QTLminer: identifying genes regulating quantitative traits.

SOFTWARE Open Access QTLmine : iden i ying genes egula ing quan i a i e ai s Rudi Albe s, Klaus Schugha * Abs ac Backg ound: Quan i a i e ai locus (QTL) mapping iden i ies genomic egions ha likely con ain genes egula ing a quan i a i e ai . Howe e , QTL egions may encompass ens o hund eds o genes. To ind he mos p omising candida e genes ha egula e he ai , he biologis ypically collec s in o ma ion om mul iple esou ces abou he genes in he QTL in e al. This p ocess is e y labo ious and ime consuming. Resul s: QTLmine is a bioin o ma ics ool ha au oma ically pe o ms QTL egion analysis. I is a ailable in GeneNe wo k and i in eg a es in o ma ion such as gene anno a ion, gene exp ession and sequence polymo phisms o all he genes wi hin a gi en genomic in e al. Conclusions: QTLmine subs an ially speeds up disco e y o he mos p omising candida e genes wi hin a QTL egion. Backg ound Quan i a i e ai locus (QTL) mapping is a powe ul me hod o iden i y genes egula ing complex ai s. By combining molecula ma ke da a o gene ically ela ed indi iduals wi h pheno ypic ai alues, genomic QTLs a e iden i ied ha likely con ain gene ic egula o s o he ai . This s a egy has bo h been applied o ‘classical’ ai s like body weigh , blood p essu e o disease suscep ibili y, as well as o ai s measu ed using high- h oughpu ech- nologies: mRNA abundances measu ed by mic oa ays [1,2], and p o ein o me aboli e abundances measu ed by mass spec ome y [3,4]. QTLs gene ally span a genomic egion con aining ens o hund eds o genes. Iden i ica ion o he mos p omising egula ing genes wi hin QTL in e - als, which can hen be unc ionally es ed, s ill emains a majo challenge. QTLmine has been implemen ed in he GeneNe wo k [5], a la ge esou ce wi h geno ypes, pheno- ypes and gene exp ession p o iles o mul iple o ganisms and gene ic e e ence popula ions. I au oma ically ana- lyses a QTL egion and in eg a es in o ma ion abou he candida e genes, so ha he bes candida e genes can be quickly iden i ied. Implemen a ion QTLmine was implemen ed in Py hon as pa o he GeneNe wo k [5]. Resul s and Discussion QTLmine akes a QTL in e al as inpu , which is de ined by he ch omosome and he s a and end posi- ions in megabases. The p og am au oma ically gene - a es a lis o genes wi hin he in e al and e ie es addi ional in o ma ion o each gene. The i s pa comp ises anno a ion da a such as gene name, desc ip- ion, genomic posi ion, Gene On ology (GO) e ms and KEGG pa hways in which he gene is implica ed. Nex , he amoun o non-synonymous single nucleo ide poly- mo phisms (nsSNPs) wi hin he gene is displayed. nsSNPs esul in amino acid changes in he co espond- ing p o ein. These changes may modi y i s s uc u e and may hus be causa i e o he pheno ypic di e ences which we e mapped. Fu he mo e, he use can selec h ee GeneNe wo k exp ession da a se s. Fo each o he da a se s, gene exp ession and in o ma ion abou cis- egula ion will be added. In his way, he use can see whe he candida e genes a e exp essed in he issue unde s udy. Genes ha a e only exp essed in he issue o in e es and no in o he s migh e en be be e candi- da es. The use should howe e be awa e ha high exp ession is no always equi ed, i.e. lowly exp essed * Co espondence: [email p o ec ed] Depa men o In ec ion Gene ics, Helmhol z Cen e o In ec ion Resea ch and Uni e si y o Ve e ina y Medicine Hanno e , Inho ens asse 7, 38104 B aunschweig, Ge many Albe s and Schugha BMC Bioin o ma ics 2010, 11:516 h p://www.biomedcen al.com/1471-2105/11/516 © 2010 Albe s and Schugha ; licensee BioMed Cen al L d. This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License (h p://c ea i ecommons.o g/licenses/by/2.0), which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. genes may also ha e a s ong in luence on ai s. In addi ion, in o ma ion abou cis- egula ion is added. A gene is cis- egula ed i i s exp ession maps close o i s own genomic posi ion. Any cis- egula ed gene in he QTL egion is a good candida e o egula e he ai , since cis- egula ion indica es a di e ence in gene exp es- sion le els, which may egula e he ai . Albe s e al. [6] ha e shown ha ‘ghos ’cis-eQTLs (exp ession QTLs) can be de ec ed i he e a e SNPs o o he sequence a ian s in he p obe egions ha cause a di - e ence in hyb idiza ion signal. These cases migh also be in e es ing since he sequence a ian s in he p obe egions migh esul in changes in he p o ein s uc u e which can egula e he ai . To demons a e he u ili y o he p og am, we ook a QTL ho spo on mouse dis al ch omosome 1 as exam- ple. This egion, also called Q 1, con ains many QTLs ha con ol neu al and beha io al pheno ypes, including mo o beha io , escape la ency, emo ionali y, and sei- zu e suscep ibili y (Szs1) [7]. Mozhui e al. ha e u he in es iga ed his egion and e ealed a highly complex gene exp ession egula o y in e al in Q 1, composed o mul iple loci modula ing he exp ession o unc ion- ally cogna e se s o genes. In he dis al pa o Q 1, hey ha e iden i ied he gene Fmn2 as a s ong candi- da e. To e-analyze his in e al using QTLmine , open a web b owse and go o h p://genene wo k.helmhol z- hzi.de and click ‘Sea ch’and ‘QTLmine ’.In he o m ha appea s, choose Ch omosome 1, iew om 173 Mb o 177 Mb. Selec wo mouse s ains o which nsSNPs should be analyzed. Choose h ee GeneNe wo k da a se s and click ‘Analyze QTL in e al’. Th ee hippocam- pus da a se s in BXD, CXB and LXS mice a e chosen by de aul . Figu e 1A displays he esul s o QTLmine . I shows he genes wi hin he Q 1 in e al, as well as hei desc ip ions and posi ions. To ob ain mo e in o ma ion abou he unc- ion o he genes, Gene On ology e ms and all KEGG pa hways in which each gene occu s a e displayed (excluded om he igu e). Nex , he amoun s o (s ain speci ic) nsSNPs a e shown. Clicking his numbe links o GeneNe wo k’s SNP b owse whe e de ailed in o ma ion abou he SNPs can be sea ched. In he exp ession and cis columns, one can di ec ly see which genes wi hin he QTL in e al a e exp essed and cis- egula ed, and also compa e exp ession and cis- egula ion o o he issues (da a se s). The use should, howe e , be awa e ha o en gene exp ession alues a e measu ed in whole o gans. These alues migh no e lec he exp ession alues in a speci ic cell ype o in e es , especially i he amoun o his cell ype is only a small ac ion o he o al numbe o cells in he o gan. In he Q 1 in e al, o e 100 genes we e ound. Indeed he gene Fmn2 shows up as a e y good candida e, because i exhibi s a high exp ession alue, s ong cis- egula ion and con ains nsSNPs be ween he pa en al s ains C57BL/6J and DBA/2J. To help he use o in e p e he esul s, we added a sco e o each gene and he possibili y o so he genes acco ding o ei he posi- ion o sco e. The sco e anges be ween 0 and 4 and inc eases by s eps o one uni i 1) he gene has an exp es- sion alue g ea e han 8 in he i s da a se ; 2) he gene is cis- egula ed in he i s da a se ; 3) he genes con ains non-synonymous SNPs; 4) he gene con ains indels. I should be no ed ha hese sco es a e jus helping he biologis use o so he lis and o ge he mos in e es - ing candida es a he op o he esul s able. The use should s ill s udy gene desc ip ions, pa hways and o he in o ma ion o all genes o ge he bes candida es. A gene wi h a sco e o 1 which is known o be in ol ed in he biological p ocess unde s udy migh be a much be e candida e han a gene wi h unknown unc ion and a highe sco e. An addi ional ea u e o QTLmine is he isualiza ion o he haplo ypes wi hin a QTL egion. In he haplo ype plo (Figu e 1B), indi iduals a e so ed acco ding o hei quan i a i e ai alue, and hei haplo ypes a e indica ed by colo s. The gene Kcnj9 is one o he genes wi h a s ong ans eQTL in he Q 1 in e al in hippo- campus. To ob ain BXD haplo ypes o his gene, sea ch he BXD hippocampus da a se in GeneNe wo k o Kcnj9, click p obese 1450712 a , click he bu on ‘In e - al Mapping’and ill in Ch omosome 1, 173 un il 177 Mb. Selec ‘Haplo ype Analys ’and click ‘Remap’.I is immedia ely ob ious ha mice wi h a C57BL/6J ( ed) allele a he QTL loca ion ha e a low ai alue, whe eas mice wi h a DBA/2J (g een) allele ha e a high ai alue. This isualiza ion may be used o ine map- ping QTLs. Suppose ha a QTL sepa a es wo g oups o indi iduals wi h low and high ai alue o mos indi iduals, bu some indi iduals (which we e no ye s udied) ha e a ecombina ion wi hin he in e al. Then hese indi iduals may be used o u he geno yping and in his way, a QTL egion can be u he na owed down. Conclusions QTLmine au oma ically in eg a es gene anno a ion, Gene On ology e ms, KEGG pa hway in o ma ion, gene exp ession and cis- egula ion da a o all genes wi hin a QTL in e al. Wi h only a ew mouse clicks on he Gen- eNe wo k websi e, he mos p omising candida e genes wi hin a gi en QTL egion a e quickly highligh ed. A ailabili y and equi emen s P ojec name: QTLmine P ojec home page: h p://genene wo k.helmhol z-hzi. de (click Sea ch - QTLmine ) Ope a ing sys em(s): Pla o m independen Albe s and Schugha BMC Bioin o ma ics 2010, 11:516 h p://www.biomedcen al.com/1471-2105/11/516 Page 2 o 3 P og amming language: Py hon O he equi emen s: none License: none Any es ic ions o use by non-academics: none Acknowledgemen s We hank Robe W. Williams o e y help ul commen s and sugges ions. Funding: This wo k was suppo ed by he Helmhol z Cen e o In ec ion Resea ch and by GeNeSys (Ge man Ne wo k o Sys ems Gene ics), a Helmhol z Vi ual Ins i u e, inanced by he Helmhol z Associa ion. Au ho s’con ibu ions RA concei ed he s udy, w o e he manusc ip and implemen ed he p og am. KS concei ed he s udy and w o e he manusc ip . All au ho s ead and app o ed he inal manusc ip . Recei ed: 7 May 2010 Accep ed: 15 Oc obe 2010 Published: 15 Oc obe 2010 Re e ences 1. B em RB, Y e G, Clin on R, K uglyak L: Gene ic dissec ion o ansc ip ional egula ion in budding yeas . Science 2002, 296:752-755. 2. Bys ykh L, Wee sing E, Don je B, Su on S, Ple che MT, Wil shi e T, Su AI, Vellenga E, Wang J, Manly KF, Lu L, Chesle EJ, Albe s R, Jansen RC, Williams RW, Cooke MP, de Haan G: Unco e ing egula o y pa hways ha a ec hema opoie ic s em cell unc ion using ‘gene ical genomics’.Na Gene 2005, 37:225-232. 3. Foss EJ, Radulo ic D, Sha e SA, Rude e DM, Bedalo A, Goodle DR, K uglyak L: Gene ic basis o p o eome a ia ion in yeas . Na Gene 2007, 39:1369-1375. 4. Keu en jes JJ, Fu J, de Vos CH, Lommen A, Hall RD, Bino RJ, an de Plas LH, Jansen RC, V eugdenhil D, Koo nnee M: The gene ics o plan me abolism. Na Gene 2006, 38:842-849. 5. GeneNe wo k. [h p://genene wo k.helmhol z-hzi.de]. 6. Albe s R, Te ps a P, Li Y, B ei ling R, Nap JP, Jansen RC: Sequence polymo phisms cause many alse cis eQTLs. PLoS ONE 2007, 2:e622. 7. Mozhui K, Ciobanu DC, Schiko ski T, Wang X, Lu L, Williams RW: Dissec ion o a QTL ho spo on mouse dis al ch omosome 1 ha modula es neu obeha io al pheno ypes and gene exp ession. PLoS Gene 2008, 4: e1000260. doi:10.1186/1471-2105-11-516 Ci e his a icle as: Albe s and Schugha : QTLmine : iden i ying genes egula ing quan i a i e ai s. BMC Bioin o ma ics 2010 11:516. Figu e 1 QTLmine esul s and haplo ype isualiza ion. A) Pa o he esul s e u ned by QTLmine . The genes in he QTL in e al and hei desc ip ions and posi ions a e p esen ed. Nex , he amoun o non-synonymous SNPs wi hin all a ailable mouse s ains is gi en, and he amoun o nsSNPs be ween wo chosen s ains, in his case C57BL/6J and DBA/2J. The exp ession columns display he mean exp ession alues wi hin h ee GeneNe wo k da a se s. No e ha mul iple alues pe gene a e gi en when he gene is measu ed by mul iple p obes. The nex columns show cis- egula ion. I a gene is cis- egula ed, he LRS alue (likely a io s a is ic) ha indica es he signi icance o he cis QTL is also gi en. B) Haplo ype isualiza ion o BXD mice o he Q 1 egion. A ed (g een) line indica es ha he mouse ca ies he C57BL/6J (DBA/2J) geno ype. The e ical lines a e exis ing geno ypic da a and colo ed lines in be ween a e in e ed geno ypes based on he mapping o su ounding ma ke s. Blue lines indica e a egion con aining a ecombina ion b eak poin . S ains a e o de ed acco ding o ai alue. Genes in he in e al a e displayed a he op (colo ed ba s). Albe s and Schugha BMC Bioin o ma ics 2010, 11:516 h p://www.biomedcen al.com/1471-2105/11/516 Page 3 o 3