Advertisement

AdvertisementHigh-quality and unlimited genotyping from preciousDNA samplesGenetic analyses often require large amounts of genomic DNA. Since the availability of DNA from individual genomes isoften limited, accurate replication of large amounts of high-quality genomic DNA is required. QIAGEN’s REPLI-g ® Kits andService use multiple displacement amplification (MDA) to overcome the limitations of small amounts of template DNA.Effect of Heat and Alkaline Denaturationon Loci RepresentationC T value5040302010095°CHeat denaturationLocus 1Locus 2REPLI-g protocolAlkaline denaturationHigh-Molecular-Weight DNA6.05.04.03.02.01.00.0gDNA95°C denaturationAlkaline denaturation––––––––– 5.8 kb– 5.2 kb– 3.0 kb– 1.0 kbReliable SNP genotyping0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0In contrast to traditional PCR-based whole genome amplification,QIAGEN’s REPLI-g Kits and Service provide highly uniform DNAamplification across the entire genome, with minimal sequence bias.Reproducible, high yields of DNA can be obtained from various clinicaland non-clinical research samples, including genomic DNA, fresh or driedblood, buffy coat, and tissue culture cells.A number of kit formats are available to suit all DNA yield andthroughput requirements. Typical yields per 50 µl reaction are up to 8 µg(Screening Kit), 10 µg (Mini Kit) and 40 µg (Midi Kit). A reproducible yieldof amplified DNA is usually achieved regardless of the quantity of templateDNA, enabling immediate downstream genetic analysis, including SNPgenotyping, STR analysis, Southern blotting and RFLP analysis, subcloning,and DNA sequencing. The average product length is typically greater than10 kb, with a range between 2 kb and 100 kb.DNA denaturation is a critical step in the amplification of whole genomicDNA. Excessive heat can damage the DNA resulting in incomplete andless consistent locus representation. In contrast, the alkaline denaturation(Mini and Midi Kit) or gentle thermal denaturation (Screening Kit) providedby QIAGEN’s REPLI-g Kits results in uniform DNA amplification.The REPLI-g Screening Kit combines a number of features making it idealfor use in manual or automated high-throughput mutation screening,including high volume pipetting steps for increased accuracy, compatibilitywith 96-well microplates, and a streamlined protocol for fast and easy set up.The REPLI-g Service provides amplification of unlimited amounts of DNAfrom limited samples. A stringent quality control assay provides informationon the quality of the amplified DNA, enabling you to make reliablepredictions for the success of your downstream assay.Visit www.qiagen.com/goto/REPLI-g to find out more about unlimitedgenotyping from precious samples!WWW.QIAGEN.COM

Table of contentsLetter from the Editor . . . . . . . . . . . . . . . . . . . . . . . 4Index of Experts . . . . . . . . . . . . . . . . . . . . . . . . . . . 4Q1: What do you consider whenchoosing platforms? . . . . . . . . . . . . . . . . . . 6Q2: What criteria do you use for selectingSNPs or potentially functional sites? . . . . . 8Q3: Which factors are most important toachieving high genotyping accuracy? . . . 11Q4: What quality control measuresdo you employ? . . . . . . . . . . . . . . . . . . . . . . 13Q5: What methods or tools do you use forthe analysis and visualization of data? . .16List of Resources . . . . . . . . . . . . . . . . . . . . . . . . . 18

Announcing TaqMan ® Drug Metabolism GenotypingAssays for basic and clinical research. For moreinformation, please visit dme.appliedbiosystems.comChoose your genome.Choose your SNP.Ready-to-use TaqMan ® SNP Genotyping Assays.Make your disease-research mapping and screening efforts moreproductive—and more reliable—with single-tube, ready-to-useTaqMan ® assays for the mutations of your choice. For unprecedentedgenomic coverage, select from our library of human mapping and coding SNP assays orhave custom assays built for any possible SNP in any genome. A simple, scalable workflow runs on AppliedBiosystems real-time PCR systems and gives you the accurate and reproducible results you’re looking for.Your SNP assays are ready when you are at: http://www.allsnps.comFor Research Use Only. Not for use in diagnostic procedures. The PCR process and the 5' nuclease process are covered by patents owned by Roche Molecular Systems, Inc., andF. Hoffman-La Roche Ltd. Applied Biosystems is a registered trademark and AB (Design) is a trademark of Applera Corporation or its subsidiaries in the US and/or certain other countries.TaqMan is a registered trademark of Roche Molecular Systems, Inc. © 2006 Applied Biosystems. All rights reserved.

What do you considerwhen choosing platforms?In general, we perform SNP genotyping by restrictionfragment length polymorphism (RFLP), sequencing,PCR-CTPP. In our published freeware, SNP-RFLPing,the RFLP available restriction enzymes for input ofSNP ID (rs#, ss#), gene name, gene ID, and anyformats of sequences are provided. Users can get theRFLP information for SNP genotyping online.— Hsueh-Wei ChangSome of the things we consider include throughputand performance as measured by reliability,reproducibility, and accuracy. We also consider costand things such as ease-of-use.In addition to that, we may not have the righttechnology, and so we may ask, “Is there a bettertechnology available elsewhere?” — whether it befrom a collaborator or even a fee-for-use service.[The last] in our consideration of approaches is[asking whether] there are there new, more efficienttechnologies on the horizon. Efficient is defined byany one or more [aspects such as] throughput,performance, cost, and ease-of-use. So a lot goesinto considering or choosing approaches to thesetypes of studies.In my lab, we actually have a number of SNPgenotyping technologies, including technology fromSequenom, Illumina, and Affymetrix. We also have,although we're not using them all that much,Applied Biosystems' TaqMan and SNPlex. At timeswe still find ourselves using PCR followed byrestriction enzyme digestion.Being a genotyping and genomics researchinstitute, we're closely watching the latestdevelopments in resequencing technology.Resequencing is, after all, the ultimate genotypingtechnology. It's just cost prohibitive right now.— David DugganThe Australian Genome Research Facility is a majornational research facility, which has been set up bythe federal government to be a provider of genomicservices for the Australasian region. This basicallymeans that we are a service provider and as such donot perform our own research studies. Rather, ourclients request us to run specific SNPs on theirsamples and provide them with the data. Whenchoosing platforms for SNP genotyping, we'reactually caught in a 'tug-of-war' between two keyfactors: the quality and reliability of a specific SNPgenotyping platform versus the cost. Our clients arelargely cost motivated, but at the same time, weneed to ensure that the data that’s generated isaccurate. Our primary motivator is the accuracy andreliability of the platform, as well as its throughput(because we are a high-throughput laboratory),followed very closely by the cost.— Darryl IrwinSNP genotyping platforms each have associatedcosts and benefits, and will vary from project toproject and from lab to lab.Arguably, the most important consideration isthe scale of the project: How many SNPs are to beassayed, and how many individuals are to be tested?Say you want to test 10 SNPs in 10,000 DNA samples.Highly multiplexed technologies are not required, butone would probably like to avoid running 10,000lanes on a sequencer. An appropriate alternative is FP-SBE, or fluorescence polarization detection of singlebaseextension products (Chen et al. 1999). This6 SNP Genotyping Tech Guide Genome Technology

method has the advantage that the detectionhappens right in the microtiter plate (noelectrophoresis is required), the assays are fairlyrobust, and the initial cost of the oligonucleotideprobe is low. If you want to genotype 10,000 or moreSNPs, then the ultra-high-throughput platformsoffered by Illumina (Shen et al. 2005) and Affymetrix(Matsuzaki et al. 2004) are ideal.The real difficult decisions arise when you wantto genotype 100 to 200 SNPs in 1,000 to 2,000individuals, for example in a high-power candidategene-based association mapping experiment. Thereis no obvious commercially available technology thatis fully compatible with a 200 SNP, 2,000 sampleproject: 200 SNPs is not sufficient to make use of thepower of the ultra-high-throughput systems, andyou need access to a large number of PCR machinesto make singleplex assays workable. This is exactlythe problem we faced in our lab, and so wedeveloped an open-source genotyping system basedon the OLA (Oligonucleotide Ligation Assay,Landegren et al. 1988), using 16-plex genotypingreactions, and array-based genotype detection(Macdonald et al. 2005). This permitted us to collectgenotypes on large panels of 2,000 individuals,and to assay 200 SNPs cheaply, efficiently, and withhigh accuracy.— Stuart MacdonaldAt the French National Genotyping Center, a numberof platforms are already in place, including massspectrometry, TaqMan, Amplifluor, SNPlex, andIllumina. The first consideration to find theappropriate platform for a given project is thenumber of samples and the number of SNPs to begenotyped. This allows us to choose among thedifferent technologies. Then among the appropriateplatforms, the accuracy of the genotyping method isthe second consideration.For a large number of SNPs in a single gene,with a small number of samples, sequencing is veryefficient. For a small number of SNPs on a relativelylarge number of samples, we have found TaqMan tobe the most accurate method among the three wehave tested (mass spectrometry using the GOODassay, Amplifluor, and TaqMan; Giancola et al.,2006). For a large number of SNPs and a largenumber of samples, we have chosen SNPlex andhave obtained very good results. Starting with 384SNPs, Illumina is the system chosen.— Heather McKhannAccuracy, cost, user-friendliness, and possibility oftechnical support from the company providing theinstruments.— Louise NordforsI choose the platform according to the requirements:(i) many SNPs in a small subset of patients or (ii) fewSNPs in thousands of patients. [It also depends on]the budget and the price-per-genotype calculated bythe technical director.— Jose Luis RoyoAccuracy always first, then cost and flexibility, andothers. Although all are important, the accuracy isthe most important for general considerations; costand flexibility are important for large-scalegenotyping projects.— Huanming YangGenome Technology SNP Genotyping Tech Guide 7

databases to review potential alleles and minor allelefrequency.— Darryl IrwinAll else being equal, one would prefer to designassays for those SNPs most likely to convert toworking assays. For each query SNP, the sequencearound it will be available from at least two (andhopefully more) alleles. There are several propertiesof the flanking sequence that will influence whetherthe query SNP yields accurate genotypes: GCcontent, sequence repetitiveness, and sequencepolymorphism are three of the important ones. SNPsin very AT-rich sequence and those within highlyrepetitive regions are less likely to result in convertingassays, due to instability of hybridization betweengenotyping probes and their targets. Also, ifthe sequence is very polymorphic around the querySNP, it may not be possible to design allele- orSNP-specific genotyping probes. Segregatingpolymorphisms can and do interfere with thehybridization of genotyping probes to their targets,and can adversely impact the accuracy of thegenotype data.— Stuart MacdonaldFor mapping using SNPs, it is important to haveSNPs that are evenly distributed throughout thegenome. We have made consensus maps forArabidopsis recombinant inbred lines with arecurrent parent. In this case, it was important tochoose SNPs that distinguish the recurrent parentfrom all the other accessions. For functional studies,SNPs that cause non-conservative amino acidchanges are prioritized. Another possibility whenfunctional information is lacking is to choose asubset of SNPs to genotype that represent thedifferent haplotypes. Once SNPs have been chosen,there may be technical constraints for theirgenotyping; additional adjacent SNPs are usually notallowed and for multiplex methods, there must becompatibility between the SNPs.— Heather McKhannLiterature searches in related research areas giveinput on interesting candidate genes and SNPs.Databases such as NCBI and HapMap provide toolsfor selecting frequent SNPs and showing haplotypeblocks and tag SNPs.— Louise NordforsSearching for candidate SNPs on the basis of theirputative functionality is a lottery. I try to capture thehaplotype diversity using SNPs according to (i)Celera, (ii) Perlegen, and (iii) HapMap, usually in nonconserved,intronic regions and taking into accountthe informativity (frequency of 30 percent to 50percent) and covering the different haplotype blocks.— Jose Luis RoyoSNP scoring, such as the criteria given by Illumina SNPscoring algorithm, and htSNPs, generated by theHapMap Project. Generally speaking, we willconsider genotyping SNPs at a 10 kb to 50 kb densityfor general studies on human diseases.— Huanming YangGenome Technology SNP Genotyping Tech Guide 9

SNP Genotypingas easy as1, 2, 31 - Hybridize2 - Extend12-Plex48-Plex3 - DetectTMGenomeLab SNPstreamSNP Genotyping SystemsIn an automated, scalable package, the GenomeLabSNPstream Genotyping System combines multiplexingabilities with single-base primer extension technology toprovide accurate, cost-effective, medium to high throughputanalysis. This system provides a scalable solution forprocessing a range of 4,600 to over 3 million genotypes perday. Just the kind of power you need to accelerate geneticdiscoveries, validation and testing of signature panels.• Two levels of multiplexing reduces cost per genotype andper sample at any throughput• As little as 0.04 ng DNA per genotype required• Fast and easy optimized primer design with Autoprimer.com• Automated image processing into genotype callswww.beckmancoulter.com/123Genomics Proteomics Cell Analysis Centrifugation Lab Tools Particle Characterization Bioseparation Lab AutomationSimplify . Automate . Innovate© 2006 Beckman Coulter, Inc.

Which factors are mostimportant to achievinghigh genotyping accuracy?DNA quality. Purity is the number one factor.Molecular weight is also important. However, while itmay be important, it has become less of an issuetoday, especially since we're moving from RFLPs toSSTRs and now to SNPs. The region of the genomethat we're interrogating is actually getting smaller andsmaller, thus there's really not the requirement to haveextremely high molecular weight DNA samples likethere were when we were doing Southern blots.— David DugganThe number one factor isDNA quality. If you start withpoor quality DNA, you'regoing to get a poor-qualityresult. As these reactions arebecoming more highlymultiplexed, the quality ofthe starting material isbecoming more and moreimportant. We have qualitycontrol steps in our process where we thoroughlycheck the DNA before we put it in the reactions tomake sure that it is high quality.We do make some DNA quality recommendationsto our clients, though we don't recommend a specifictechnology for DNA extraction. We recommend theA260/280 ratio being 1.8-2.2 and the A260/230ration being 1.6-2.4. We also recommend that theDNA is high molecular weight and available foramplification. This is a key point, the DNA may bepresent and quantifiable by OD but it may notamplify. To check this we recommend using a realtimePCR system such as SYBR green to check DNAactivity and normalize the concentration.“We believe that the more factorstaken into consideration forimproving accuracy, the higheraccuracy will be achieved.”We also recommend that there is minimal EDTAin the elution buffers, because the EDTA caninterfere, particularly when it's at high concentration.If [EDTA] is up around 0.5 mM, then it can interferewith the activity of some of the enzymes that we use,so we recommend that they should elute in eitherwater, 10mM TRIS or reduced TE.— Darryl IrwinThe whole process of genotyping (from organism togenotype calling) should bemade as routine and— Huanming Yangstreamlined as possible.Some form of liquid-handlingrobot is essential for anythingother than modest amountsof genotyping, and willmarkedly reduce errorsintroduced by manualpipetting. Reducing thenumber of pipette steps andsample movement among plates is also important tominimize the possibility of contamination across wells.To minimize handling of our valuable DNA samples,using our liquid-handling robot we create manyreplicate sets of aliquoted DNA plates. Each set ofplates holds the entire panel of DNA samples, andeach well represents one sample, and containssufficient DNA for a single PCR reaction. The DNAaliquot plates are then dried and stored at -80ºC untilPCR. In this way we eliminate the need to repeatedlyfreeze/thaw DNA, and radically streamline the processof setting up PCR reactions.— Stuart Macdonald(continued on p.18)Genome Technology SNP Genotyping Tech Guide 11

Unwind.Introducing the Rosetta Syllego system, a genetic data managementand analysis solution. The path to breakthrough science just became morestraightforward. With the Rosetta Syllego system, you can better plan and executegenetic studies and ensure high-quality data to prevent loss of time in downstreamanalyses. It also helps you connect relevant genotypic and phenotypic study dataand catalog these data and analysis results for drawing solid conclusions. All so youcan focus on what really matters—advancing your research and drug developmentprograms. Go to www.rosettabio.com/syllego for more information.1.800.737.6583 (U.S.)+1.206.926.1220 (Outside the U.S.)sales@rosettabio.comCopyright © 2006 Rosetta Inpharmatics LLC. All rights reserved. Syllego is a trademark of Rosetta Inpharmatics LLC.The Rosetta Biosoftware convergence icon is a registered trademark of Rosetta Inpharmatics LLC.

What quality controlmeasures do you employ?We always perform sequencing after RFLP assay forsome samples to ensure high reproducibility.Alternatively, another restriction enzyme is selectedto test the RFLP result again. Real-time PCR canprovide high-throughput genotyping, but it isexpensive in our experiments.— Hsueh-Wei ChangThe first quality control measure would be theDNA purity. DNA purity is measured, for example, byA260/280 standard UV spectra or, more recently,fluorescence (e.g., picogreen). Also, DNA quality —although that's becoming, depending on thequestion, less of an issue — is measured by agarosegel electrophoresis. Lately, we've also foundourselves switching to the Agilent Bioanalyzer,which consumes considerably less DNA than anagarose gel does.Next are quality control and quality assurancemeasures that are provided to us by the technologyproviders. For example, in the Affymetrix 500Kprotocol, there are several QC and QA steps alongthe procedure that would tip us off to somethingbeing wrong with the sample or the batch ofsamples. That includes, in the case of the Affymetrix500K, an agarose gel and a quantification of PCRproducts following the PCR amplification step.There's also another agarose gel following thefragmentation reaction. Then, finally, in regards tothe technology provider QCs and QAs, there arepost-hybridization statistics that we look at,including the MDRs, MCRs, and SNP call rates.The third of four measures that we have in placeincludes the more traditional replicate and duplicatecontrols on each plate that we process. On everysingle 96-well microtiter plate, we include positiveand, if the technology allows, negative controls inthe form of water. In every 96-well microtiter plate,we include replicates and duplicates. By replicate, Imean the same DNA is found on each 96-wellmicrotiter plate. So, if we have a collection of athousand samples spread out over 10 96-wellmicrotiter plates, there's at least one DNA incommon on all 10 of those plates. That is used tolook at inter-plate reproducibility and reliability. Then,on each one of those 96-well plates, we also haveduplicate controls to account for reliability andreproducibility within a plate. In the case of duplicatecontrols, we have no less than two randomlyselected DNAs per 96-well plate.When it's available, we can look at theconcordance of our SNP genotype with that of publicdata, such as the International HapMap project.Those are the four measures we have in place,and all are used at some point in [the] process toassess the quality and control of the actual sample,the experiments, and the data at the end.— David DugganWe have a variety of quality control steps. We'reNATA accredited to international standard ISO17025, and as such quality is a priority for us.The first QC step when we get the DNA is to testthat the DNA is at the right OD ratios that the clientshave been advised to provide. We also do real-timePCR to quantify the DNA and ensure that even if DNAis quantified by OD it is also amplifiable. We use SYBRGreen on the real-time PCR platform for this.Once our DNA has passed, we design and orderour SNP assay as previously described. When thoseGenome Technology SNP Genotyping Tech Guide 13

assays arrive, we firstly check the primer sequenceand quality by mass spectrometry. We then run theassays on a small number of client samples, as wellas our in-house single DNA controls and our inhousepooled DNA controls. Quite a number of SNPsin the public databases are non-polymorphic, [so] byrunning these assays on pooled DNA we can adviseour clients of expected minor allele frequencies priorto incurring genotyping charges from their fullproject. We also include negative controls to monitorfor potential contamination.On larger projects, we request that clientsleave specific plate wells empty. When they arrive,we place a combination of oligonucleotides forwhich we know the masses. When the data isbeing analyzed we check the contents of thesewells [by mass spec] to ensure there hasn't been alab mix-up.— Darryl IrwinOn the microtiter plates, among the DNA samplesof unknown genotype, we place some blanks.These are valuable to indicate any assay-specificoddities during PCR amplification or genotyping.We also use various DNA controls of knowngenotype on the plates. If the assay discriminatespoorly among individuals with different knowngenotypes, this indicates that the assay isproblematic. In the past we have also conductedresequencing of subsets of individuals and SNPs toensure that the error rate of our entire SNPgenotyping pipeline is low.— Stuart MacdonaldInitially, we tested every method (GOOD assay,Amplifluor, TaqMan, sequencing SNPlex) usingduplicate plates. With the exception of the GOODassay, we found very high repeatability.Subsequently, each plate is genotyped one timeexcept in cases where there are obvious problems.Prior to genotyping, the quality and quantity of theDNA is assessed. We have found that very lowconcentrations of DNA may affect genotyping resultsand further, it is important that the samples arehomogeneous in concentration. Then, in each plate,a number of controls are essential, includingnegative controls (water) and positive controls(previously sequenced individuals with all possiblegenotypes).— Heather McKhannResults are checked and confirmed regularly usingdifferent methods: SNPlex, pyrosequencing, TaqMan,and sequencing.— Louise NordforsFirst, checking the Hardy-Weinberg equilibrium law.Then we perform 5 percent to 10 percent reextractionfrom blood and re-typing with doubleblindcheck. For association studies, 2 percentmistyping is assumable. If more, you must discardthe marker.— Jose Luis RoyoFive to 10 percent duplicates of the total samplesused for QA. If analyzing pedigree samples, we canalso estimate the genotyping quality from thepedigree information. And 1 percent of the totalsamples are always [included] as negative controls.— Huanming Yang14 SNP Genotyping Tech Guide Genome Technology

Relax. We’ve already built aworld-class microarray facility.So you don’t have to.Looking to expand your capacityor for a backup to your internalcapabilities?Rely on Gene Logic for superiorservices and results. Our servicesinclude:•SNP genotyping from DNA,blood, or tissue•Gene expression profiling fromtissue, cells, whole blood, orRNA•Optimized gene expressionprofiling of whole blood•qRT-PCR•RNA isolation from low-yieldtissues•Bioinformatic analysesWhy build it when you canoutsource it?610 Professional DriveGaithersburg, Maryland 20879Phone: 1.800.GENELOGIC or+1.301.987.1709Fax: 301.987.1701Email: info@genelogic.comWeb:www.genelogic.com© 2006 Gene Logic Inc. All rights reserved.

What methods or tools doyou use for the analysisand visualization of data?Excel and SPSS analysis.—Hsueh-Wei ChangWe are using home-grown visualization software to,for example, display coverage of a gene of interest ina candidate gene study. In some cases, we plugit into graphical user interfaces like that offered bythe HapMap.Secondly, we are using commercial software forlinkage analysis. We do a lot of linkage analysis withAffymetrix's 10K. We are using commercial softwareas well as relying on ourstatistical collaborators toanalyze the linkage data.We're relying on manymore statistical collaboratorsand evolving software forwhole-genome scans. Whilewe have ideas about how toanalyze that data today, thisis really an evolving area ofresearch, and I rely on mystatistical collaborators fordirection in analysis.Some of us researchers play a key role inproviding statisticians with direction, and some of usare taking a stab at the analysis itself, but that's alittle dangerous. We're not statisticians and we couldbe creating more havoc than help. Just as we sawwith the microarray world 10 years ago, we're seeingsoftware companies bring, in some cases, userfriendlysoftware to market that a lot of statisticiansare having a hard time swallowing. Because ease-ofusedoesn't mean statistical robustness. Just becauseyou can select an algorithm or a test, it doesn't mean“Some of us researchers play akey role in providing statisticianswith direction, and some of us aretaking a stab at the analysis itself,but that's a little dangerous.”it's the right test. It doesn't substitute for having astatistician on your team.— David DugganFor analyzing the genotyping data, we actually usethe Sequenom Typer software for the Sequenomplatform, and we use the Applied BiosystemsGeneMapper software for the Applied Biosystemsplatform. We supply back to our clients thegenotypes only. Clients may request the mass spectraor electropherograms; however, these are notroutinely supplied as themajority of our clients do— David Duggannot have the software toanalyze these.We also analyze ascatter plot of the peakareas of the two alleles toensure that our genotypesare clustered very clearly.They fall into three separategroups: homozygous wildtype, heterozygous, andhomozygous mutant. Thisplot gives us a lot of information about the quality ofthe DNA that's going through and the accuracy ofthe genotypes. We always expect to see threeseparate groups, but on occasion, we have observeda split in the heterozygous groups, which indicatesthat there is a copy number variant and this is alsoadvised back to our clients.We find the Sequenom platform is a very usefultool for looking at copy number variants because itcan perform quantitative genotyping to a high levelof accuracy. Sequenom's platform can perform16 SNP Genotyping Tech Guide Genome Technology

quantitative genotyping to look at allelic frequenciesin pooled DNA samples. We can use the samemethodology to analyze individual DNA samples forcopy number variants with little additional cost.— Darryl IrwinThe genotypes we collect are all derived fromintensity measurements on spotted arrays. We usethe ArrayVision software (GE Healthcare) to semiautomaticallyextract the intensity data, and with thisdata in hand use custom scripts written in the freelyavailable statistical programming language R tocluster the individuals and call genotypes. Thesescripts are described in Macdonald et al. (2005). Theadvantage of R over licensed commercial software isthat scripts can be easily customized. Since there is awide community of academics developing R scriptsand depositing them in accessible databases, oftenthere are useful scripts and functions readily availablefor downstream analysis of the genotype data.— Stuart MacdonaldIn general we use the methods provided with a givenplatform. For Amplifluor and TaqMan, we have usedApplied Biosystem's SDS2.0 software, and forSNPlex, the Genemapper software. For the GOODassay, software called SNPmaster has beendeveloped for internal use at the French NationalGenotyping Center. Following analysis, genotypingdata is submitted to a database located at the INRA'sUnité de Recherche Génomique-Info.— Heather McKhannWe have created our own database for storing thedata. Statistical analyses are mainly performed inJMP IN, which also provides graphs, etc.— Louise NordforsThe traditional statistical ones: Excel sometimes, andmainly SPSS, STATA, Episheet, and EPI-Info.— Jose Luis RoyoIllumina Gencall, Sequenom MassArray typer, andPerkinElmer SNPscorer. All work reasonably well.— Huanming YangGenome Technology SNP Genotyping Tech Guide 17

Q3: Which factors are mostimportant to achievinghigh genotypingaccuracy?(continued from p.11)This depends very much on the method being used.We found that the GOOD assay, which separatesalleles based on their molecular weight, is highlyaccurate. However, as it comprises many steps, thereis a loss in repeatability. In fluorescent methods, thekey factor is the design of the primers/probes.Depending on the method, this may be done by themanufacturer, in which case there is little controlover design; in certain cases, the researcher maydesign their own primers/probes, for example forTaqMan, using PrimerExpress.— Heather McKhannA highly accurate genotyping method together withcareful research about the gene and SNPs to beinvestigated to avoid, for example, multiplesequence variation sites and genotyping ofpseudogenes. Of course, highly skilled and carefultechnicians are crucial.— Louise NordforsAutomation in all processes. In DNA extraction,working in 96-well or 384-well formats and withrobots preparing the PCRs and the following postprocess.— Jose Luis RoyoBasically, we believe that the more factors taken intoconsideration for improving accuracy, the higheraccuracy will be achieved. However, too manyfactors may lead to difficulties in sample handlingand automation. Right now, we prefer ASO- (allelespecificoligonucleotides) and LSO- (locus-specificoligonucleotides) based hybridization, primerextension and ligation.What's more, primer design algorithms, oligosynthesis protocols, redundancy of reactions, andquality of DNA samples are also very important inensuring maximum accuracy in SNP genotypingmethods.— Huanming YangList of resourcesOur panel of experts referred to a number publications,which we’ve compiled in the list below.PublicationsChen X, Levine L, Kwok PY (1999) Fluorescencepolarization in homogeneous nucleic acid analysis.Genome Res. 9:492-498.Giancola S, McKhann HI, Bérard A, et al. (2006)High throughput genotyping in plants:comparison of three current technologies. TheorAppl Genet. 112(6):1115-1124Landegren U, Kaiser R, Sanders J, Hood L (1988) Aligase-mediated gene detection technique.Science. 241:1077-1080.Macdonald SJ, Pastinen T, Genissel A, et al. (2005)A low-cost open-source SNP genotypingplatform for association mapping applications.Genome Biol. 6:R105.Matsuzaki H, Dong S, Loi H, et al. (2004)Genotyping over 100,000 SNPs on a pairof oligonucleotide arrays.Nat Methods.1:109-11.Sauer S, Lechner D, Berlin K, et al. (2000) A novelprocedure for efficient genotyping of singlenucleotide polymorphisms.Nucleic Acids Res. 28:E13Shen R, Fan JB, Campbell D, et al. (2005) HighthroughputSNP genotyping on universalbead arrays. Mutat Res. 573:70-82.Syvänen AC (2001) Accessing genetic variation:genotyping single nucleotide polymorphisms.Nat Rev Genet 2:930-942.AcknowledgmentsMany thanks to Aurélie Bérard of INRA for advisingon the answers submitted by Heather McKhann.18 SNP Genotyping Tech Guide Genome Technology