Longevity Variant Database (LVDB)
Authors:Kristen Fortney [a] [b] †, Daniel Wuttke [c] †, Eric Morgen [d], Jessica Choi [a] [e], Navneet Ramesh [a] [f], Dmitri Shytikov [a], Nick Schaum [g], John M. Adams [a], Barry Goldberg [a], Amir Levine [a], Maria Litovchenko [a], Aiste Narkeviciute [a], Emily Quist [a], Jimi Vyas [a], Justin Rebo [h] [i], Joe Betts-Lacroix [a]
[a](1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12) Health Extension Research Foundation, Mountain View, CA, 94043, USA
[b]Department of Developmental Biology, Stanford University, Stanford, CA, 94305, USA
[c]Institute of Integrative Biology, University of Liverpool, Liverpool L69 7ZB, UK
[d]Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON, M5S 1A1, Canada
[e]University of California, Berkeley, CA, USA
[f]UCLA Longevity Center, University of California, Los Angeles, Los Angeles, CA, USA
[g]Lawrence Berkeley National Laboratory, Life Sciences Division, Berkeley, CA 94720, USA.
[h]SENS Research Foundation, Mountain View, CA, 94041, USA
[i]Department of Bioengineering, University of California at Berkeley, Berkeley, CA, 94720, USA
¬To whom correspondence should be addressed. Tel: +415 794 4323; Fax: +415 794 4323; Email: joe@healthextension.co
†The authors wish it to be known that, in their opinion, the first 2 authors should be regarded as joint First Authors.
ABSTRACT
INTRODUCTION
DATABASE CONSTRUCTION AND DESIGN
Genetics data curation
Database construction
USING THE DATABASE
Browsing LVDB
Example use cases
DISCUSSION
ACKNOWLEDGEMENT
FUNDING
REFERENCES
TABLE AND FIGURES LEGENDS
The Longevity Variant Database (LVDB) is a collaborative effort to catalogue all genetic variants that have been studied for their relation to human longevity. We have systematically identified and curated all genome-wide association, candidate gene, and candidate chromosome region studies conducted on long-lived human populations from the scientific literature. We have included both those studies linking new variants to human longevity, as well as those refuting putative longevity variants identified in previous works. In total, our database currently stores over 1000 distinct genetic variant associations curated from roughly 300 publications. We anticipate that LVDB will be a useful resource for unraveling and deciphering the complex genetics of human longevity and age-related disease.
Human longevity is heritable, especially at the extremes. For the general population, the genetic component of aging lies in the range 0.25-0.33, while the heritability of being a centenarian - living to age 100 or older - is estimated to be higher at 0.33-0.48 [1]. And extended lifespan is often accompanied by extended healthspan. Relative to age-matched controls, centenarian offspring show significantly delayed onset of multiple age-related diseases, including heart disease and diabetes [2] [3]. Longevity is strongly related to the underlying rate of aging and presumably to factors affecting age-related diseases; age is the major risk factor for some of the largest causes of adult morbidity and mortality in developed countries [4].
It follows that determining the genetic factors underlying human longevity will help us to better understand the mechanism of and ultimately treat a whole suite of degenerative age-related diseases. There are multiple examples from the literature showing the promise of this approach. Consider the case of Apoliprotein E, perhaps the single gene most consistently associated to human longevity in both genome-wide association studies (GWAS) and candidate studies [5]. APOE is highly pleiotropic and APOE alleles have also been associated with incidence of Alzheimer's disease [6], macular degeneration [6], and coronary heart disease [7]. Similarly, variants in insulin/IGF-1 pathway genes including AKT, FOXO3A, and IGF-1R associate with human longevity [8] [9], and also appear to be involved in the development of diabetes, again demonstrating the close relationship between longevity and age-related diseases. Understanding the specific functional effects of genetic variants that associate with longevity may provide new drug targets for age-related diseases. A variant in CETP that results in reduced protein function has been linked to exceptional longevity in Ashkenazi Jewish populations [10]. Pharmaceutical companies are developing CETP inhibitors to mimic the beneficial effects of this mutation on late-life cognitive function and risk of Alzheimer's disease [4].
Consequently, it is crucial to identify the genetic factors that underlie human longevity. The literature in this field is large and growing quickly. In the past few decades, hundreds of candidate gene studies have tested individual genes, variants, and regions for their association with human longevity. In more recent years, GWAS have begun to provide unbiased surveys of the entire genome, though to date their sample sizes have been too small to discover many significant effects [11] [5]. For researchers, keeping track of all these discoveries is very challenging. Further complicating interpretation, results from different studies of the same variant are sometimes inconsistent [9], and many findings have not yet been replicated in large, independent populations [5] [12] [13].
We built the Longevity Variant Database (LVDB;) to help researchers navigate the large and complex literature on the genetics of human longevity. LVDB unifies all existing knowledge on genetic variants that have been tested for their association with human longevity. LVDB’s detailed records store both the positive and negative evidence for several hundred putative human longevity genes, annotated with genetic and functional information to facilitate cross-study comparisons. We anticipate that LVDB will help the research community to prioritize future genetic as well as functional work to determine the effects of these beneficial or dementrial mutations.
Several disease-specific variant databases have been created to curate and aggregate relevant genomic variants, including for example AlzGene [14] and AutismKB [15], but no comparable resource exists for aging. Existing aging-related databases (HAGR [16], Lifespan Observations Database [http://lifespandb.sageweb.org/]) focus on curating genetic manipulations and small molecule treatments in model organisms, and have not systematically catalogued the large number of genetic variants associated with exceptional human longevity.
Paper selection. We designed a Boolean expression for identification of scientific articles with relevance to genetic variants of longevity in humans. Specifically, we retrieved all PubMed abstracts ([http://www.ncbi.nlm.nih.gov/pubmed/]) returned with the query:
“(longevity[Title/Abstract] OR lifespan[Title/Abstract] OR centenarian[Title/Abstract] OR aging[Title/Abstract]) AND (snp[Title/Abstract] OR snps[Title/Abstract] OR variant[Title/Abstract] OR polymorphism[Title/Abstract] OR allele*[Title/Abstract] OR locus[Title/Abstract] OR loci[Title/Abstract])”
These 3957 abstracts were assessed for their relevance to LVDB, and 300 papers were selected for curation. Positive and negative studies were included.
Paper curation. In each article we focused on curating those genetic associations considered by study authors to be their main results. In candidate studies, this may have been a new association of a variant with longevity, the confirmation of a previously associated variant, or the failure of a previously identified variant to replicate in a new population. In genome-wide studies, this was all variants reported by the authors to be genome-wide significant or nominally significant (due to small sample size, GWAS studies to date have only reported variants linked to APOE as genome-wide significant; in LVDB we have included nominally significant results from these studies, as they are likely enriched for true associations [17] [18]. In the database, we coded all variant associations as “Significant”, “Nominally significant”, or “Not significant”. Where available, for each reported association we curated all relevant data including: population size and ethnicity, age of long-lived cases, variant type [single nulcleotide polymorphism (SNP), haplotype, etc.] and ID, odds ratio, P value, short- and long-lived allele, associated gene name, study type, genotyping technology, and whether the result was replicated in an independent population. For each variant association we also report whether the authors judged the result to be positive or negative. Any additional relevant study-specific data were recorded in an informal free-form description field.
Data annotation. Variant identifiers, gene symbols and names, and other relevant data were reported in a variety of different formats in different publications. To standardize data across studies, we mapped all gene names to Entrez Gene identifiers [19] and SNPs to rsids (using Ensembl variation [20] and dbSNP [19]). We queried online databases to supplement the raw data provided in the individual publications with functional annotations, including Gene Ontology terms [21], and protein structure and function descriptions from Uniprot [22]. For each gene, we have integrated data across studies into a gene-specific summary page that collates information on all tests conducted on the genes’ variants. Similarly, for each article LVDB provides a study-specific summary page listening all investigated variants and genes. Link outs to relevant gene and functional annotation databases are provided for each polymorphism entry.
Web interface. The LVDB was implemented with a relational database in Denigma which is written in Python based on Django and runs on the Amazon Web Service. List and detail views were designed in HTML/JavaScript for all relevant models. The main list of variant associations allows sorting, filtering and querying. The LVDB has full create, read, update and delete as well as reversion control support. Users are encouraged to submit comments, updates, or correction through links from the front page. LVDB will be updated continously and at least biannually with data from the all latest publications.
LVDB is user-friendly, with flexible searching and browsing capabilities.
From the LVDB front page, key results for variant-publication pairs are displayed as rows of data in a table. Columns include polymorphism ID, associated gene, odds ratio, P value, size of population cohort, ethnicity, age of long-lived cases, shorter- and longer-lived alleles, study type, and PMID. As a default, rows are sorted in order of P value (the most significant findings are displayed at the top of the page); users click on any column heading to sort by its values instead. Users can access full results for any row (variant-publication pair) by clicking on the polymorphism ID. Clicking on a gene name will load a summary page integrating all results for that gene; similarly, clicking on a PMID will load a paper-specific page including a study description and all variant association results.
On the front page, LVDB has a Search field and a Term field. Through the Search field, LVDB can be searched any variant-associated data such as polymorphism ID, gene name, ethnicity, study type, genotyping technology, etc. Through the Term field, LVDB can be searched by Gene Ontology (21) terms to return results for genes sharing a common function, process, or location (as well as biological pathway). Clicking on buttons next to the front-page chromosome images will restrict results to those genomic locations. The Export button allows the results of any search to be exported as a tab-delimited text file.
Example 1: Nominally significant candidate SNPs from all longevity GWAS. Longevity GWAS to date have been underpowered to detect many associations at genome-wide significance; nominally significant SNPs from these studies may be enriched for true positives and should be prioritized for further study in replication cohorts [23] [18]. Querying LVDB for GWAS results yields 611 SNPs considered genome-wide significant or nominally significant in at least one GWAS study, meta-analysis, or data re-analysis [12] [24] [25] [26] [27] [28] [29] [18] [30]. While genome-wide significant hits rs2075650 and rs4420638 are both linked to APOE, there are suggestive signals at many other locations in the genome (Figure 1A).
Example 2: Significant genes in the insulin signaling pathway. Mutations in genes related to insulin signaling can modify lifespan in many species, including worms and mice [31] [32]. For this reason insulin signaling has been heavily studied as a candidate pathway in humans, and there is compelling evidence for its contribution from both candidate gene [33] and GWAS studies [34]. Querying LVDB for the Gene Ontology term GO:0008286, the identifier for "insulin receptor signaling pathway", retrieves 9 genes with variant associations that have been linked to human longevity (Figure 1B).
Example 3: Browsing results for CETP. LVDB provides gene summary pages that collate information on all related variants that have been tested for their association with human longevity. Clicking on the gene name CETP loads the CETP summary page (Figure 2), which provides basic information about protein structure and function, displays variant IDs, PMIDs, and P values for all tested variants, and links to more detailed information for each finding.
Example 4: Analysing longevity-associated genes Ontology terms can be used to understand the meaning of our genome and for instance identify shared functionalities and properties of genes in humans that are singificantly associated with longevity (Table 1).
The Longevity Variants Database is the first and most comprehensive resource on the genetics of human longevity. There is a great potential for future development and applications especially in regard with functional consequences of individual variants and the interactions between different variants. With the ever increasing sequencing capacity and associated falling cost, personal medicine will benefit from the knoweldge on genetic variants that determine one individuals lifespan. Deciphering the genetics of human aging of pivote importance to improve the health of mankind in the long term. Eventually also genetic variants of individuals that are stuck in development and do not appear to age might help to decode the basic aging process and solve the enigma of aging.
The authors thank the team members of the Global Computing Initiative.
Funding for open access charge: Health Extension Research Foundation.
[1]Sebastiani,P. and Perls,T.T. (2012) The genetics of extreme longevity: lessons from the New England Centenarian study. Front. Genet. Aging, 3, 277.
[2]Terry,D.F., Wilcox,M.A., McCormick,M.A., Pennington,J.Y., Schoenhofen,E.A., Andersen,S.L. and Perls,T.T. (2004) Lower All-Cause, Cardiovascular, and Cancer Mortality in Centenarians’ Offspring. J. Am. Geriatr. Soc., 52, 2074–2076.
[3]Adams,E.R., Nolan,V.G., Andersen,S.L., Perls,T.T. and Terry,D.F. (2008) Centenarian Offspring: Start Healthier and Stay Healthier. J. Am. Geriatr. Soc., 56, 2089–2092.
[4](1, 2) Harman,D. (1991) The aging process: major risk factor for disease and death. Proc. Natl. Acad. Sci., 88, 5360–5363.
[5](1, 2, 3) Newman,A.B. and Murabito,J.M. (2013) The Epidemiology of Longevity and Exceptional Survival. Epidemiol. Rev., 10.1093/epirev/mxs013.
[6](1, 2) McKay,G.J., Patterson,C.C., Chakravarthy,U., Dasari,S., Klaver,C.C., Vingerling,J.R., Ho,L., de Jong,P.T.V.M., Fletcher,A.E., Young,I.S., et al. (2011) Evidence of association of APOE with age-related macular degeneration - a pooled analysis of 15 studies. Hum. Mutat., 32, 1407–1416.
[7]Bennet,A.M., Di Angelantonio,E., Ye,Z., Wensley,F., Dahlin,A., Ahlbom,A., Keavney,B., Collins,R., Wiman,B., de Faire,U., et al. (2007) Association of apolipoprotein E genotypes with lipid levels and coronary risk. JAMA J. Am. Med. Assoc., 298, 1300–1311.
[8]Kenyon,C.J. (2010) The genetics of ageing. Nature, 464, 504–512.
[9](1, 2) Heijmans,B.T., Westendorp,R.G. and Slagboom,P.E. (2000) Common gene variants, mortality and extreme longevity in humans. Exp. Gerontol., 35, 865–877.
[10]Barzilai,N., Atzmon,G., Schechter,C., Schaefer,E.J., Cupples,A.L., Lipton,R., Cheng,S. and Shuldiner,A.R. (2003) Unique lipoprotein phenotype and genotype associated with exceptional longevity. JAMA J. Am. Med. Assoc., 290, 2030–2040.
[11]Deelen,J., Beekman,M., Capri,M., Franceschi,C. and Slagboom,P.E. (2013) Identifying the genomic determinants of aging and longevity in human population studies: Progress and challenges. BioEssays, 35, 386–396.
[12](1, 2) Beekman,M., Blanché,H., Perola,M., Hervonen,A., Bezrukov,V., Sikora,E., Flachsbart,F., Christiansen,L., De Craen,A.J.M., Kirkwood,T.B.L., et al. (2013) Genome-wide linkage analysis for human longevity: Genetics of Healthy Aging Study. Aging Cell, 12, 184–193.
[13]Tazearslan,C., Cho,M. and Suh,Y. (2012) Discovery of Functional Gene Variants Associated With Human Longevity: Opportunities and Challenges. J. Gerontol. A. Biol. Sci. Med. Sci., 67A, 376–383.
[14]Bertram,L., McQueen,M.B., Mullin,K., Blacker,D. and Tanzi,R.E. (2007) Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nat. Genet., 39, 17–23.
[15]Xu,L.-M., Li,J.-R., Huang,Y., Zhao,M., Tang,X. and Wei,L. (2012) AutismKB: an evidence-based knowledgebase of autism genetics. Nucleic Acids Res., 40, D1016–1022.
[16]Tacutu,R., Craig,T., Budovsky,A., Wuttke,D., Lehmann,G., Taranukha,D., Costa,J., Fraifeld,V.E. and de Magalhães,J.P. (2013) Human Ageing Genomic Resources: integrated databases and tools for the biology and genetics of ageing. Nucleic Acids Res., 41, D1027–1033.
[17]Li,M.J., Wang,P., Liu,X., Lim,E.L., Wang,Z., Yeager,M., Wong,M.P., Sham,P.C., Chanock,S.J. and Wang,J. (2011) GWASdb: a database for human genetic variants identified by genome-wide association studies. Nucleic Acids Res., 10.1093/nar/gkr1182.
[18](1, 2, 3) Sebastiani,P., Solovieff,N., Dewan,A.T., Walsh,K.M., Puca,A., Hartley,S.W., Melista,E., Andersen,S., Dworkis,D.A., Wilk,J.B., et al. (2012) Genetic signatures of exceptional longevity in humans. PloS One, 7, e29848.
[19](1, 2) NCBI Resource Coordinators (2013) Database resources of the National Center for Biotechnology Information. Nucleic Acids Res., 41, D8–D20.
[20]Flicek,P., Ahmed,I., Amode,M.R., Barrell,D., Beal,K., Brent,S., Carvalho-Silva,D., Clapham,P., Coates,G., Fairley,S., et al. (2013) Ensembl 2013. Nucleic Acids Res., 41, D48–55.
[21]Gene Ontology Consortium (2013) Gene Ontology annotations and resources. Nucleic Acids Res., 41, D530–535.
[22]UniProt Consortium (2013) Update on activities at the Universal Protein Resource (UniProt) in 2013. Nucleic Acids Res., 41, D43–47.
[23]Christensen,K., Johnson,T.E. and Vaupel,J.W. (2006) The quest for genetic determinants of human longevity: challenges and insights. Nat. Rev. Genet., 7, 436–448.
[24]Shi,H., Belbin,O., Medway,C., Brown,K., Kalsheker,N., Carrasquillo,M., Proitsi,P., Powell,J., Lovestone,S., Goate,A., et al. (2012) Genetic variants influencing human aging from late-onset Alzheimer’s disease (LOAD) genome-wide association studies (GWAS). Neurobiol. Aging, 33, 1849.e5–18.
[25]Walter,S., Atzmon,G., Demerath,E.W., Garcia,M.E., Kaplan,R.C., Kumari,M., Lunetta,K.L., Milaneschi,Y., Tanaka,T., Tranah,G.J., et al. (2011) A genome-wide association study of aging. Neurobiol. Aging, 32, 2109.e15–28.
[26]Malovini,A., Illario,M., Iaccarino,G., Villa,F., Ferrario,A., Roncarati,R., Anselmi,C.V., Novelli,V., Cipolletta,E., Leggiero,E., et al. (2011) Association study on long-living individuals from Southern Italy identifies rs10491334 in the CAMKIV gene that regulates survival proteins. Rejuvenation Res., 14, 283–291.
[27]Lunetta,K.L., D’Agostino,R.B.,Sr, Karasik,D., Benjamin,E.J., Guo,C.-Y., Govindaraju,R., Kiel,D.P., Kelly-Hayes,M., Massaro,J.M., Pencina,M.J., et al. (2007) Genetic correlates of longevity and selected age-related phenotypes: a genome-wide association study in the Framingham Study. BMC Med. Genet., 8 Suppl 1, S13.
[28]Deelen,J., Beekman,M., Uh,H.-W., Helmer,Q., Kuningas,M., Christiansen,L., Kremer,D., van der Breggen,R., Suchiman,H.E.D., Lakenberg,N., et al. (2011) Genome-wide association study identifies a single major locus contributing to survival into old age; the APOE locus revisited. Aging Cell, 10, 686–698.
[29]Yashin,A.I., Wu,D., Arbeev,K.G. and Ukraintseva,S.V. (2012) Polygenic effects of common single-nucleotide polymorphisms on life span: when association meets causality. Rejuvenation Res., 15, 381–394.
[30]Nebel,A., Kleindorp,R., Caliebe,A., Nothnagel,M., Blanché,H., Junge,O., Wittig,M., Ellinghaus,D., Flachsbart,F., Wichmann,H.-E., et al. (2011) A genome-wide association study confirms APOE as the major gene influencing survival in long-lived individuals. Mech. Ageing Dev., 132, 324–330.
[31]Barbieri,M., Bonafè,M., Franceschi,C. and Paolisso,G. (2003) Insulin/IGF-I-signaling pathway: an evolutionarily conserved mechanism of longevity from yeast to humans. Am. J. Physiol. Endocrinol. Metab., 285, E1064–1071.
[32]Kenyon,C. (2011) The first long-lived mutants: discovery of the insulin/IGF-1 pathway for ageing. Philos. Trans. R. Soc. B Biol. Sci., 366, 9–16.
[33]Willcox,B.J., Donlon,T.A., He,Q., Chen,R., Grove,J.S., Yano,K., Masaki,K.H., Willcox,D.C., Rodriguez,B. and Curb,J.D. (2008) FOXO3A genotype is strongly associated with human longevity. Proc. Natl. Acad. Sci., 105, 13987–13992.
[34]Deelen,J., Uh,H.-W., Monajemi,R., van Heemst,D., Thijssen,P.E., Böhringer,S., van den Akker,E.B., de Craen,A.J.M., Rivadeneira,F., Uitterlinden,A.G., et al. (2013) Gene set analysis of GWAS data for human longevity highlights the relevance of the insulin/IGF-1 signaling and telomere maintenance pathways. Age Dordr. Neth., 35, 235–249.
[35]Franceschini,A., Szklarczyk,D., Frankild,S., Kuhn,M., Simonovic,M., Roth,A., Lin,J., Minguez,P., Bork,P., von Mering,C., et al. (2012) STRING v9.1: protein-protein interaction networks, with increased coverage and integration. Nucleic Acids Res., 41, D808–D815.
Figure 1. LVDB examples. A. Manhattan plot of nominally significant findings from published longevity GWAS. LDVB retrieved 611 distinct nominally significant SNPs from published GWAS, with maximum P value 0.01. The red line indicates a genome-wide significant cut-off (P https://github.com/stephenturner/qqman]). B. Network of genes from the insulin signaling pathway with one or more variants associated to longevity; lines indicate functional interactions (interaction data and image generated with STRING [35]).
Figure 2. LVDB gene summary page. The LVDB summary page for CETP lists all tested variants, P values, and PMIDs, as well as basic gene information.
Table 1: LVDB functional enrichment of longevity-associated genes. LVDB contains build-in analysis tools that allow to utilize ontologies. Genes that are assoicated via genetic variants to longevity are enriched for specific ontological terms.
Term listHits Percent foldEnrichment FDR
disease mutation 89 22.36 2.78 8.66E-16
signal 119 29.90 1.82 2.37E-08
mutagenesis site 87 21.86 2.10 3.97E-08
signal peptide 119 29.90 1.81 4.17E-08
sequence variant 303 76.13 1.25 4.70E-08
disulfide bond 110 27.64 1.87 4.06E-08
plasma 17 4.27 9.09 6.83E-08
glycoprotein 143 35.93 1.65 1.19E-07
disulfide bond 106 26.63 1.86 1.94E-07
polymorphism 290 72.86 1.25 6.43E-07
transmembrane protein 39 9.80 3.02 4.05E-06
hsa05332:Graft-versus-host disease 14 3.52 8.45 3.53E-06
Secreted 71 17.84 2.09 5.38E-06
lymphokine 8 2.01 28.40 5.44E-06
hsa04940:Type I diabetes mellitus 14 3.52 7.85 9.91E-06
dna repair 20 5.03 5.20 1.58E-05
glycosylation site:N-linked (GlcNAc...) 131 32.91 1.57 3.67E-05
DNA damage 20 5.03 4.85 5.02E-05
hsa05330:Allograft rejection 12 3.02 7.85 1.69E-04
pharmaceutical 10 2.51 11.30 2.64E-04
liver 11 2.76 9.27 3.56E-04
h_longevityPathway:The IGF-1 Receptor and Longevity 8 2.01 10.35 2.15E-03
immunoregulation 5 1.26 41.42 3.34E-03
active site:Proton acceptor 33 8.29 2.48 7.66E-03
atherosclerosis 5 1.26 27.61 2.68E-02
lipid binding 6 1.51 16.57 3.16E-02
obesity 7 1.76 11.60 3.58E-02
h_inflamPathway:Cytokines and Inflammatory Response 10 2.51 5.47 3.95E-02
