differentiation of mouse multipotent cells. Oncotarget 8, 20588–20601 (2017).

    PubMed  PubMed Central  Google Scholar 

  17. 17.

    Castaño-Betancourt, M. C. et al. Novel genetic variants for cartilage thickness and hip osteoarthritis. PLoS Genet. 12, e1006260 (2016).

    PubMed  PubMed Central  Google Scholar 

  18. 18.

    Usmani, S. E. et al. Transforming growth factor alpha controls the transition from hypertrophic cartilage to bone during endochondral bone growth. Bone 51, 131–141 (2012).

    CAS  PubMed  Google Scholar 

  19. 19.

    Appleton, C. T., Usmani, S. E., Bernier, S. M., Aigner, T. & Beier, F. Transforming growth factor alpha suppression of articular chondrocyte phenotype and Sox9 expression in a rat model of osteoarthritis. Arthritis Rheum. 56, 3693–3705 (2007).

    PubMed  Google Scholar 

  20. 20.

    Appleton, C. T., Usmani, S. E., Mort, J. S. & Beier, F. Rho/ROCK and MEK/ERK activation by transforming growth factor-alpha induces articular cartilage degradation. Lab. Invest. 90, 20–30 (2010).

    CAS  PubMed  Google Scholar 

  21. 21.

    Usmani, S. E. et al. Context-specific protection of TGFα null mice from osteoarthritis. Sci. Rep. 6, 30434 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  22. 22.

    Miettinen, P. J. et al. Epidermal growth factor receptor function is necessary for normal craniofacial development and palate closure. Nat. Genet. 22, 69–73 (1999).

    CAS  PubMed  Google Scholar 

  23. 23.

    Chu, J. et al. A mouse forward genetics screen identifies LISTERIN as an E3 ubiquitin ligase involved in neurodegeneration. Proc. Natl Acad. Sci. USA 106, 2097–2103 (2009).

    CAS  PubMed  Google Scholar 

  24. 24.

    Wang, M. & Casey, P. J. Protein prenylation: unique fats make their mark on biology. Nat. Rev. Mol. Cell Biol. 17, 110–122 (2016).

    CAS  Google Scholar 

  25. 25.

    Gonzales, P. A. et al. Large-scale proteomics and phosphoproteomics of urinary exosomes. J. Am. Soc. Nephrol. 20, 363–379 (2009).

    CAS  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Palmieri, A. et al. Comparison between osteoblasts derived from human dental pulp stem cells and osteosarcoma cell lines. Cell Biol. Int. 32, 733–738 (2008).

    CAS  PubMed  Google Scholar 

  27. 27.

    Ochiai, S. et al. Oxidative stress reaction in the meniscus of Bach 1 deficient mice: potential prevention of meniscal degeneration. J. Orthop. Res. 26, 894–898 (2008).

    PubMed  Google Scholar 

  28. 28.

    Takada, T. et al. Bach1 deficiency reduces severity of osteoarthritis through upregulation of heme oxygenase-1. Arthritis Res. Ther. 17, 285 (2015).

    PubMed  PubMed Central  Google Scholar 

  29. 29.

    Capellini, T. D. et al. Ancient selection for derived alleles at a GDF5 enhancer influencing human growth and osteoarthritis risk. Nat. Genet. 49, 1202–1210 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  30. 30.

    Daans, M., Luyten, F. P. & Lories, R. J. GDF5 deficiency in mice is associated with instability-driven joint damage, gait and subchondral bone changes. Ann. Rheum. Dis. 70, 208–213 (2011).

    PubMed  Google Scholar 

  31. 31.

    Miyamoto, Y. et al. A functional polymorphism in the 5′ UTR of GDF5 is associated with susceptibility to osteoarthritis. Nat. Genet. 39, 529–533 (2007).

    CAS  PubMed  Google Scholar 

  32. 32.

    Southam, L. et al. An SNP in the 5′-UTR of GDF5 is associated with osteoarthritis susceptibility in Europeans and with in vivo differences in allelic expression in articular cartilage. Hum. Mol. Genet. 16, 2226–2232 (2007).

    CAS  PubMed  Google Scholar 

  33. 33.

    Davey Smith, G. & Ebrahim, S. ‘Mendelian randomization’: can genetic epidemiology contribute to understanding environmental determinants of disease? Int. J. Epidemiol. 32, 1–22 (2003).

    PubMed  Google Scholar 

  34. 34.

    Prieto-Alhambra, D. et al. An increased rate of falling leads to a rise in fracture risk in postmenopausal women with self-reported osteoarthritis: a prospective multinational cohort study (GLOW). Ann. Rheum. Dis. 72, 911–917 (2013).

    PubMed  Google Scholar 

  35. 35.

    Baldwin, J. N. et al. Self-reported knee pain and disability among healthy individuals: reference data and factors associated with the knee injury and osteoarthritis outcome score (KOOS) and KOOS-Child. Osteoarthritis Cartilage 25, 1282–1290 (2017).

    CAS  PubMed  Google Scholar 

  36. 36.

    Callahan, L. F. et al. Associations of educational attainment, occupation and community poverty with knee osteoarthritis in the Johnston County (North Carolina) osteoarthritis project. Arthritis Res. Ther. 13, R169 (2011).

    PubMed  PubMed Central  Google Scholar 

  37. 37.

    Hussain, S. M. et al. How are obesity and body composition related to patellar cartilage? a systematic review. J. Rheumatol. 44, 1071–1082 (2017).

    PubMed  Google Scholar 

  38. 38.

    Fewell, Z., Davey Smith, G. & Sterne, J. A. The impact of residual and unmeasured confounding in epidemiologic studies: a simulation study. Am. J. Epidemiol. 166, 646–655 (2007).

    PubMed  Google Scholar 

  39. 39.

    Marchini, J., Howie, B., Myers, S., McVean, G. & Donnelly, P. A new multipoint method for genome-wide association studies by imputation of genotypes. Nat. Genet. 39, 906–913 (2007).

    CAS  PubMed  Google Scholar 

  40. 40.

    Willer, C. J., Li, Y. & Abecasis, G. R. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics 26, 2190–2191 (2010).

    CAS  PubMed  PubMed Central  Google Scholar 

  41. 41.

    Franklin, J., Ingvarsson, T., Englund, M. & Lohmander, S. Association between occupation and knee and hip replacement due to osteoarthritis: a case-control study. Arthritis Res. Ther. 12, R102 (2010).

    PubMed  PubMed Central  Google Scholar 

  42. 42.

    Styrkarsdottir, U. et al. Whole-genome sequencing identifies rare genotypes in COMP and CHADL associated with high risk of hip osteoarthritis. Nat. Genet. 49, 801–805 (2017).

    CAS  PubMed  Google Scholar 

  43. 43.

    Wesseling, J. et al. CHECK (Cohort Hip and Cohort Knee): similarities and differences with the Osteoarthritis Initiative. Ann. Rheum. Dis. 68, 1413–1419 (2009).

    CAS  PubMed  Google Scholar 

  44. 44.

    Steinberg, J. et al. Integrative epigenomics, transcriptomics and proteomics of patient chondrocytes reveal genes and pathways involved in osteoarthritis. Sci. Rep. 7, 8935 (2017).

    PubMed  PubMed Central  Google Scholar 

  45. 45.

    Ritchie, M. E. et al. limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Res. 43, e47 (2015).

    PubMed  PubMed Central  Google Scholar 

  46. 46.

    Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).

    PubMed  PubMed Central  Google Scholar 

  47. 47.

    Tischler, G. & Leonard, S. biobambam: tools for read pair collation based algorithms on BAM files. Source Code Biol. Med. 9, 13 (2014).

    PubMed Central  Google Scholar 

  48. 48.

    Patro, R., Duggal, G., Love, M. I., Irizarry, R. A. & Kingsford, C. Salmon provides fast and bias-aware quantification of transcript expression. Nat. Methods 14, 417–419 (2017).

    CAS  PubMed  PubMed Central  Google Scholar 

  49. 49.

    Law, C. W., Chen, Y., Shi, W. & Smyth, G. K. voom: precision weights unlock linear model analysis tools for RNA-seq read counts. Genome Biol. 15, R29 (2014).

    PubMed  PubMed Central  Google Scholar 

  50. 50.

    UK10K Consortium. et al. The UK10K project identifies rare variants in health and disease. Nature 526, 82–90 (2015).

    Google Scholar 

  51. 51.

    1000 Genomes Project Consortium. et al. A global reference for human genetic variation. Nature 526, 68–74 (2015).

    Google Scholar 

  52. 52.

    Chen, W. et al. Fine mapping causal variants with an approximate Bayesian method using marginal test statistics. Genetics 200, 719–736 (2015).

    PubMed  PubMed Central  Google Scholar 

  53. 53.

    Kircher, M. et al. A general framework for estimating the relative pathogenicity of human genetic variants. Nat. Genet. 46, 310–315 (2014).

    CAS  PubMed  PubMed Central  Google Scholar 

  54. 54.

    Ionita-Laza, I., McCallum, K., Xu, B. & Buxbaum, J. D. A spectral approach integrating functional genomic annotations for coding and noncoding variants. Nat. Genet. 48, 214–220 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  55. 55.

    Iotchkova, V. et al. Discovery and refinement of genetic loci associated with cardiometabolic risk using dense imputation maps. Nat. Genet. 48, 1303–1312 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  56. 56.

    Zheng, J. et al. LD Hub: a centralized database and web interface to perform LD score regression that maximizes the potential of summary level GWAS data for SNP heritability and genetic correlation analysis. Bioinformatics 33, 272–279 (2017).

    CAS  PubMed  Google Scholar 

  57. 57.

    Hemani, G. et al. MR-Base: a platform for systematic causal inference across the phenome using billions of genetic associations. Preprint at https://www.biorxiv.org/content/early/2016/12/16/078972 (2016).

  58. 58.

    Barban, N. et al. Genome-wide analysis identifies 12 loci influencing human reproductive behavior. Nat. Genet. 48, 1462–1472 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

This research was conducted by using the UK Biobank Resource under application no. 9979. This work was funded by the Wellcome Trust (WT098051). We are grateful to R. Brooks, J. Choudhary and T. Roumeliotis for their contribution to the functional genomics data collection. The Human Research Tissue Bank is supported by the NIHR Cambridge Biomedical Research Centre. The Rotterdam Study is funded by Erasmus Medical Center and Erasmus University, Organization for Health Research and Development (ZonMw); the Netherlands Genomics Initiative (NGI)/Netherlands Organisation of Scientific Research (NWO); the Netherlands Consortium for Healthy Aging (NCHA), Research Institute for Diseases in the Elderly (RIDE); the Ministry of Education, Culture and Science; the Ministry for Health, Welfare and Sports; the European Commission (DG XII); and the Municipality of Rotterdam.

Author information

Affiliations

Authors

Contributions

Association analyses: E. Zengini, K.H., I.T., L.S., J.S., S.H. and A.G. Mendelian randomization: F.P.H. and G.D.S. Functional genomics sample collection: A.McC., J.M.W. and E. Zeggini. Functional genomics analyses: J.S. and L.S. Endophenotype analyses: C.G.B., A.G.U. and J.B.J.v.M. Replication analyses: U.S., T.I., H.J., U.T. and K.S. Bioinformatics: A.G., D.S. and B.K. Student supervision: K.H., G.C.B., G.D.S., J.M.W. and E. Zeggini. Manuscript writing: E. Zengini, K.H., I.T., J.S., F.P.H., L.S., C.G.B., U.S., D.S., J.B.J.v.M., G.D.S., J.M.W. and E. Zeggini.

Corresponding author

Correspondence to Eleftheria Zeggini.

Ethics declarations

Competing interests

U.T., U.S. and K.S. are employees of deCODE genetics/Amgen.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Integrated supplementary information

Supplementary Figure 1

Study-design flowchart. OA: Osteoarthritis; UKBB: UK Biobank; n: sample size; SNP: single nucleotide polymorphism; QC: quality control; 1000G: 1000 Genomes Project; AF: allele frequency; PC: principal component.

Supplementary Figure 2 Manhattan plots

a) Self-reported OA; b) Hospital diagnosed OA; c) Hospital diagnosed hip OA; d) Hospital diagnosed knee OA; e) Hospital diagnosed hip and/or knee OA; the horizontal line denotes the genome-wide significance threshold. OA: osteoarthritis.

Supplementary Figure 3 QQ plots

a) Self-reported OA; b) Hospital diagnosed OA; c) Hospital diagnosed hip OA; d) Hospital diagnosed knee OA; e) Hospital diagnosed hip and/or knee OA. OA: Osteoarthritis; N: number of variants analysed following quality control; MAF: minor allele frequency; Lambda: genomic inflation factor.

Supplementary Figure 4 Genome-wide-association summary statistics of the self-reported and sensitivity analysis for self-reported OA

a) qq plot for self-reported OA; b) Manhattan plot for self-reported OA; c) qq plot for the sensitivity analysis of self-reported OA; d) Manhattan plot for the sensitivity analysis of self-reported OA. OA: Osteoarthritis; N: number of variants analysed following quality control; MAF: minor allele frequency; Lambda: genomic inflation factor; the horizontal line denotes the genome-wide significance threshold.

Supplementary Figure 5 Correlation across annotations and corresponding correlations across PPAs in an exemplary fine-mapped region (indexed by rs2820436)

a: Histogram of all annotations (CADD, CADD-Phred, Eigen, Eigen-Phred, EigenPC, EigenPC-Phred, Combined CADD-Phred and Eigen-Phred) and scatter plots between pairs of annotations. b: Pearson correlation coefficient of the posterior probability of association (PPA) across variants in the fine-mapped region for the different annotations used.

Supplementary Figure 6 Gene-set representation onto protein–protein-interaction networks

Large coloured circles represent the proteins of each gene set, while grey small circles represent proteins found in the InWeb database (high-confidence interaction set) that have a direct (proteins that directly bind each other) or indirect (any other proteins connected to the direct network) connectivity. a) Anatomical structure morphogenesis b) Ion channel transport c) Activation of MAPK activity d) Histidine metabolism e) Recruitment of mitotic centrosome proteins and complexes.

Supplementary Figure 7 LD-regression Venn diagram

Using LD regression, we identify 23 phenotypes with significant genetic correlation with self-reported OA, 31 with hospital diagnosed OA, 26 with hospital diagnosed hip and/or knee OA, 23 with hospital diagnosed knee OA, and 4 with hospital diagnosed hip OA, with large overlap between the identified phenotypes (35 phenotypes in total; significance at 5% FDR). OA: osteoarthritis.

Supplementary Figure 8 ORs and risk-allele frequencies for established and novel OA-associated variants

The sibling relative risk ratio (λs) for each variant is shown by the different size of each circle. The darker color of the circles shows the percentage of variance explained on the liability scale (h2L) (calculated assuming 10% prevalence of OA). Totals of λs and h2L are shown at the top. OA: osteoarthritis.

Supplementary information

Supplementary Text and Figures

Supplementary Figures 1–8, Supplementary Note and Supplementary Tables 1, 4–6, 8–10, 19–21 and 28–30

Life Sciences Reporting Summary

Supplementary Table 27: Results of one-sample MR analyses in UK Biobank

Regression coefficients are shown for continuous exposure phenotypes, and odds ratio are shown for binary exposure phenotypes. MR: Mendelian randomization; OA: osteoarthritis; OR: odds ratio; SE: standard error; CI: confidence interval; N: sample size

Supplementary Table 2

ICD10 code exclusion criteria for the definition of controls

Supplementary Table 3

ICD10 code inclusion criteria for the definition of cases

Supplementary Table 7: Association summary statistics and power calculations for established OA loci present in UK Biobank

RA, risk allele; RAF, risk allele frequency; OR, odds ratio; CI, confidence interval; OA, osteoarthritis; EA, effect allele; EAF, effect allele frequency; N/A, not available a: variant studied separately in Asian and European populations b: RA is defined as the allele with OR >1 in the respective study c: Allele on reverse strand d: Power is calculated on the basis of the EAF and OR estimates from the UK Biobank OA datasets and assuming 4 controls per case e: Power is calculated on the basis of the reported OR estimates from each discovery study (referenced in the last column) and effect allele frequencies from the UK Biobank OA datasets respectively. For rs143383 we used the OR from the European population. f: Power is calculated on the basis of the replication OR estimates from each discovery study and effect allele frequencies from the UK Biobank OA datasets, respectively. For rs11842874 we used the OR from the "arcOGEN replication set1" replication dataset. For rs12901071 we used the relevant OR from the deCODE replication dataset. For rs143383 we used the relevant OR from the European population and particularly the OR from the "Ten studies from large-scale meta-analysis" replication cohort. For rs7639618 we used the OR from the Japanese replication dataset. g: The number of cases required for 80% power is calculated on the basis of EAF and OR estimates from the UK Biobank OA datasets and assuming 4 controls per case

Supplementary Table 11: Association summary statistics for the 173 prioritised SNPs in the discovery, replication and overall meta-analysis for self-reported OA

OA: Osteoarthritis; EA: Effect allele; EAF: Effect allele frequency; OR: Odds ratio; PV: p-value; chr: chromosome

Supplementary Table 12: Association summary statistics for the 173 prioritised SNPs in the discovery, replication and overall meta-analysis for hospital diagnosed OA

OA: osteoarthritis; EA: Effect allele; EAF: Effect allele frequency; OR: Odds ratio; PV: p-value; chr: chromosome

Supplementary Table 13: Association summary statistics for the 173 prioritised SNPs in the discovery, replication and overall meta-analysis for hospital diagnosed knee and/or hip OA

OA: osteoarthritis; EA: Effect allele; EAF: Effect allele frequency; OR: Odds ratio; PV: p-value; chr: chromosome

Supplementary Table 14: Association summary statistics for the 173 prioritised SNPs in the discovery, replication and overall meta-analysis for hospital diagnosed hip OA

OA: osteoarthritis; EA: Effect allele; EAF: Effect allele frequency; OR: Odds ratio; PV: p-value; chr: chromosome

Supplementary Table 15: Association summary statistics for the 173 prioritised SNPs in the discovery, replication and overall meta-analysis for hospital diagnosed knee OA

OA: osteoarthritis; EA: Effect allele; EAF: Effect allele frequency; OR: Odds ratio; PV: p-value; chr: chromosome

Supplementary Table 16: eQTL analysis in GTEx

Significant eQTL associations (at 5% FDR) for the 9 osteoarthritis signals as provided by the GTEx portal

Supplementary Table 17: Detailed fine-mapping results for the OA-associated loci

For each variant (columns A-D), we report its association summary statistics in UK Biobank (G-K), its predicted consequence (L) and the nearest gene (M), as well as the region centred around the variant used for the fine mapping (N-O). For each region, we report the Bayes’ factor (F) and the posterior probability that there be at least 1 causal variant in the region (E), followed by the variants (P) in the 95% credible intervals (defined to contain the minimum set of variants that jointly have at least 95% probability of including the causal variant). Summary statistics (S-W), Bayes’ factors (Q) and the posterior probabilities of association (PPA, column R) of each variant in the credible intervals are also reported. For regions in which the sum of probabilities of causality of all variants in the fine-mapped region is less than 0.95, we report the lower confidence interval achieved. EA: effect allele; NEA: non-effect allele; EAF: affect allele frequency; SE: standard error; N: sample size; BF: Bayes factor; PPA: posterior probability of association; OA: osteoarthritis

Supplementary Table 18: Fine-mapping results for the OA-associated loci

For each index variant (columns A-D), we report its predicted consequence (E) and the nearest protein coding gene (F). For each region constructed around the index variant, we report the Bayes’ factor (H) and the posterior probability that there be at least 1 causal variant in the region (G). Column I reports the number of variants in the resulting 95% credible intervals (defined to contain the minimum set of variants that jointly have at least 95% probability of including the causal variant). For each region, the number of variants in the credible interval with posterior probability >0.1, and their combined probability are listed in column J and K, respectively. Properties (L-Q, T-X), summary statistics (Y-AA), Bayes’ factors (R) and the posterior probabilities of association (S) of the top variants in the credible intervals are also reported. Only variants with probability of causality >0.1 are shown, or the top variant if its probability of causality is

Supplementary Table 22: Description of studies used to identify genetic instruments for MR analyses

MR: Mendelian randomization. aNot available in MR-Base for this record, so this value was extracted from another GWAS of the same phenotype (PMID 22504420) available in MR-Base

Supplementary Table 23: Summary association results used in the two-sample MR analyses

MR: Mendelian randomization; EA: effect allele; NEA: non-effect allele; EAF: effect allele frequency; SE: standard error; OA: osteoarthritis

Supplementary Table 24: Summary statistics of each set of genetic instruments used in the two-sample MR analyses

MR: Mendelian randomization; OA: osteoarthritis

Supplementary Table 25: Causal OR estimates of OA for each exposure phenotype from two-sample MR analyses

For exposure phenotypes with at least 10 genetic instruments, all methods were applied. For exposure phenotypes with less than 10 genetic instruments, only the inverse-variance weighting method was applied to calculate a pooled causal effect estimate, and only the heterogeneity test was used to assess horizontal pleiotropy. For exposure phenotypes with only one genetic instrument, the ratio method was used to calculate the causal effect estimate. bStandard error of the ln(OR). ϕ: tunning parameter of the mode-based estimate. The smaller the value of ϕ, the more stringent the analysis is. OR: odds ratio; OA: osteoarthritis; MR: Mendelian randomization; SE: standard error; CI: confidence interval

Supplementary Table 26: Power (%) to detect different OR of OA in MR analyses using the inverse-variance weighting method

OR: odds ratio; OA: osteoarthritis; MR: Mendelian randomization

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Zengini, E., Hatzikotoulas, K., Tachmazidou, I. et al. Genome-wide analyses using UK Biobank data provide insights into the genetic architecture of osteoarthritis. Nat Genet 50, 549–558 (2018). https://doi.org/10.1038/s41588-018-0079-y

Download citation

Further reading

Search

Quick links

Sign up for the Nature Briefing newsletter for a daily update on COVID-19 science.
Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing