gm333

7/31/2019 gm333

1/7

Complex diseases: omics and genome-wideassociation studies

Common, severe human diseases such as cancer,diabetes, asthma, or mental and cardiovascular disordershave complex etiologies and complex mechanisms. o

uncover the causal events leading to these diseases,in ormation on the actors that challenge human healthand the immediate responses to these challenges isneeded. Yet, un ortunately, the dataset is never complete.In most cases, studies o humans are restricted toobservations a ter a disease has occurred, except inclinical cases when individuals with particular diseasesare treated or take part in randomized controlledintervention trials. Outside clinical trials, longitudinalstudies (observational studies tracking the same individ-uals) that analyze phenotypes can also be undertaken.Both o these types o studies are hampered by unknownand uncontrolled exposure to the environment (such asdi erences in nutrition, medication, environmental endo-crine disruptors and li estyle) even in well-phenotypedcohorts (where weight, height and health status, orexample, are known).

Cohorts can be analyzed or speci c eatures such asgenomic variance (variants in the DNA sequence) ormetric parameters (concentrations or comparative levels)o RNA, proteins or metabolites. I the eatures analyzedand disease phenotypes coincide (and the requency o coincidence is biostatistically valid), then it would bepossible to identi y the pathways involved. Tere ore, acurrent approach to unveiling the etiology and mecha-nism o complex diseases is to employ sophisticatedanalysis methodologies (omics) that allow or the inte-gration o multiple layers o molecular and organismaldata. Data acquired with omics have already contributedconsiderably to the understanding o homeostasis inhealth and disease. Genome-wide association studies(GWAS), in particular, have contributed substantially to

the eld in the past 6 years [1]. Tis approach hasidenti ed numerous genetic loci that are associated withcomplex diseases. However, the number o geneticmechanisms that have been identi ed to explain complexdiseases has not increased signi cantly [2].

In this review, I will highlight the current limitations o GWAS and how issues such as the large sample sizerequired can be overcome by adding metabolomicsin ormation to these studies. I will explain the principlesbehind the combination o metabolomics and GWAS(mGWAS) and how together they can provide a more

AbstractGenome-wide association studies (GWAS) analyze thegenetic component o a phenotype or the etiologyo a disease. Despite the success o many GWAS,little progress has been made in uncovering theunderlying mechanisms or many diseases. The use o metabolomics as a readout o molecular phenotypeshas enabled the discovery o previously undetectedassociations between diseases and signaling andmetabolic pathways. In addition, combining GWASand metabolomic in ormation allows the simultaneousanalysis o the genetic and environmental impacts onhomeostasis. Most success has been seen in metabolicdiseases such as diabetes, obesity and dyslipidemia.Recently, associations between loci such as FADS1 ,ELOVL2 or SLC16A9 and lipid concentrations have beenexplained by GWAS with metabolomics. CombiningGWAS with metabolomics (mGWAS) provides therobust and quantitative in ormation required or thedevelopment o specifc diagnostics and targeteddrugs. This review discusses the limitations o GWASand presents examples o how metabolomicscan overcome these limitations with the ocus onmetabolic diseases.Keywords complex disease, genetics, metabolomics,metabolic traits, pathway in erence

2010 BioMed Central Ltd

Genome-wide association studies withmetabolomicsJerzy Adamski1,2*

REVIEW

*Correspondence: [email protected] Zentrum Mnchen, German Research Center or EnvironmentalHealth, Institute o Experimental Genetics, Genome Analysis Center, 85764Neuherberg, GermanyFull list o author in ormation is available at the end o the article

Adamski Genome Medicine 2012, 4:34http://genomemedicine.com/content/4/4/34

2012 BioMed Central Ltd

7/31/2019 gm333

2/7

power ul analysis. I conclude by exploring how mGWAShas been used to identi y the metabolic pathwaysinvolved in metabolic diseases.

Aims and limitations o GWAS

GWAS analyze the association between common genetic variants and speci c traits (phenotypes). Te phenotypesoriginally included weight (or body mass index), height,blood pressure or requency o a disease. More recently,speci c traits in the transcriptome, proteome or metabo-lome have been included, and these are usually quanti-tative ( or example, concentration). GWAS can also beused to explore whether common DNA variants areassociated with complex diseases ( or example, cancer ortype 2 diabetes mellitus). Te common variants might besingle nucleotide polymorphisms (SNPs), copy numberpolymorphisms (CNPs), insertions/deletions (indels) orcopy number variations (CNVs), but most GWAS employ SNPs [3]. At present, SNPs are used most requently because o coverage o a large raction o genome, through-put o assay, quality assurance and cost-e ective ness.Because the concept o GWAS is hypothesis- ree, theanalyses o GWAS are generally genetically unbiased, butthey assume a genetic cause that might not be the mostsigni cant contributor.

In the past, candidate gene and pedigree analyses were very success ul in the study o diseases o monogeneticorigin: heritable dysregulation o certain metabolomictraits (inborn errors o metabolism) were among the rstto be associated with speci c genes [4]. However, theseapproaches are not use ul in complex diseases becausecandidate regions contain too many genes or there are nogroups o related individuals with a clear inheritancepattern o the disease phenotype. Inspired by the successo the Mendelian inheritance (genetic characteristicspassed rom the parent organism to o spring) approach,a great e ort was undertaken to generate a humanre erence database o common genetic variant patternsbased on a haplotype survey - the haplotype map(HapMap) [5]. Tis resource indeed improved, throughlinkage disequilibrium (LD) analyses, both the quality and the speed o GWAS, but it has not solved the majorissue o study outcome. Te common limitation o

GWAS is that they do not provide mechanisms ordisease; in other words, GWAS are unable to detectcausal variants. Speci cally, a GWAS provides in orma-tion about an association between a variant ( or example,SNP) and a disease, but the connection between a SNPand a gene is sometimes unclear. Tis is becauseannotated genes in the vicinity o a SNP are used in anattempt to explain the association unctionally. However,proximity to a gene (without any unctional analyses)should not be taken as the only sign that the identi edgene contributes to a disease.

It should be urther noted that the current analysistools or SNPs do not include all possible variants, butrather only common ones with a major allele requency greater than 0.01. SNPs with requencies o less than 1%are not visible (or hardly discernible) in GWAS at present

[3], and there ore some genetic contributions mightremain undiscovered. So ar, associations discovered by GWAS have had almost no relevance to clinical prognosisor treatment [6], although they might have contributed torisk strati cation in the human population. However,common risk actors ail to explain the heritability o human disease [7]. For example, a heritability o 40% hadbeen estimated or type 2 diabetes mellitus [8,9], but only 5 to 10% o the type 2 diabetes mellitus heritability can beexplained by the more than 40 con rmed diabetes lociidenti ed by GWAS [9,10].

Overcoming the limitationsTere are several ways to improve GWAS per ormance.Instead o searching or a single locus, multipleindependent DNA variants are being selected to identi y those responsible or the occurrence o a disease [2].Odds ratios could be more use ul than P -values or theassociations [6] in the interpretation o mechanisms andthe design o replication or unctional studies. Tis isespecially true i highly signi cant (but spurious) asso-ciations are observed in a small number o samples,which might originate rom a strati ed population. Tedesign o GWAS is also moving rom tagging a singlegene as a cause o disease to illuminating the pathway involved. Tis pathway might then be considered as atherapeutic target. In this way, GWAS comes back to itsroots. Te term post-GWAS is used to describe GWAS-inspired experiments designed to study disease mecha-nisms. Tis usually involves exploration o expressionlevels o genes close to the associated variants, orknockout experiments in cells or animals [11]. In otherwords, post-GWAS analyses bring unctional validationto associations [12].

Although omics approaches are power ul, they do notprovide a complete dataset. Each omic technology provides a number o speci c eatures ( or example, trans-cript level old change, protein identity or metabolite

concentration, concentration ratios). At present, experi-mental datasets consisting o thousands o eatures un-ortunately do not encompass all the eatures present in

vivo . With incomplete data, only imper ect conclusionscan be expected. However, the coverage o di erentomics eatures is expanding rapidly to overcome bothgenetic and phenotypic limitations o GWAS. As or thegenetic aspects, progress in whole genome sequencing( or example, the 1000 Genomes Project [13,14]) isbeginning to provide more in-depth analyses or less

requent (but still signi cant), and multiple, co-existing


Page 2 of 7

7/31/2019 gm333

3/7

disease loci. In addition, epigenetic eatures ( or example,methylation, histone deacetylation) will soon be expandedin GWAS [15-17].

Improvements in the interpretation o phenotypes arelikely to come rom causal DNA variants showing

signi cant and multiple associations with di erent omicsdata [11]. GWAS can be applied to intermediate pheno-types (including traits measured in the transcriptome,proteome or metabolome). Te resulting associations canidenti y SNPs related to molecular traits and providecandidate loci or disease phenotypes related to suchtraits. Disease-associated alleles might modulate distincttraits such as transcript levels and splicing, thus acting onprotein unction, which can be monitored directly ( orexample, by proteomics) or by metabolite assays. Tisleads to the conclusion that another way to improve theoutcomes o GWAS is the application o versatile andunbiased molecular phenotyping. Te choice o molecu-lar phenotyping approach will be driven by its quality regarding eature identi cation, coverage, through putand robustness.

Metabolomic phenotyping or GWASMetabolomics deals with metabolites with molecularmasses below 1,500 Da that refect unctional activitiesand transient e ects, as well as endpoints o biologicalprocesses, that are determined by the sum o a personsor tissues genetic eatures, regulation o gene expression,protein abundance and environmental infuences. Ideally all metabolites will be detected by metabolomics.Metabolomics is a very use ul tool that complementsclassical GWAS or several reasons. Tese includequanti cation o metabolites, unequivocal identi cationo metabolites, provision o longitudinal (time-resolved)dynamic datasets, high throughput ( or example, 500samples a week, with 200 metabolites or each sample),implementation o quality measures [18-21] and stan-dard ized reporting [22].

Enhancing classical GWAS or disease phenotypes withmetabolomics is better than metabolomics alone orunequivocal description o individuals, strati cation o test persons, and provision o multiparametric datasetswith independent metabolites or identi cation o whole

pathways a ected (including co-dependent metabolites).It is also instrumental in quantitative trait locus (Q L) ormetabolite quantitative trait locus (mQ L) analyses. Inthese studies quantitative traits ( or example, weight orconcentrations o speci c metabolites) are linked to DNAstretches or genes. Tis in ormation is important orassessing the extent o the genetic contribution to theobserved changes in phenotypes.

A part o the metabolome could be computed rom thegenome [23], but the in ormation would be static andhardly usable in biological systems except or annotation

purposes. Te time dynamics o the metabolome provides a means to identi y the relative contributions o genes and environmental impact in complex diseases.Tere ore, combining mGWAS expands the window o phenotypes that can be analyzed to multiple quantitative

eatures, namely total metabolite concentrations.

Metabolomic approachesMetabolomics mostly uses two major technologicalapproaches: non-targeted metabolomics by nuclear mag-netic resonance (NMR) or mass spectrometry (MS) [24];and targeted metabolomics by MS [20,25].

Non-targeted metabolomics provides in ormation onthe simultaneous presence o many metabolites or

eatures ( or example, peaks or ion traces). Samplethrough put may reach 100 samples a week on a singleNMR spectrometer, gas chromatography-mass spectro-meter (GC-MS) or liquid chromatography-tandem massspectrometer (LC-MS/MS) [20,25]. Te number o metabolites identi ed varies depending on the tissue andis usually between 300 (blood plasma) and 1,200 (urine)[26]. Te major advantage o non-targeted metabolomicsis its unbiased approach to the metabolome. Tequanti cation is a limiting issue in non-targeted metabo-lomics as it provides the di erences in the abundance o metabolites rather than absolute concentrations. In silico analyses (requiring access to public [27-30] or proprietary [31,32] re erence databanks) are required to annotate theNMR peaks, LC peaks or ion traces to speci c meta-bolites. Tere ore, i a metabolite mass spectrum is notavailable in the databases, the annotation is not automaticbut requires urther steps. Tese may include analysesunder di erent LC conditions, additional mass ragmen-tation or high-resolution (but slow) NMR experiments.

argeted metabolomics work with a de ned set o metabolites and can reach a very high throughput ( orexample, 1,000 samples per week on a single LC-MS/MS). Te set might range rom 10 to 200 metabolites in aspeci c ( or example, only or lipids, prostaglandins,steroids or nucleotides) GC-MS or LC-MS/MS assay [33-37]. o cover more metabolites, samples are divided intoaliquots and parallel assays are run under di erentconditions or GC- or LC-MS/MS. In each o the assays

the analyzing apparatus is tuned or one or more speci cchemical classes and stable isotope labeled standards areused to acilitate concentration determination. Te majoradvantages o targeted metabolomics are the throughputand absolute quanti cation o metabolites.

Both approaches (that is, targeted and non-targeted)reveal a large degree o common metabolite coverage[38] or allow or quantitative comparisons o the samemetabolites [21,39]. Metabolomics generates large-scaledatasets, in the order o thousands o metabolites, whichare easily included in bioin ormatics processing [40,41].


Page 3 of 7

7/31/2019 gm333

4/7

GWAS with metabolomics traitsTe outcome o GWAS depends very much on thesample size and the power o the study, which increaseswith the sample size. Some criticisms o GWAS haveaddressed this issue by questioning whether GWAS are

theoretically big enough to overcome the threshold o P -values and associated odds ratios. Initial GWAS or asingle metabolic trait (that is, plasma high-density lipoprotein (HDL) concentration [42]) were unable todetect the genetic component even with 100,000 samples.Tis indicates low genetic penetrance or this trait andsuggests that another approach should be used todelineate the underlying mechanism. More recently,metabolomics was ound to reveal valuable in ormationwhen combined with GWAS. Studies with a muchsmaller sample size (284 individuals) but with a largermetabolic set (364 eatured concentrations) demon-strated the advantage o GWAS combined with targetedmetabolomics [34]. In this study the genetic variantswere able to explain up to 28% o the metabolic ratio variance (that is, the presence or absence o a genetic variant coincided with up to 28% o changes in concen-tration ratios o metabolites rom the same pathway).Moreover, the SNPs in metabolic genes were indeed

unctionally linked to speci c metabolites converted by the enzymes, which are gene products o the associatedgenes.

In another study on the impact o genetics in humanmetabolism [35], involving 1,809 individuals but only 163metabolic traits, ollowed by targeted metabolomics(LC-MS/MS), it was shown that in loci with previously known clinical relevance in dyslipidemia, obesity ordiabetes ( FADS1 , ELOVL2 , ACADS , ACADM , ACADL , SPTLC3 , ETFDH and SLC16A9 ) the genetic variant islocated in or near genes encoding enzymes or solutecarriers whose unctions match the associating metabolictraits. For example, variants in the promoter o FADS1 , agene that encodes a atty acid desaturase, coincided withchanges in the conversion rate o arachidonic acid. In thisstudy, the metabolite concentration ratios were used asproxies or enzymatic reaction rates, and this yielded very robust statistical associations, with a very small P -valueo 6.5 10-179 or FADS1 . Te loci explained up to 36% o

the observed variance in metabolite concentrations [35].In a recent ascinating study on the genetic impact on thehuman metabolome and its pharmaceutical implicationswith GWAS and non-targeted metabolomics (GC orLC-MS/MS), 25 genetic loci showed unusually highpene trance in a population o 1,768 individuals (repli-cated in another cohort o 1,052 individuals) andaccounted or up to 60% o the di erence in metabolitelevels per allele copy. Te study generated many new hypotheses or biomedical and pharmaceutical research[21] or indications such as cardiovascular and kidney

disorders, type 2 diabetes , cancer, gout, venous thrombo-embolism and Crohns disease.

A speci c subset o the metabolome dealing with lipidstermed lipidomics has provided important insights intohow genetics contributes to modulated lipid levels. Tis

area is o particular interest or cardiovascular diseaseresearch, as about 100 genetic loci (without causalexplanation as yet) are associated with serum lipidconcentrations [42]. Lipidomics increases the resolutiono mGWAS over that with complex endpoints such astotal serum lipids ( or example, HDL only). For example,a NMR study showed that eight loci ( LIPC , CETP , PLTP ,

FADS1 , -2 , and -3 , SORT1 , GCKR , APOB , APOA1 ) wereassociated with speci c lipid sub ractions ( or example,chylomicrons, low-density lipoprotein (LDL), HDL),whereas only our loci ( CETP , SORT1 , GCKR , APOA1 )were associated with serum total lipids [43]. GWAS havealready enabled tracing o the impact o human ancestry on n -3 polyunsaturated atty acid (PUFA) levels. Tese

atty acids are an important topic in nutritional sciencein trying to explain the impact o PUFA levels onimmunological responses, cholesterol biosynthesis andcardiovascular disease [44-47]. It has been shown thatthe common variation in n -3 metabolic pathway genesand in the GCKR locus, which encodes the glucosekinase regulator protein, infuences the levels o plasmaphos pholipid o n -3 PUFAs in populations o Europeanancestry, whereas in other ancestries ( or example,A rican or Chinese) there is an impact on the infuencesin the FADS1 locus [48]. Tis explains the mechanismso di erent responses to diet in these populations.GWAS with NMR-based metabolomics can also beapplied to large cohorts. An example is the analysis o 8,330 indi viduals in whom signi cant associations( P < 2.31 10 -10) were identi ed at 31 loci, including 11new loci or cardiometabolic disorders (among thesemost were allocated to the ollowing genes: SLC1A4 ,

PPM1K , F12 , DHDPSL , TAT , SLC2A4 , SLC25A1 , FCGR2B , FCGR2A ) [49]. A comparison o 95 known lociwith 216 metabolite concentrations uncovered 30 new genetic or metabolic associations ( P < 5 10 -8) andprovides insights into the underlying processes involvedin the modulation o lipid levels [50].mGWAS can also

be used in the assignment o new unctions to genes. Inmetabolite quantitative trait locus (mQ L) analyses withnon-targeted NMR-based metabolomics, a previously uncharacterized amilial com ponent o variation inmetabolite levels, in addition to the heritability contribution rom the corresponding mQ L e ects, wasdiscovered [38]. Tis study demon strated that the so- ar

unctionally unannotated genes NAT8 and PYROXD2 arenew candidates or the mediation o changes in themetabolite levels o tri ethylamine and dimethylamine.Serum-based GWAS with LC/MS targeted metabolomics


Page 4 of 7

7/31/2019 gm333

5/7

has also contributed to eld o unction annotation:SLC16A9 , PLEKHH1 and SYNE2 have been assigned totransport o acylcarnitine C5 and metabolism o phos-phatidylcholine PCae36:5 and PCaa28:1, respectively [34,35].

mGWAS has recently contributed to knowledge onhow to implement personalized medicine by analysis o the background o sexual dimorphism [51]. In 3,300independent individuals 131 metabolite traits werequanti ed, and this revealed pro ound sex-speci c asso-ciations in lipid and amino acid metabolism - or example,in the CPS1 locus (carbamoyl-phosphate synthase 1;

P = 3.8 10 -10) or glycine. Tis study has importantimplications or strategies concerning the developmento drugs or the treatment o dyslipidemia and theirmonitoring; an example would be statins, or whichdi erent predispositions should now be taken intoaccount or women and men.

GWAS and metabolic pathway identifcationBy integrating genomics, metabolomics and complexdisease data, we may be able to gain important in or-mation about the pathways that are involved in thedevelop ment o complex diseases. Tese data arecombined in systems biology [52] and systems epidemi-ology evaluations [53,54]. For example, SNP rs1260326 inGCKR lowers asting glucose and triglyceride levels andreduces the risk o type 2 diabetes [55]. In a recentmGWAS [35], this locus was ound to be associated withdi erent ratios between phosphatidylcholines, thus pro- viding new insights into the unctional background o theoriginal association. Te polymorphism rs10830963 inthe melatonin-receptor gene MTNR1B has been ound tobe associated with asting glucose [56], and the same SNPassociates with tryptophan:phenylalanine ratios inmGWAS [35]: this is noteworthy because phenylalanineis a precursor o melatonin. Tis may indicate a

unctional relationship between the phenylalanine-mela-tonin pathway and the regulation o glucose homeostasis.Te third example is SNP rs964184 in the apolipoproteincluster APOA1-APOC3-APOA4-APOA5, which associatesstrongly with blood triglyceride levels [57]. Te sameSNP associates with ratios between di erent phos pha-

tidylcholines in mGWAS [35]: these are biochemically connected to triglycerides by only a ew enzymaticreaction steps.

ConclusionsBy combining metabolomics as a phenotyping tool withGWAS, the studies gain more precision, standardization,robustness and sensitivity. Published records worldwideillustrate the power o mGWAS. Tey provide new insights into the genetic mechanisms o diseases that isrequired or personalized medicine.

AbbreviationsGC, gas chromatography; GWAS, genome-wide association study; HDL, high-density lipoprotein; LC, liquid chromatography; LDL, low-density lipoprotein;mGWAS, metabolomics with genome-wide association study; mQLT,metabolite quantitative trait locus; MS, mass spectrometry; MS/MS, tandemmass spectrometer; NMR, nuclear magnetic resonance; PUFA, polyunsaturated

atty acid; QTL, quantitative trait locus; SNP, single nucleotide polymorphism.

Competing interests The author has no competing interests to declare.

AcknowledgementsI thank Dr Eva Lattka and Dr Gabriele Mller rom Helmholtz ZentrumMuenchen, Germany or help ul discussions and critical reading o themanuscript. This work was supported in part by a grant rom the GermanFederal Ministry o Education and Research (BMBF) to the German Center orDiabetes Research (DZD e.V.).

Author details1Helmholtz Zentrum Mnchen, German Research Center or EnvironmentalHealth, Institute o Experimental Genetics, Genome Analysis Center, 85764Neuherberg, Germany. 2Lehrstuhl r Experimentelle Genetik, TechnischeUniversitt Mnchen, 85350 Freising-Weihenstephan, Germany.

Published: 30 April 2012

References1. Ku CS, Loy EY, Pawitan Y, Chia KS: The pursuit o genome-wide association

studies: where are we now? J Hum Genet 2010,55:195-206.2. Visscher PM, Brown MA, McCarthy MI, Yang J:Five years o GWAS discovery.

Am J Hum Genet 2012,90:7-24.3. Altshuler D, Daly MJ, Lander ES:Genetic mapping in human disease. Science

2008,322:881-888.4. Palosaari PM, Kilponen JM, Hiltunen JK:Peroxisomal diseases. Ann Med 1992,

24:163-166.5. de Bakker PI, Yelensky R, Peer I, Gabriel SB, Daly MJ, Altshuler D:E ciency

and power in genetic association studies. Nat Genet 2005,37:1217-1223.6. McClellan J, King MC:Genetic heterogeneity in human disease. Cell 2010,

141:210-217.7. Manolio TA, Collins FS, Cox NJ, Goldstein DB, Hindor LA, Hunter DJ,

McCarthy MI, Ramos EM, Cardon LR, Chakravarti A, Cho JH, Guttmacher AE,Kong A, Kruglyak L, Mardis E, Rotimi CN, Slatkin M, Valle D, Whittemore AS,Boehnke M, Clark AG, Eichler EE, Gibson G, Haines JL, Mackay TF, McCarroll SA,Visscher PM:Finding the missing heritability o complex diseases. Nature2009,461:747-753.

8. So HC, Yip BH, Sham PC:Estimating the total number o susceptibilityvariants underlying complex diseases rom genome-wide associationstudies. PLoS One2010,5:e13898.

9. Park KS: The search or genetic risk actors o type 2 diabetes mellitus.Diabetes Metabol J 2011,35:12-22.

10. McCarthy MI:Genomics, type 2 diabetes, and obesity. N Engl J Med 2010,363:2339-2350.

11. Freedman ML, Monteiro AN, Gayther SA, Coetzee GA, Risch A, Plass C, CaseyG, De Biasi M, Carlson C, Duggan D, James M, Liu P, Tichelaar JW, Vikis HG, YouM, Mills IG:Principles or the post-GWAS unctional characterization o cancer risk loci. Nat Genet 2011,43:513-518.

12. Doucle M:Select: GWAS gets unctional.Cell 2010,143:177-178.

13. Via M, Gignoux C, Burchard EG: The 1000 Genomes Project: newopportunities or research and social challenges. Genome Med 2010,2:3.

14. Huang J, Ellinghaus D, Franke A, Howie B, Li Y:1000 Genomes-basedimputation identi es novel and re ned associations or the Wellcome Trust Case Control Consortium phase 1 Data. Eur J Hum Genet 2012. doi:10.1038/ejhg.2012.3

15. Rakyan VK, Down TA, Balding DJ, Beck S:Epigenome-wide associationstudies or common human diseases. Nat Rev Genet 2011,12:529-541.

16. Bataille V, Lens M, Spector TD: The use o the twin model to investigate thegenetics and epigenetics o skin diseases with genomic, transcriptomicand methylation data. J Eur Acad Dermatol Venereol 2012, doi:10.1111/j.1468-3083.2011.04444.x

17. Krueger F, Kreck B, Franke A, Andrews SR:DNA methylome analysis usingshort bisul te sequencing data. Nat Methods2012,9:145-151.


Page 5 of 7

7/31/2019 gm333

6/7

18. Gri n JL, Nicholls AW:Metabolomics as a unctional genomic tool orunderstanding lipid dys unction in diabetes, obesity and relateddisorders. Pharmacogenomics2006,7:1095-1107.

19. Grebe SK, Singh RJ:LC-MS/MS in the clinical laboratory - where to romhere? Clin Biochem Rev 2011,32:5-31.

20. Gri ths WJ, Koal T, Wang Y, Kohl M, Enot DP, Deigner HP: Targetedmetabolomics or biomarker discovery. Angew Chem Int Ed Engl 2010,49:5426-5445.

21. Suhre K, Shin SY, Petersen AK, Mohney RP, Meredith D, Wagele B, Altmaier E,Deloukas P, Erdmann J, Grundberg E, Hammond CJ, de Angelis MH,Kastenmller G, Kttgen A, Kronenberg F, Mangino M, Meisinger C, Meitinger T, Mewes HW, Milburn MV, Prehn C, Rafer J, Ried JS, Rmisch-Margl W,Samani NJ, Small KS, Wichmann HE, Zhai G, Illig T, Spector TD, Adamski J,Soranzo N, Gieger C: Human metabolic individuality in biomedical andpharmaceutical research. Nature2011,477:54-60.

22. Goodacre R, Broadhurst D, Smilde AK, Kristal BS, Baker JD, Beger RD, BessantC, Connor SC, Capuani G, Craig A, Ebbels T, Kell DB, Manetti C, Newton J,Paternostro G, Somorjai R, Sjstrm M, Trygg J, Wul ert F:Proposed minimumreporting standards or data analysis in metabolomics. Metabolomics2007,3:231-241.

23. Weckwerth W:Unpredictability o metabolism - the key role o metabolomics science in combination with next-generation genomesequencing. Anal Bioanal Chem2011,400:1967-1978.

24. Malet-Martino M, Holzgrabe U:NMR techniques in biomedical andpharmaceutical analysis. J Pharm Biomed Anal 2011,55:1-15.25. Koal T, Deigner HP:Challenges in mass spectrometry based targeted

metabolomics. Curr Mol Med 2010,10:216-226.26. Sreekumar A, Poisson LM, Rajendiran TM, Khan AP, Cao Q, Yu J, Laxman B,

Mehra R, Lonigro RJ, Li Y, Nyati MK, Ahsan A, Kalyana-Sundaram S, Han B, CaoX, Byun J, Omenn GS, Ghosh D, Pennathur S, Alexander DC, Berger A, ShusterJR, Wei JT, Varambally S, Beecher C, Chinnaiyan A:Metabolomic pro lesdelineate potential role or sarcosine in prostate cancer progression.Nature2009,457:910-914.

27. Williams AJ:Public chemical compound databases. Curr Opin Drug Discov Devel 2008,11:393-404.

28. Babushok VI, Linstrom PJ, Reed JJ, Zenkevich IG, Brown RL, Mallard WG, SteinSE:Development o a database o gas chromatographic retentionproperties o organic compounds. J Chromatogr 2007,1157:414-421.

29. Wishart DS, Knox C, Guo AC, Eisner R, Young N, Gautam B, Hau DD,Psychogios N, Dong E, Bouatra S, Mandal R, Sinelnikov I, Xia J, Jia L, Cruz JA,

Lim E, Sobsey CA, Shrivastava S, Huang P, Liu P, Fang L, Peng J, Fradette R,Cheng D, Tzur D, Clements M, Lewis A, De Souza A, Zuniga A, Dawe M,et al.:HMDB: a knowledgebase or the human metabolome. Nucleic Acids Res2009,37 (Database issue): D603-610.

30. Little JL, Williams AJ, Pshenichnov A, Tkachenko V:Identi cation o knownunknowns utilizing accurate mass data and ChemSpider. J Am Soc MassSpectrom2012,23:179-185.

31. Evans AM, DeHaven CD, Barrett T, Mitchell M, Milgram E:Integrated,nontargeted ultrahigh per ormance liquid chromatography/electrosprayionization tandem mass spectrometry plat orm or the identi cation andrelative quanti cation o the small-molecule complement o biologicalsystems. Anal Chem2009,81:6656-6667.

32. Lawton KA, Berger A, Mitchell M, Milgram KE, Evans AM, Guo L, Hanson RW,Kalhan SC, Ryals JA, Milburn MV:Analysis o the adult human plasmametabolome. Pharmacogenomics2008,9:383-397.

33. Unterwurzacher I, Koal T, Bonn GK, Weinberger KM, Ramsay SL:Rapid samplepreparation and simultaneous quantitation o prostaglandins and

lipoxygenase derived atty acid metabolites by liquid chromatography-mass spectrometry rom small sample volumes. Clin Chem Lab Med 2008,46:1589-1597.

34. Gieger C, Geistlinger L, Altmaier E, Hrabe de Angelis M, Kronenberg F,Meitinger T, Mewes HW, Wichmann HE, Weinberger KM, Adamski J, Illig T,Suhre K:Genetics meets metabolomics: a genome-wide association studyo metabolite pro les in human serum. PLoS Genet 2008,4:e1000282.

35. Illig T, Gieger C, Zhai G, Rmisch-Margl W, Wang-Sattler R, Prehn C, Altmaier E,Kastenmller G, Kato BS, Mewes HW, Meitinger T, de Angelis MH, KronenbergF, Soranzo N, Wichmann HE, Spector TD, Adamski J, Suhre K:A genome-wideperspective o genetic variation in human metabolism. Nat Genet 2010,42:137-141.

36. Ceglarek U, Shackleton C, Stanczyk FZ, Adamski J:Steroid pro ling andanalytics: going towards sterome. J Steroid Biochem Mol Biol 2010,

121:479-480.37. Dunn WB, Broadhurst D, Begley P, Zelena E, Francis-McIntyre S, Anderson N,

Brown M, Knowles JD, Halsall A, Haselden JN, Nicholls AW, Wilson ID, Kell DB,Goodacre R; Human Serum Metabolome (HUSERMET) Consortium:Procedures or large-scale metabolic pro ling o serum and plasma usinggas chromatography and liquid chromatography coupled to massspectrometry. Nat Protoc 2011,6:1060-1083.

38. Nicholson G, Rantalainen M, Li JV, Maher AD, Malmodin D, Ahmadi KR, FaberJH, Barrett A, Min JL, Rayner NW, To t H, Krestyaninova M, Viksna J, Neogi SG,Dumas ME, Sarkans U; MolPAGE Consortium, Donnelly P, Illig T, Adamski J,Suhre K, Allen M, Zondervan KT, Spector TD, Nicholson JK, Lindon JC,Baunsgaard D, Holmes E, McCarthy MI, Holmes CC:A genome-widemetabolic QTL analysis in Europeans implicates two loci shaped by recentpositive selection. PLoS Genet 2011,7:e1002270.

39. Suhre K, Meisinger C, Doring A, Altmaier E, Belcredi P, Gieger C, Chang D,Milburn MV, Gall WE, Weinberger KM, Mewes HW, Hrab de Angelis M,Wichmann HE, Kronenberg F, Adamski J, Illig T:Metabolic ootprint o diabetes: a multiplat orm metabolomics study in an epidemiologicalsetting. PLoS One2010,5:e13953.

40. Oresic M:Metabolomics, a novel tool or studies o nutrition, metabolismand lipid dys unction. Nutr Metab Cardiovasc Dis2009,19:816-824.

41. Krumsiek J, Suhre K, Illig T, Adamski J, Theis FJ:Gaussian graphical modelingreconstructs pathway reactions rom high-throughput metabolomics

data. BMC Syst Biol 2011,5:21.42. Teslovich TM, Musunuru K, Smith AV, Edmondson AC, Stylianou IM, Koseki M,Pirruccello JP, Ripatti S, Chasman DI, Willer CJ, Johansen CT, Fouchier SW,Isaacs A, Peloso GM, Barbalic M, Ricketts SL, Bis JC, Aulchenko YS, Thorlei ssonG, Feitosa MF, Chambers J, Orho-Melander M, Melander O, Johnson T, Li X,Guo X, Li M, Shin Cho Y, Jin Go M, Jin Kim Y,et al :Biological, clinical andpopulation relevance o 95 loci or blood lipids. Nature2010,466:707-713.

43. Petersen AK, Stark K, Musameh MD, Nelson CP, Rmisch-Margl W, Kremer W,Rafer J, Krug S, Skurk T, Rist MJ, Daniel H, Hauner H, Adamski J, TomaszewskiM, Dring A, Peters A, Wichmann HE, Kaess BM, Kalbitzer HR, Huber F, P ahlertV, Samani NJ, Kronenberg F, Dieplinger H, Illig T, Hengstenberg C, Suhre K,Gieger C, Kastenmller G:Genetic associations with lipoproteinsub ractions provide in ormation on their biological nature. Hum Mol Genet 2012,21:1433-1443.

44. Moodley T, Vella C, Djahanbakhch O, Bran ord-White CJ, Craw ord MA:Arachidonic and docosahexaenoic acid de cits in preterm neonatalmononuclear cell membranes. Implications or the immune response at

birth. Nutr Health2009,20:167-185.45. Choi YS, Goto S, Ikeda I, Sugano M:E ect o dietary n-3 polyunsaturatedatty acids on cholesterol synthesis and degradation in rats o di erent

ages. Lipids1989,24:45-50.46. Misra A, Khurana L, Isharwal S, Bhardwaj S:South Asian diets and insulin

resistance. Br J Nutr 2009,101:465-473.47. Singer P, Berger I, Moritz V, Forster D, Taube C:N-6 and N-3 PUFA in liver

lipids, thromboxane ormation and blood pressure rom SHR during dietssupplemented with evening primrose, sunfowerseed or sh oil.Prostaglandins Leukot Essent Fatty Acids1990,39:207-211.

48. Lemaitre RN, Tanaka T, Tang W, Manichaikul A, Foy M, Kabagambe EK,Nettleton JA, King IB, Weng LC, Bhattacharya S, Bandinelli S, Bis JC, Rich SS,Jacobs DR Jr, Cherubini A, McKnight B, Liang S, Gu X, Rice K, Laurie CC,Lumley T, Browning BL, Psaty BM, Chen YD, Friedlander Y, Djousse L, Wu JH,Siscovick DS, Uitterlinden AG,et al .:Genetic loci associated with plasmaphospholipid n-3 atty acids: a meta-analysis o genome-wide associationstudies rom the CHARGE Consortium.PLoS Genet 2011,7:e1002193.

49. Kettunen J, Tukiainen T, Sarin AP, Ortega-Alonso A, Tikkanen E, Lyytikinen LP,Kangas AJ, Soininen P, Wrtz P, Silander K, Dick DM, Rose RJ, Savolainen MJ,Viikari J, Khnen M, Lehtimki T, Pietilinen KH, Inouye M, McCarthy MI, JulaA, Eriksson J, Raitakari OT, Salomaa V, Kaprio J, Jrvelin MR, Peltonen L, PerolaM, Freimer NB, Ala-Korpela M, Palotie A, Ripatti S:Genome-wide associationstudy identi es multiple loci infuencing human serum metabolite levels.Nat Genet 2012,44:269-276.

50. Tukiainen T, Kettunen J, Kangas AJ, Lyytikinen LP, Soininen P, Sarin AP, Tikkanen E, OReilly PF, Savolainen MJ, Kaski K, Pouta A, Jula A, Lehtimki T,Khnen M, Viikari J, Taskinen MR, Jauhiainen M, Eriksson JG, Raitakari O,Salomaa V, Jrvelin MR, Perola M, Palotie A, Ala-Korpela M, Ripatti S:Detailedmetabolic and genetic characterization reveals new associations or 30known lipid loci. Hum Mol Genet 2012,21:1444-1455.

51. Mittelstrass K, Ried JS, Yu Z, Krumsiek J, Gieger C, Prehn C, Roemisch-Margl W,


Page 6 of 7

7/31/2019 gm333

7/7

Polonikov A, Peters A, Theis FJ, Meitinger T, Kronenberg F, Weidinger S,Wichmann HE, Suhre K, Wang-Sattler R, Adamski J, Illig T:Discovery o sexualdimorphisms in metabolic and genetic biomarkers. PLoS Genet 2011,7:e1002215.

52. Hood L, Heath JR, Phelps ME, Lin B:Systems biology and new technologiesenable predictive and preventative medicine. Science2004,306:640-643.

53. Adourian A, Jennings E, Balasubramanian R, Hines WM, Damian D, Plasterer TN, Clish CB, Stroobant P, McBurney R, Verheij ER, Bobeldijk I, van der Gree J,Lindberg J, Kenne K, Andersson U, Hellmold H, Nilsson K, Salter H, Schuppe-Koistinen I:Correlation network analysis or data integration andbiomarker selection. Mol Biosyst 2008,4:249-259.

54. Haring R, Wallascho ski H:Diving through the -omics: the case or deepphenotyping and systems epidemiology. OMICS2012. doi: 10.1089/ omi.2011.0108

55. Vaxillaire M, Cavalcanti-Proenca C, Dechaume A, Tichet J, Marre M, Balkau B,Froguel P: The common P446L polymorphism in GCKR inverselymodulates asting glucose and triglyceride levels and reduces type 2diabetes risk in the DESIR prospective general French population. Diabetes2008,57:2253-2257.

56. Prokopenko I, Langenberg C, Florez JC, Saxena R, Soranzo N, Thorlei sson G,Loos RJ, Manning AK, Jackson AU, Aulchenko Y, Potter SC, Erdos MR, Sanna S,Hottenga JJ, Wheeler E, Kaakinen M, Lyssenko V, Chen WM, Ahmadi K,Beckmann JS, Bergman RN, Bochud M, Bonnycastle LL, Buchanan TA, Cao A,Cervino A, Coin L, Collins FS, Crisponi L, de Geus EJ,et al :Variants in MTNR1Binfuence asting glucose levels. Nat Genet 2009,41:77-81.

57. Kathiresan S, Melander O, Guiducci C, Surti A, Burtt NP, Rieder MJ, Cooper GM,Roos C, Voight BF, Havulinna AS, Wahlstrand B, Hedner T, Corella D, Tai ES,Ordovas JM, Berglund G, Vartiainen E, Jousilahti P, Hedblad B, Taskinen MR,Newton-Cheh C, Salomaa V, Peltonen L, Groop L, Altshuler DM, Orho-Melander M:Six new loci associated with blood low-density lipoproteincholesterol, high-density lipoprotein cholesterol or triglycerides inhumans. Nat Genet 2008,40:189-197.

doi:10.1186/gm333Cite this article as : Adamski J:Genome-wide association studies withmetabolomics. Genome Medicine2012,4:34.


Page 7 of 7

gm333

Documents