Open Access Method2005ZipperlenetVolume al. 6, Issue 2, Article R19

A universal method for automated mapping comment Peder Zipperlen¤*, Knud Nairz¤†, Ivo Rimann†, Konrad Basler*, Ernst Hafen†, Michael Hengartner* and Alex Hajnal†

Addresses: *Institute of Molecular , University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland. †Institute of Zoology, University of Zürich, Winterthurerstrasse 190, CH-8057 Zürich, Switzerland.

¤ These authors contributed equally to this work. reviews

Correspondence: Peder Zipperlen. E-mail: [email protected]. Knud Nairz. E-mail: [email protected]

Published: 17 January 2005 Received: 9 September 2004 Revised: 15 November 2004 Biology 2005, 6:R19 Accepted: 9 December 2004 The electronic version of this article is the complete one and can be found online at http://genomebiology.com/2005/6/2/R19 reports

© 2005 Zipperlen et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Mapping

Amaps forhigh-throughput DrosophilaInDel sequence and method C.polymorphisms. elegans.

for genotyping by mapping InDels. This method has been used to create fragment-length deposited research deposited Abstract

Small insertions or deletions (InDels) constitute a ubiquituous class of sequence polymorphisms found in eukaryotic . Here, we present an automated high-throughput genotyping method that relies on the detection of fragment-length polymorphisms (FLPs) caused by InDels. The protocol utilizes standard sequencers and genotyping software. We have established genome-wide FLP maps for both Caenorhabditis elegans and Drosophila melanogaster that facilitate genetic mapping

with a minimum of manual input and at comparatively low cost. refereed research

Background where -specific, dual-labeled fluorescent TaqMan For humans and model organisms, such as worms and , probes guarantee specificity [7]. However, the need for two the availability of high-density sequence polymorphism maps dual-labeled fluorescent probes, expensive specialized chem- greatly facilitates the rapid mapping and cloning of [1- istry and specialized machinery increase the costs per assay of 3]. Key advantages of most molecular polymorphisms are the this approach significantly. Similarly, denaturing high-per- fact that they are codominant and in general phenotypically formance liquid chromatography (DHPLC) also analyses the interactions neutral. The vast majority of sequence polymorphisms are primary amplification product [8]. This approach is based on single-nucleotide polymorphisms (SNPs). melting differences of homo- versus heteroduplex DNA frag- ments under increasingly denaturing conditions and requires The most direct approach for SNP detection is sequencing of no specific labeling of the PCR fragments. However, condi- a PCR product spanning the polymorphism, but this is too tions have to be optimized for every assay, throughput is lim- costly and labor intense for high-throughput genotyping. For ited and specialized equipment is required. DHPLC has been this reason, several different strategies and methods have used in small-scale genotyping projects in Drosophila mela- been developed in order to detect SNPs more efficiently. In nogaster [9]. information general, assays can be grouped into strategies, where the nature of the SNP is determined by directly analyzing the pri- Of the methods that detect the SNP in a secondary assay, mary PCR product and those that require a secondary assay restriction fragment length polymorphism (RFLP) analysis performed on the primary amplification product [4-6]. An are very popular [10]. For this purpose, only those SNPs that important strategy of the first group is the 5' nuclease assay, alter a restriction site are analyzed. A great advantage of

Genome Biology 2005, 6:R19 R19.2 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. http://genomebiology.com/2005/6/2/R19

(a) ZH1-01: 3bp InDel; no A addition (b) ZH2-01: 12bp InDel; with A addition (c) ZH3-05a: 2bp InDel; with A addition Fragment length (bp) Fragment length (bp) Fragment length (bp) 117118 119 120 121 122 123 134 136 138 140 142 144 146 148 150 174175 176 177 178 179 180 1600 2000 6000 1600 1200 4000 1200 800 800 2000 400 400

Bristol Bristol Bristol

117118 119 120 121 122 123 134 136 138 140 142 144 146 148 150 [bp] 174175 176 177 178 179 180 300 12000 4000 10000 3000 8000 200 6000 2000 4000 100 1000 2000 Flourescence

Hawaii Hawaii Hawaii

117118 119 120 121 122 123 134 136 138 140 142 144 146 148 150 [bp] 174175 176 177 178 179 180 600 5000 4000

4000 400 3000 3000 2000 2000 200 1000 1000

Bristol Hawaii Bristol Hawaii Bristol Hawaii

(d) ZH3-23: 1bp InDel; with A addition (e) 3R160: 1bp InDel; poly-T stretch (f) ZHX-22: 6bp InDel; poly-C stretch Fragment length (bp) Fragment length (bp) Fragment length (bp) 120 125 130 171172 173 174 175 176 177 201 203 205 207 209 211 213 6000 3000 12000 5000 10000 4000 2000 8000 3000 6000 1000 4000 2000 1000 1 2 2000 1 2 3

Bristol EP Bristol

120 125 130 171172 173 174 175 176 177 201 203 205 207 209 211 213 5000 14000 10000 12000 4000 8000 10000 3000 8000 6000 2000 6000 4000 4000 2000 1000 2000

Flourescence 1 2 1 2 3

Hawaii FRT Hawaii

120 125 130 171172 173 174 175 176 177 201 203 205 207 209 211 213 10000 20000 5000 16000 8000 4000 12000 6000 3000 8000 4000 2000 4000 2000 1000 1 2 3 1 2 3 4

Bristol Hawaii EP FRT Hawaii Bristol

FLPFigure detection 1 of InDels of various sizes in homozygotes and heterozygotes FLP detection of InDels of various sizes in homozygotes and heterozygotes. In each panel the top two graphs show the homozygotes and the bottom graph the heterozygote. Gray shaded areas mark the defined expected allele lengths and red lines indicate the borders of a predefined window of expected allele lengths. (a-c) Detection of InDels in C. elegans that show increasing levels of adenosine (A) addition. (a) 3-bp InDel ZH1-01 with no A addition; (b) 12-bp InDel ZH2-01 with A addition; (c) 2-bp InDel ZH3-05a with A addition. (d) 1-bp InDel ZH3-23 in C. elegans with A addition. An unambiguous allele- call can be made, irrespectively of the level of A addition: both homozygous samples consist of two peaks at different positions, whereas the heterozygous animal exhibits three peaks. (e) The 1-bp InDel 3R160 in Drosophila runs over a 12-13 nucleotide poly(T) stretch and exhibits stutter bands. Even in this case, a clear allele-call can be made (three peaks in homozygous and four peaks in heterozygous animals). (f) The 6-bp InDel ZHX-22 in C. elegans occurs in a poly(C) stretch and the FLP graph displays stutter bands. As expected, the longer fragment exhibits a higher degree of stuttering.

Genome Biology 2005, 6:R19 http://genomebiology.com/2005/6/2/R19 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. R19.3

RFLP analysis is that no specialized equipment is needed and formed automatically using the Applied Biosystems GeneMa- it can be carried out in every laboratory. RFLP maps recently pper software commonly used for genotyping STRs established for Caenorhabditis elegans and Drosophila are (Materials and methods). comment used regularly in genotyping projects [2,3,11]. However, RFLP analysis requires significant manual input. Moreover, To demonstrate the feasibility of this strategy, we have vali- the use of different restriction enzymes with different reac- dated 112 evenly spaced FLP assays at 3 centimorgan (cM) tion requirements adds another level of complexity that resolution in C. elegans (one every 0.9 megabase-pair (Mbp)) makes this method difficult to automate. Primer-extension- and 54 FLP assays at 4 cM resolution for the Drosophila auto- based technologies have also gained some prominence [12]. somes. This set of FLP assays allows us to rapidly map muta- Here, a primer that anneals right next to the polymorphism is tions to small chromosomal subregions with a minimum of extended by one polymorphism-specific terminator nucle- manual input. Furthermore, we provide a list of predicted reviews otide. Extension products are analyzed by size or, alterna- InDels for which additional assays can be readily designed in tively, by differences in the behavior of incorporated versus the chromosomal subregion of interest. Those non-validated non-incorporated terminator nucleotides under polarized FLPs enhance the resolution of the map by a factor of 5.6 and fluorescent light [13]. Swan and colleagues [14] have devel- 17.9, respectively. oped a set of fluorescence polarization-template directed incorporation (FP-TDI) assays for C. elegans. However, this We show the usefulness of this approach by identifying novel approach is labor intensive and requires specialized chemis- of previously characterized genes. In summary, we try and equipment. Using DNA microarrays, large numbers of have taken advantage of a publicly available dataset to adapt reports SNPs can be analyzed in parallel, but the number of individu- a technology widely used for STR analysis to genetic mapping. als that can be analyzed is low because of the high cost per Thanks to the complete automation of genotyping, this chip [15,16]. approach is considerably faster, more reliable and cheaper than previously used mapping strategies in C. elegans or Dro- Besides SNPs, short tandem repeats (STRs) or microsatellites sophila. represent another class of sequence polymorphisms used for genotyping [17-21]. STRs result in fragment length differ- research deposited ences that are either detected on gel-based or capillary Results and discussion sequencers or high-resolution hydrogels (Elchrom Scientific Detection of fragment length polymorphisms (FLPs) Inc.). One advantage of STRs over SNPs is that they are highly To detect a FLP, the region of interest is amplified in a stand- polymorphic and are thus ideal for measuring the degree of ard PCR reaction with one fluorescently labeled primer, and variability in natural populations. STRs are, however, present the PCR products are directly analyzed on a capillary at much lower density than SNPs and are therefore not suita- sequencer. Fragment sizes are determined automatically rel- ble for high-resolution mapping of genes. ative to an internal size standard with AppliedBiosystem's

GeneMapper software (for details see Materials and meth- refereed research Interestingly, a significant proportion of the currently availa- ods). The software then allocates fragment sizes to previously ble polymorphisms are caused by small insertions or dele- calibrated . tions (InDels). Weber et al. [22] identified a genome-wide set of about 2,000 human InDel polymorphisms and estimated Taq polymerase has the tendency to catalyze the addition of that InDels comprise at least 8% and up to 20% of all human adenosine (A) to the 3' end of PCR products. This activity polymorphisms. This is in line with the findings of Berger and could make it difficult to achieve the single base-pair resolu- co-workers [2] who found that 16.2% of polymorphisms in tion required to assay all available InDels and may hamper Drosophila are of the InDel type. Also, two independent stud- allele-calling [23]. However, we have found that the sensitiv- ies in C. elegans found that InDels constitute between 25% ity of a capillary sequencer and the genotyping software is suf- interactions and 28% of all polymorphisms [3,14]. In addition, those stud- ficient to allow for unambiguous allele assignment even for ies found that the vast majority of InDels are due to 1-2 base- 'difficult' sequences exhibiting 3' A addition. The examples pair (bp) differences (65% in Drosophila [2], 84% in C. ele- shown in Figure 1a-d illustrate that robust genotyping is fea- gans [3]). sible for 1-bp InDels even when 3' A addition occurs. Another problem is the stuttering of the polymerase when it encoun- To take full advantage of this class of small InDel polymor- ters poly(N) stretches. However, larger InDels are reliably phisms, we have developed a strategy that allows us to detect detected by the software in poly(N) stretches (Figure 1f), and most, if not all, InDels by analyzing the lengths of primary in a few difficult cases visual inspection can even resolve and information PCR products on a capillary sequencer at single base-pair res- unambiguously assign 'stuttering' 1-bp InDels according to olution. We call these assays fragment length polymorphism the location and number of peaks (Figure 1e). (FLP) assays. Importantly, this approach can easily be auto- mated on standard robotic pipetting platforms as it involves a Genotyping with FLP assays is extremely accurate. In a con- simple PCR reaction setup. Furthermore, allele calling is per- trol experiment, we genotyped all 96 samples of the strains

Genome Biology 2005, 6:R19 R19.4 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. http://genomebiology.com/2005/6/2/R19

FRT42B and EP0755 for the 1-bp InDel 2R090 and 231 sam- assays were evenly spaced on the genetic map with an average ples homo- and heterozygous for the C. elegans Bristol and distance between neighboring assays of about 4 cM, corre- Hawaii backgrounds, respectively, for the 1-bp InDel ZH5-16. sponding to an average resolution of 1.77 Mbp on the physical 2R090 exhibits both stuttering and A addition and hence is map encompassing 95,55 Mbp [27,28]. Taking into account especially difficult to resolve (see Additional data file 8). The the non-validated InDels, the maximal average resolution is was correctly and automatically assigned by Gen- currently 314 kb or 0.7 cM. On the left arm of chromosome 3, eMapper in all 423 assays. Thus, automated genotyping based where the genetic map is inexact, FLPs were spaced on the on FLPs is sensitive down to single base-pair resolution and physical map assuming colinearity between the two maps. is extremely robust. The accuracy of FLP mapping is compa- The length of amplicons ranges from 99 to 365 bp, and the rable to other methods such as TaqMan (error rate less than 1 size difference ranges from 1 to 54 bp (Additional data file 9). in 2,000 [24]), minisequencing (99.5% [25]), and pyrose- quencing (97.3 % [25]). Our Drosophila FLP assays are in part derived from a set of InDels of size difference 7 bp or more (termed PLPs by Berger C. elegans and Drosophila FLP maps et al. [2]). However, since 86.8% of all Drosophila InDels In C. elegans, genetic experiments are performed almost exhibit a length difference of one to six nucleotides [2], so far exclusively in the background of the standard wild-type strain only a small subset of the available InDels has been covered. N2 (C. elegans variety Bristol) [26]. For gene mapping exper- The approach presented here significantly increases the iments, the polymorphic strain CB4856 (C. elegans, variety number of possible FLP assays for genotyping and offers a Hawaii) has proved extremely useful [3]. When compared to greater flexibility and higher resolution. N2, CB4856 contains on average one SNP every 840 bp and approximately 25% of all polymorphisms are InDels [14]. FLP mapping of C. elegans genes Starting from the dataset previously published by Wicks et al. To demonstrate the usefulness of the C. elegans FLP map, we [3], 112 FLPs that are evenly spaced on the physical map of C. mapped three previously characterized mutations on chro- elegans were validated to date (Figure 2a). The confirmation mosome II that exhibit diverse . Those were the rate of the predicted InDels was 88% (n = 169). Most failures centrally located let-23(sy1) allele that causes an 80% pene- to detect a FLP are probably due to original sequencing trant vulvaless [29], rol-1(e91) in the middle of the errors. The average distance between neighboring FLP assays left chromosome arm, which causes the animals to roll is about 0.9 Mbp. This physical distance corresponds to about around their body axis [30], and the unc-52(e444) mutation 3 cM, assuming 300 kb per map unit, and encompasses located at the right end of the chromosome, which results in a between 100 and a maximum of 500 genes (Figure 2a). The paralyzed phenotype [31]. Mutant were

length of the amplicons ranges from 100 to 444 bp, and the crossed with CB4856 males, and wild-type F1 cross-progeny fragment length differences are between 1 and 21 bp (Addi- was selected (F1 self-progeny would exhibit a mutant pheno- tional data file 9). If necessary, another 2,454 predicted type). Finally, mutant self-progeny was isolated in the F2 gen- InDels are available to increase the mapping resolution down eration and used for genotyping (Figure 3a). To minimize the to 50 kbp on average (Additional data files 12-17). number of PCR reactions, we pursued a two-step strategy. First, we determined chromosomal linkage by analyzing 16

To establish a Drosophila FLP map, a set of 54 FLP assays (12 individual F2 animals (corresponding to 32 chromosomes in to 17 per arm of the two major ) was validated from total) with one centrally located FLP assay per chromosome the list of polymorphisms identified by Berger et al. [2] (Fig- (Tier 1, Figure 2a). This allowed us to establish clear linkage ure 2b, and Additional data file 10); high-resolution X-chro- to chromosome 2 for all three mutations (Additional data file mosomal SNP and FLP maps have yet to be established. 2). Surprisingly, the rol-1(e91) mutation showed linkage to Similarly to C. elegans, the confirmation rate of the predicted the X chromosome of N2 in addition to chromosome II. This Drosophila InDels was 80% (n = 30). Furthermore, another pseudo-linkage could be due to a suppressor of the Rol phe- 509 InDels are predicted at 248 sites for which an assay can notype present on the CB4856 X chromosome. In a second

be established to discriminate between EP and FRT strains step, 48 F2 animals for each mutation were analyzed with (Additional data file 18). The validated Drosophila FLP eight FLP assays along chromosome 2 (Tier 2, Figure 2a). In

C.Figure elegans 2 and(see Drosophilafollowing page) FLP maps C. elegans and Drosophila FLP maps. (a) The C. elegans FLP map. Marker names comprise a ZH prefix followed by the chromosome number and a unique identifier number. Markers used in first-level assays (Tier 1) for determination of chromosomal linkage are in red, those used for second-level assays (Tier 2) for higher resolution mapping are in black. (b) The Drosophila FLP map of chromosomes 2 and 3. The FRT sites and EP elements are symbolized by blue and green triangles, respectively. The strains that were genotyped are shown below each chromosome. Green indicates the EP genotype, blue the FRT genotypes and new alleles are shown in other colors.

Genome Biology 2005, 6:R19 http://genomebiology.com/2005/6/2/R19 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. R19.5

Assays used for (a) LG chromosomal linkage

ZH1-16ZH1-17 ZH1-10aZH1-25ZH1-07 ZH1-18a ZH1-03 ZH1-27 ZH1-34 ZH1-21 ZH1-01 ZH1-22ZH1-23 ZH1-15ZH1-05 ZH1-08ZH1-09ZH1-06ZH1-24 comment (tier 1) I

ZH2-15ZH2-04a ZH2-05ZH2-16ZH2-06aZH2-07ZH2-17ZH2-13 ZH2-19 ZH2-01ZH2-02 ZH2-20 ZH2-25 ZH2-27 ZH2-09 ZH2-28ZH2-10 ZH2-11ZH2-12ZH2-23 II

ZH3-17aZH3-25ZH3-06ZH3-07ZH3-08ZH3-26ZH3-28ZH3-15 ZH3-04 ZH3-02 ZH3-05a ZH3-32ZH3-35 ZH3-10aZH3-23ZH3-11ZH3-12ZH3-13 III reviews ZH4-04a ZH4-05 ZH4-06 ZH4-07ZH4-16 ZH4-08 ZH4-02ZH4-03ZH4-17 ZH4-18ZH4-09 ZH4-19 ZH4-20 ZH4-10aZH4-21ZH4-11 ZH4-12 ZH4-22 IV

ZH5-02a ZH5-13 ZH5-03a ZH5-14 ZH5-04ZH5-15 ZH5-05 ZH5-16 ZH5-01ZH5-17 ZH5-18ZH5-06 ZH5-11 ZH5-23 ZH5-12ZH5-20ZH5-08 ZH5-21 ZH5-09 ZH5-22 V

ZHX-16ZHX-17 ZHX-03 ZHX-08 ZHX-13ZHX-15 ZHX-10 ZHX-02 ZHX-24 ZHX-07 ZHX-12 ZHX-11 ZHX-05 ZHX-21a ZHX-06 ZHX-22ZHX-23 X reports Mb 0 5 10 15 20

EP0511 FRT40A FRT42D EP0755

(b) research deposited 2L017 2L0302L0382L0512L057 2L0692L0752L0882L0902L093 2L119 2L143 2R017 2R0392R0512R0602R068 2R083 2R096 2R1092R1182R1242R1302R139 2L CEN 2R 2.5 5 7.5 10 12.5 15 17.5 20 2.5 5 7.5 10 12.5 15 17.5 20 Mb

EP2L EP2R FRT2L FRT2R

FRT40A,w+, cl FRT42D,w+, cl EP2R EP2L

FRT2R FRT2L refereed research yw(WG) yw(WG) yw(GT1) yw(GT1)

EP3104 FRT80A FRT82B EP0381

3L0213L0313L0413L0583L0643L0763L0833L0863L0943L105 3L127 3L148 3R061 3R074 3R092 3R1223R1513R1603R1693R1863R1923R204 3R2213R224

3L CEN 3R 2.5 57.5 10 12.5 15 17.5 20 22.5 2.557.5 10 12.5 1517.5 20 22.5 25 27.5 Mb interactions

EP3L EP3R FRT3L FRT3R

FRT80A,w+, cl FRT82,w+, cl EP3R EP3L FRT3R FRT3L yw(WG) yw(WG) yw(GT1) yw(GT1) information

EP FRT Novel alleles No amplification

Figure 2 (see legend on previous page)

Genome Biology 2005, 6:R19 R19.6 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. http://genomebiology.com/2005/6/2/R19

this way, we could narrow down the three mutations to the arms [35], as well as the FRT and EP reference strains at both correct chromosomal subregions (Additional data files 3-5). relevant autosomal arms (Figure 2b). Surprisingly, the FRT We used the same strategy to map the zh41 mutation that was and EP lines are largely not of FRT or EP genotype on the identified in a forward genetic screen for mutants exhibiting chromosome arm for which they have not been calibrated. a loss of egl-17::gfp expression in the vulval cell linage ([32] Overall, we found novel alleles for 18 of the 48 assays, and in and I. Rimann and A. Hajnal, unpublished work). Analysis an extreme case, we even observed five different alleles in five with Tier 1 established linkage to chromosome 1 (Figure 3b), examined strains (2R017, Figure 2b). This result further high- and Tier 2 narrowed down the candidate region to an interval lights the heterogeneity of Drosophila strains (see Additional of 2.2 Mbp containing 498 genes (Figure 3c). The phenotype data file 1 for further details on FLP calibration and fly of zh41 animals is similar to the phenotype caused by loss-of- genetics). function mutations in lin-11, which maps to the same interval in the center of chromosome I [33]. Like lin-11 mutants, zh41 FLP mapping in Drosophila animals exhibit a penetrant protruding vulva (Pvl) pheno- In a genetic screen devised to isolate genes that regulate type, and staining of the adherens junctions with the MH27 growth and are situated on chromosome 2R, we found a com- antibody showed defects in the formation of the vulval torroid plementation group characterized by a mild overgrowth phe- rings (Figure 3d) [33]. Subsequent sequencing of the lin-11 notype (Figure 4b (2), and C. Rottig and E.H., unpublished in zh41 animals revealed a point mutation that results in work). From a cross between allele VI.29 and EP0755 we a change of leucine 274 to phenylalanine. Furthermore, zh41 recovered three types of recombinant chromosomes: failed to complement lin-11(n389), indicating that the zh41 recombinants with a crossover proximal or distal to the muta- mutation in the lin-11 open reading frame (ORF) is responsi- tion, respectively, and double-crossovers (Figure 4a, see also ble for the vulval phenotype. Additional data file 1 for further details on the crossing scheme). The mutation could be placed 16.9 cM proximal to In cases where a mutation maps to an interval that contains EP0755 and 38.7 cM distal to FRT42D. The FLPs in the no obvious candidate gene, we first screen for additional recombinant flies were directly analyzed without backcross- informative recombinants by FLP analysis and then refine the ing the recombinant chromosome into a parental strain back- map position by extracting more FLPs from our set of non- ground. DNA was prepared from recombinants by a novel validated InDels (Additional data files 12-17) and by high-throughput protocol (see Materials and methods). We genotyping existing SNPs in the candidate interval [3]. In genotyped 34 distal crossover events, 40 proximal crossovers, many cases, this resolution is sufficient to identify the and eight double-crossovers. This analysis placed the muta- affected gene through RNA interference (RNAi) analysis of tion between markers 2R096 and 2R109 (Figure 4c). This the genes in the corresponding interval [34]. (See Additional interval includes the tumor suppressor hippo [36], and subse- data file 6 for a detailed flowchart of the mapping process). quent complementation analysis confirmed VI.29 as a weak hippo allele (data not shown). Furthermore, data from this In summary, FLP mapping in C. elegans allows us to rapidly and other FLP mappings in this region allowed us to further map a mutation down to a small region containing, on aver- refine the genetic map (Additional data file 11). This kind of age, 200 candidate genes by crossing a mutant strain to experimental data is helpful to space new FLP assays more

CB4856 and analyzing 48 F2 animals with 300 to 500 PCR evenly on the genetic map should the available map turn out reactions. to be inexact.

Genotyping Drosophila strains with FLP assays If the resolution of the validated FLP map is too low to iden- In contrast to the well defined genetic backgrounds used for tify a candidate gene, we further refine the map position by C. elegans, zebrafish (Danio rerio) or Arabidopsis genetics, several approaches. First, we design novel FLP-assays in the Drosophila strains are very heterogeneous and of ill-defined region of interest and genotype the most informative recom- origin [2,9,11]. In this respect, gene mapping in Drosophila binants from the first round of FLP mapping (Additional data resembles human genetics in that standard inbred lines do file 18). Second, we genotype recombinants with SNPs avail- not exist and the genotypes of the parental lines have to be able in the region of interest and resolve them by RFLP, determined first. As genome-wide polymorphism databases sequencing or DHPLC [2,9]. Third, we perform complemen- for reference strains are available [2,11], a line of interest can tation analysis with recently established Drosophila lines be crossed with two reference strains, such as EP and FRT with molecularly defined deletions [37,38]. (See Additional (see below). Owing to the codominant character of sequence data file 7 for a detailed flowchart illustrating the mapping polymorphisms, at least one of the two respective crosses will process.) distinguish between the mutant and the mapping chromo- somes. To further facilitate mapping with our set of FLP assays, we genotyped several common laboratory lines such Conclusions as two 'wild-type' yw strains for the whole set, four FRT- We have developed an automated method to detect most nat- Minute or FRT-cell-lethal strains at the relevant autosomal urally occurring InDel polymorphisms at single base-pair res-

Genome Biology 2005, 6:R19 http://genomebiology.com/2005/6/2/R19 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. R19.7

(a) (c) Tier2 zh41 subchromosomal region comment N2 (Bristol) CB4856 (Hawaii) m* P x Lysate Informative Informative Recombinanats 0 ZH1-07 ZH1-03 ZH1-01 ZH1-15 ZH1-05 ZH1-08 ZH1-06 ZH1-10a m* Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 1 Hawaii Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 2 Hawaii Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Isolation of wild-type cross-progeny 3 Hawaii Hawaii Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 4 Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 5 Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 6 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 7 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii m* Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 8 Bristol Bristol Bristol Bristol Hawaii Hawaii Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol F1 9 Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol reviews Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 10 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii BristolBristolBristolBristolBristolBristolBristol 11 ND Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Isolation of mutant self-progeny 12 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii 13 Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 14 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 15 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii 16 Hawaii Hawaii Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 17 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol m* m* m* 18 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol F 19 ND NDHawaii Bristol ND BristolND Bristol 2 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 20 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii m* m* m* Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 21 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 22 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 23 Bristol Bristol Bristol BristolND Bristol Bristol Bristol reports Crossovers to Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Crossover to right Crossover to left 24 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol right and left of Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 25 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol of mutation of mutation Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol mutation 26 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 27 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 28 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 29 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol wild-type Bristol Bristol Bristol Bristol Bristol Bristol Bristol (b) Tier1 zh41 chromosomal linkage (d) 30 BristolND Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 31 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii Hawaii Hawaii 100% 32 Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 33 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol

Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii Hawaii research deposited 34 Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol 80% Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 35 Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 36 Bristol Bristol Bristol Bristol Bristol Bristol Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 37 Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol 60% Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 38 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 39 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Hawaii 40% 40 Hawaii Hawaii Hawaii Hawaii Bristol Bristol Bristol Bristol zh41 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol % Bristol 41 Hawaii Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 42 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 20% Bristol Bristol Bristol Bristol Bristol Bristol 43 Hawaii Hawaii Hawaii BristolND Bristol ND Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 44 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 0% 45 Hawaii Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 46 ND Bristol Bristol Bristol Bristol Bristol Bristol Hawaii 12345X Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol 47 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol Chromosome refereed research 48 Bristol Bristol Bristol Bristol Bristol Bristol Bristol Bristol

FLPFigure mapping 3 in C. elegans FLP mapping in C. elegans. (a) Crossing scheme used to map mutations generated in the N2 Bristol background. The different classes of recombinants recovered in the F2 generation are shown. (b) Analysis of the zh41 mutation with Tier 1 assays establishes linkage to chromosome I. (c) Analysis with Tier 2 places zh41 between assays ZH1-01 and ZH1-15. ND, no data as a result of PCR reaction failure. (d) Ventral views of the vulva in wild-type and zh41 L4 larvae stained with the adherens junction antibody MH27 [44]. In the wild type, the vulval cells have fused to generate the torroids that appear as concentric rings. zh41 mutants exhibit the same fusion defects observed in other lin-11 alleles [33].

olution. Since a significant fraction of polymorphisms are informative recombinants is usually the rate-limiting step, interactions caused by InDels of only a few base pairs (for example, 8% to FLP mapping is very helpful in extracting the few relevant 20% in humans [22]) the resolution of the medium-density recombinants from a large number of samples. Third, thanks FLP maps can be greatly increased where necessary, for to the standardized conditions, the low error rate and the example during the positional cloning of genes. We are there- absence of a secondary assay, FLP mapping is considerably fore continually designing new FLP assays according to our cheaper than the previously published 'manual' mapping specific needs using the predicted FLPs (Additional data files methods [2,3]. Unlike other high-throughput methods like 12-18). The full automation of the genotyping has three main TaqMan, Pyrosequencing, DHPLC, fluorescence polarization advantages when compared to manual methods. First, the or primer-extension assays, FLP mapping does not require information error rate (the number of wrongly assigned genotypes) is any investment in specialized equipment. It can be done in extremely low, as it was not measurable in 432 assays. Sec- any molecular biology lab with access to a sequencing facility ond, genotyping can be done very rapidly and at a high- equipped with a capillary- or gel-based system, which usually throughput with little manpower. The automatic allele-call- includes the genotyping software. PCR costs are marginally ing, in particular, saves much time. As the identification of

Genome Biology 2005, 6:R19 R19.8 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. http://genomebiology.com/2005/6/2/R19

- 1 (a) m FRT EP (b) × CEN F0 CEN Balancer

Isolation of EP/FRT virgins hpo42-20

2 - m × CEN F1 CEN cl*

VI.29 Isolation of red-eyed Isolation of red-eyed Isolation of white-eyed hpo mutant mosaics wild-type mosaics wild-type mosaics 3

- + + m M M

F CEN CEN 2 CEN cl* cl* cl* yw

Crossovers Crossover distal Crossover proximal distal and proximal to mutation to mutation to mutation

(c) Cross Recombinant number 2R017 2R039 2R051 2R060 2R068 2R083 2R090 2R096 2R109 2R118 2R124 2R130 2R139 Cross Recombinant number 2R017 2R039 2R051 2R060 2R068 2R083 2R090 2R096 2R109 2R118 2R124 2R130 2R139

Double R1 FRT FRT FRT ND FRT FRT FRT EP EP EP EP EP FRT Proximal R1 FRT FRT FRT FRT EP EP EP ND ND ND ND ND ND crossover R2 FRT FRT FRT EP EP EP EP EP EP FRT FRT FRT FRT crossover R2 FRT FRT FRT EP EP EP EP ND ND ND ND ND ND R5 FRT FRT FRT EP EP EP EP EP EP FRT FRT FRT FRT R3 FRT FRT FRT ND FRT EP EP ND ND ND ND ND ND R7 FRT FRT EP ND EP EP EP EP EP EP FRT FRT FRT R4 FRT EP EP ND EP ND EP ND ND ND ND ND ND R8 EP EP EP ND EP EP EP EP EP EP EP EP FRT R5 FRT FRT FRT ND FRT EP EP ND ND ND ND ND ND R11 FRT FRT FRT ND FRT FRT FRT FRT EP EP EP EP FRT R6 FRT FRT FRT ND FRT FRT EP ND ND ND ND ND ND R12 EP EP EP ND EP EP EP EP EP EP FRT FRT FRT R7 FRT EP EP ND EP EP EP ND ND ND ND ND ND R13 FRT FRT EP ND EP EP EP EP EP EP FRT FRT FRT R8 EP EP EP ND EP EP EP ND ND ND ND ND ND R9 FRT FRT FRT EP EP EP EP ND ND ND ND ND ND Distal R2 ND ND ND FRT ND ND FRT FRT ND FRT FRT FRT FRT R10 FRT FRT FRT EP EP EP EP ND ND ND ND ND ND crossover R3 ND ND ND ND FRT ND FRT FRT FRT FRT FRT EP EP R11 FRT FRT FRT EP EP EP EP ND ND ND ND ND ND R4 ND ND ND ND ND ND FRT FRT FRT ND FRT ND FRT R12 FRT FRT EP ND EP EP EP ND ND ND ND ND ND R5 ND ND ND ND ND ND FRT FRT FRT FRT FRT EP EP R13 FRT FRT FRT ND FRT FRT FRT FRT EP ND ND ND ND R6 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT EP R14 FRT EP EP ND EP EP EP ND ND ND ND ND ND R7 ND ND ND ND ND ND FRT FRT FRT FRT EP EP EP R15 FRT FRT FRT ND FRT EP EP ND ND ND ND ND ND R8 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT EP R16 FRT FRT FRT EP EP EP EP ND ND ND ND ND ND R9 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT FRT R17 FRT FRT FRT ND FRT EP EP ND ND ND ND ND ND R10 ND ND ND ND ND ND FRT FRT EP EP EP EP EP R18 FRT FRT FRT ND FRT FRT FRT FRT EP ND ND ND ND R11 ND ND ND ND ND ND FRT FRT FRT FRT EP EP EP R19 FRT FRT EP ND ND EP EP ND ND ND ND ND ND R12 ND ND ND ND ND ND FRT FRT EP EP EP EP EP R20 FRT FRT FRT EP EP ND EP ND ND ND ND ND ND R13 ND ND ND ND ND ND FRT FRT FRT FRT FRT ND FRT R21 EP EP EP ND EP EP EP ND ND ND ND ND ND R14 ND ND ND ND ND ND FRT FRT FRT EP EP ND EP R22 FRT FRT FRT EP EP EP EP ND ND ND ND ND ND R15 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT EP R23 FRT FRT FRT ND FRT FRT EP ND ND ND ND ND ND R16 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT EP R24 FRT FRT FRT FRT EP EP EP ND ND ND ND ND ND R17 ND ND ND ND ND ND FRT FRT FRT ND FRT EP EP R25 FRT FRT EP ND EP EP EP ND ND ND ND ND ND R18 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT FRT R26 FRT FRT FRT FRT EP EP EP ND ND ND ND ND ND R19 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT EP R27 FRT FRT FRT ND FRT FRT FRT FRT EP ND ND ND ND R20 ND ND ND ND ND ND FRT FRT FRT ND FRT EP ND R28 FRT FRT FRT FRT EP EP EP ND ND ND ND ND ND R21 ND ND ND ND ND ND FRT FRT FRT FRT ND FRT FRT R29 FRT FRT FRT EP EP EP EP ND ND ND ND ND ND R22 ND ND ND ND ND ND FRT FRT FRT EP ND EP EP R30 FRT FRT FRT ND FRT EP EP ND ND ND ND ND ND R23 ND ND ND ND ND ND FRT FRT FRT FRT EP EP EP R31 FRT FRT FRT ND ND FRT FRT FRT EP ND ND ND ND R24 ND ND ND ND ND ND FRT FRT FRT EP EP EP EP R32 FRT EP EP ND EP EP EP ND ND ND ND ND ND R25 ND ND ND ND ND ND FRT FRT FRT EP EP EP EP R26 ND ND FRT ND ND ND FRT FRT FRT EP EP EP EP More recs R10 EP EP EP ND EP EP EP EP EP EP EP EP EP R27 ND ND ND ND ND ND FRT FRT FRT FRT FRT EP EP R14 FRT FRT EP ND EP EP EP EP EP EP EP EP EP R28 ND ND ND ND ND ND FRT FRT FRT FRT EP EP EP R15 FRT FRT FRT ND FRT FRT FRT EP EP EP EP EP EP R29 ND ND ND ND ND ND FRT FRT FRT FRT FRT FRT FRT R16 FRT FRT FRT EP EP EP EP EP EP EP EP EP EP R30 ND ND ND ND ND ND FRT FRT FRT FRT FRT EP EP R3 FRT EP EP ND EP EP EP EP EP EP EP EP EP 29--1 FRT FRT FRT ND FRT FRT FRT FRT ND FRT ND FRT FRT R4 FRT FRT FRT ND FRT EP EP EP EP EP EP EP EP 29--2 FRT FRT FRT ND FRT FRT FRT FRT ND FRT FRT EP EP R6 FRT EP EP ND EP EP EP EP EP EP EP EP EP 29--3 FRT FRT FRT ND FRT FRT FRT FRT ND FRT EP EP EP R9 EP EP EP ND EP EP EP EP EP EP EP EP EP 29--4 FRT FRT FRT ND FRT FRT FRT FRT FRT FRT ND FRT EP 29--5 FRT FRT FRT ND FRT FRT FRT FRT FRT EP EP EP EP 29--8 ND ND ND ND ND ND ND ND FRT EP EP ND ND

Figure 4 (see legend on next page)

Genome Biology 2005, 6:R19 http://genomebiology.com/2005/6/2/R19 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. R19.9

FigureFLP mapping 4 (see in previous Drosophila page) FLP mapping in Drosophila. (a) Crossing scheme used to map mutations generated in the FRT background and recombined with an EP line. The different 42-20 comment classes of recombinants recovered in the F2 generation are shown. (b) Big head phenotypes of the hippo null allele hpo (1) and the VI.29 mutation (2). A wild-type control is shown in (3). (c) FLP mapping of the VI.29 mutation on chromosome 2R. Analysis of the different classes of recombinants places the mutation between markers 2R096 and 2R109 (dashed red line). Informative recombinants are boxed in red. ND, not determined or no data as a result of PCR reaction failure. higher because of the use of fluorescently labeled primers, but centration of proteins that might be harmful for the capillary there are no added expenses for secondary enzymatic assays. sequencer.

It seems likely that in most organisms the frequency of poly- PCR and FLP fragment analysis reviews morphisms caused by InDels is in the same range as found in Diluted single-worm lysates (2 µl samples) or single fly humans, C. elegans or Drosophila. For example, 7.3% of the extracts were added to 23 µl PCR reaction mix. Final concen- Arabidopsis sequence polymorphisms are InDels [39]. Thus, trations in the PCR reaction were: 0.4 µM forward/reverse

FLP mapping can easily be adapted to any organism for which primer, 0.2 mM dNTPs, 2 mM MgCl2, 1x PCR reaction buffer, polymorphism maps have been established, as there is no 0.25 U EuroTaq polymerase (Euroclone). PCR reaction setup conceptual difference between human, Arabidopsis, C. ele- was done in 96-well plates using a Tecan Genesis pipetting gans or Drosophila FLPs. robot with disposable tips. PCR was carried out in two MJR

thermo-cyclers that are integrated into the robot. The current reports setup allows for the sequential processing of six 96-well plates Materials and methods at a time. Cycling parameters were 2 min 95°C, 20 sec 95°C, C. elegans and Drosophila culture techniques and alleles 20 sec 61°C (-0.5°C for each cycle), 45 sec 72°C (for 10 cycles) Culturing and crossing of C. elegans was done according to followed by 24 cycles of 20 sec 95°C, 20 sec 56°C, 45 sec 72°C standard procedures described in [26]. C. elegans alleles used and a 10 min 72°C final extension. Following PCR, reactions

were: LG I: lin-11(zh41), lin-11(n389); LG II: rol-1(e91), let- were diluted 1:100 in water, and 2 µl diluted PCR products research deposited 23(sy1), unc-52(e444). Drosophila strains and the genetic were mixed with 10 µl HiDi formamide containing 0.025 µl screen have been described previously [9,35,40-42]. LIZ500 size standard (Applied Biosystems). This dilution before analysis on the capillary sequencer is necessary to Single worm DNA extraction reduce signal intensity because too strong signals Adult worms were collected in 10 µl lysis buffer (50 mM KCl, compromise data analysis. In addition, sample dilution

10 mM Tris pH 8.2, 2.5 mM MgCl2, 0.45% NP-40, 0.45% reduces the risk of damaging the capillaries with proteins or Tween-20, 100 µg/ml freshly added proteinase K) and incu- lipids present in the crude lysates. The dilution was done with bated for 60 min at 65°C followed by heat-inactivation of pro- standard tips using the Tecan Genesis pipetting station. Car- refereed research teinase K at 95°C for 10 min. Before PCR, 90 µl double- ryover of fragments was prohibited by a simple wash step distilled H2O (ddH2O) was added to obtain a total volume of with H2O. Fragments were analyzed on an ABI3730 capillary 100 µl per lysate. sequencer using POP7 polymer according to standard proce- dures. Data were analyzed using AppliedBiosystems GeneM- Fly DNA extraction apper software and raw data were treated further with DNA from recombinant flies was extracted in bulk by squish- Microsoft Excel. ing flies through mechanical force in a vibration mill (Retsch MM30) programmed to shake for 20 sec at 20 strokes per sec-

ond [43]. Single flies were placed into wells of a 96-well for- Additional data files interactions mat deep-well plate with each well filled with 200 µl The following additional data are available with the online squishing buffer (10 mM Tris-Cl pH 8.2, 1 mM EDTA, 0.2% version of this article. Additional data file 1 contains general Triton X-100, 25 mM NaCl, 200 µg/ml freshly added protei- information on fly genetics. nase K) and a tungsten carbide bead (Qiagen). The deep-well plate was then sealed with a rubber mat (Eppendorf) and Further C. elegans mapping results are given in Additional clamped into the vibration mill. (Tungsten carbide beads can data files 2,3,4 and 5. Detailed flowcharts illustrating the FLP be recycled: after an overnight incubation in 0.1 M HCl and mapping process are shown in Additional data files 6 and 7.

thorough washing in ddH2O the beads are virtually free of Additional data file 8 contains electropherograms demon- information contaminating DNA.) Debris was allowed to settle for about 5 strating the accuracy of allele-calling. Additional data files 9 min, and 50 µl of each supernatant were transferred into a 96- and 10 contain tables of primer and sequence data of experi- well PCR plate. The reactions were incubated in a thermo- mentally verified FLP assays in C. elegans and Drosophila, cycler for 30 min at 37°C and finally for 10 min at 95°C to respectively. Additional data file 11 contains a table of the heat-inactivate proteinase K. Before PCR amplification, the refined genetic distances for FLP assays on the right arm of crude DNA extracts were diluted 20-fold to reduce the con- Drosophila chromosome 2. Additional non-validated FLPs

Genome Biology 2005, 6:R19 R19.10 Genome Biology 2005, Volume 6, Issue 2, Article R19 Zipperlen et al. http://genomebiology.com/2005/6/2/R19

can be found in Additional data files 12,13,14,15,16 and 17 (C. 21. Collins JR, Stephens RM, Gold B, Long B, Dean M, Burt SK: An elegans) and Additional data file 18 (Drosophila). exhaustive DNA micro-satellite map of the human genome using high performance computing. Genomics 2003, 82:10-9. ClickAdditionalDetailedElectropherogramsTablesassaysAarmFurtherGeneral table of here in ofDrosophilaof C. information flowchartsprimerC.Drosophila the forelegansdatanon-validated elegans refinedadditional file and chromosomemapping 678954312101113141516171812demonstratingillustrating onsequencegenetic fly data FLPs genetics results distancesfile data (the C.Drosophila2 elegansthe FLPof experimentally accuracyfor mapping )FLP) assays of process allele-calling verifiedon the right FLP 22. Weber JL, David D, Heil J, Fan Y, Zhao C, Marth G: Human diallelic insertion/deletion polymorphisms. Am J Hum Genet 2002, 71:854-862. Acknowledgements 23. Smith JR, Carpten JD, Brownstein MJ, Ghosh S, Magnuson VL, Gilbert We are grateful to Carmen Rottig for providing us with the novel hippo DA, Trent JM, Collins FS: Approach to genotyping errors mutant and to DJ Pan for the hpo42-20 mutation. Angela Baer is acknowl- caused by nontemplated nucleotide addition by Taq DNA edged for excellent technical assistance. This work was funded by projects polymerase. Genome Res 1995, 5:312-317. from the Swiss National Science Foundation and the Kanton Zürich. 24. Ranade K, Chang MS, Ting CT, Pei D, Hsiao CF, Olivier M, Pesich R, Hebert J, Chen YD, Dzau VJ, et al.: High-throughput genotyping with single nucleotide polymorphisms. Genome Res 2001, 11:1262-1268. References 25. Chen DC, Saarela J, Nuotio I, Jokiaho A, Peltonen L, Palotie A: Com- 1. Sachidanandam R, Weissman D, Schmidt SC, Kakol JM, Stein LD, parison of GenFlex Tag array and pyrosequencing in SNP Marth G, Sherry S, Mullikin JC, Mortimore BJ, Willey DL, et al.: A genotyping. J Mol Diagn 2003, 5:243-249. map of human genome sequence variation containing 1.42 26. Brenner S: The genetics of Caenorhabditis elegans. Genetics 1974, million single nucleotide polymorphisms. Nature 2001, 77:71-94. 409:928-933. 27. Lindsley DL, Zimm GG: The genome of Drosophila melanogaster New 2. Berger J, Suzuki T, Senti KA, Stubbs J, Schaffner G, Dickson BJ: York: Academic Press; 1992. Genetic mapping with SNP markers in Drosophila. Nat Genet 28. Adams MD, Celniker SE, Holt RA, Evans CA, Gocayne JD, Amanati- 2001, 29:475-481. des PG, Scherer SE, Li PW, Hoskins RA, Galle RF, et al.: The 3. Wicks SR, Yeh RT, Gish WR, Waterston RH, Plasterk RH: Rapid genome sequence of Drosophila melanogaster. Science 2000, gene mapping in Caenorhabditis elegans using a high density 287:2185-2195. polymorphism map. Nat Genet 2001, 28:160-164. 29. Aroian RV, Sternberg PW: Multiple functions of let-23, a 4. Kwok PY, Chen X: Detection of single nucleotide Caenorhabditis elegans tyrosine kinase gene polymorphisms. Curr Issues Mol Biol 2003, 5:43-60. required for vulval induction. Genetics 1991, 128:251-267. 5. Syvanen AC: Accessing genetic variation: genotyping single 30. Higgins BJ, Hirsh D: Roller mutants of the nematode nucleotide polymorphisms. Nat Rev Genet 2001, 2:930-942. Caenorhabditis elegans. Mol Gen Genet 1977, 150:63-72. 6. Chen X, Sullivan PF: Single nucleotide polymorphism genotyp- 31. Mackenzie JM, Garcea RL Jr, Zengel JM, Epstein HF: Muscle devel- ing: biochemistry, protocol, cost and throughput. Pharmacoge- opment in Caenorhabditis elegans: mutants exhibiting nomics J 2003, 3:77-96. retarded sarcomere construction. Cell 1978, 15:751-762. 7. Livak KJ: Allelic discrimination using fluorogenic probes and 32. Burdine RD, Branda CS, Stern MJ: EGL-17(FGF) expression coor- the 5' nuclease assay. Genet Anal 1999, 14:143-149. dinates the attraction of the migrating sex myoblasts with 8. Wolford JK, Blunt D, Ballecer C, Prochazka M: High-throughput vulval induction in C. elegans. Development 1998, 125:1083-1093. SNP detection by using DNA pooling and denaturing high 33. Gupta BP, Wang M, Sternberg PW: The C. elegans LIM home- performance liquid chromatography (DHPLC). Hum Genet obox gene lin-11 specifies multiple cell fates during vulval 2000, 107:483-487. development. Development 2003, 130:2589-2601. 9. Nairz K, Stocker H, Schindelholz B, Hafen E: High-resolution SNP 34. Kamath RS, Fraser AG, Dong Y, Poulin G, Durbin R, Gotta M, Kana- mapping by denaturing HPLC. Proc Natl Acad Sci USA 2002, pin A, LeBot N, Moreno S, Sohrmann M, et al.: Systematic func- 99:10575-10580. tional analysis of the Caenorhabditis elegans genome using 10. White R, Lalouel JM: Chromosome mapping with DNA RNAi. Nature 2003, 421:231-237. markers. Sci Am 1988, 258:40-48. 35. Newsome TP, Asling B, Dickson BJ: Analysis of Drosophila pho- 11. Hoskins RA, Phan AC, Naeemuddin M, Mapa FA, Ruddy DA, Ryan JJ, toreceptor axon guidance in eye-specific mosaics. Develop- Young LM, Wells T, Kopczynski C, Ellis MC: Single nucleotide pol- ment 2000, 127:851-860. ymorphism markers for genetic mapping in Drosophila 36. Ryoo HD, Steller H: Hippo and its mission for growth control. melanogaster. Genome Res 2001, 11:1100-1113. Nat Cell Biol 2003, 5:853-855. 12. Syvanen AC: From gels to chips: "minisequencing" primer 37. Ryder E, Blows F, Ashburner M, Bautista-Llacer R, Coulson D, Drum- extension for analysis of point mutations and single nucle- mond J, Webster J, Gubb D, Gunton N, Johnson G, et al.: The Dros- otide polymorphisms. Hum Mutat 1999, 13:1-10. Del collection: a set of P-element insertions for generating 13. Kwok PY: High-throughput genotyping assay approaches. custom chromosomal aberrations in Drosophila Pharmacogenomics 2000, 1:95-100. melanogaster. Genetics 2004, 167:797-813. 14. Swan KA, Curtis DE, McKusick KB, Voinov AV, Mapa FA, Cancilla 38. Parks AL, Cook KR, Belvin M, Dompe NA, Fawcett R, Huppert K, Tan MR: High-throughput gene mapping in Caenorhabditis LR, Winter CG, Bogart KP, Deal JE, et al.: Systematic generation elegans. Genome Res 2002, 12:1100-1105. of high-resolution deletion coverage of the Drosophila mela- 15. Sosnowski RG, Tu E, Butler WF, O'Connell JP, Heller MJ: Rapid nogaster genome. Nat Genet 2004, 36:288-292. determination of single base mismatch mutations in DNA 39. Schmid KJ, Sorensen TR, Stracke R, Torjek O, Altmann T, Mitchell- hybrids by direct electric field control. Proc Natl Acad Sci USA Olds T, Weisshaar B: Large-scale identification and analysis of 1997, 94:1119-1123. genome-wide single-nucleotide polymorphisms for mapping 16. Gilles PN, Wu DJ, Foster CB, Dillon PJ, Chanock SJ: Single nucle- in Arabidopsis thaliana. Genome Res 2003, 13:1250-1257. otide polymorphic discrimination by an electronic dot blot 40. Xu T, Rubin GM: Analysis of genetic mosaics in developing and assay on semiconductor microchips. Nat Biotechnol 1999, adult Drosophila tissues. Development 1993, 117:1223-1237. 17:365-370. 41. Rorth P: A modular misexpression screen in Drosophila 17. Weissenbach J: Microsatellite polymorphisms and the genetic detecting tissue-specific phenotypes. Proc Natl Acad Sci USA linkage map of the human genome. Curr Opin Genet Dev 1993, 1996, 93:12418-12422. 3:414-417. 42. St Johnston D: The art and design of genetic screens: Dro- 18. A comprehensive genetic linkage map of the human sophila melanogaster. Nat Rev Genet 2002, 3:176-188. genome. NIH/CEPH Collaborative Mapping Group. Science 43. Nairz K, Zipperlen P, Dearolf C, Basler K, Hafen E: A reverse 1992, 258:67-86. genetic screen in Drosophila using a deletion-inducing 19. McCouch SR, Chen X, Panaud O, Temnykh S, Xu Y, Cho YG, Huang mutagen. Genome Biol 2004, 5:R83. N, Ishii T, Blair M: Microsatellite marker development, map- 44. Francis R, Waterston RH: Muscle cell attachment in Caenorhab- ping and applications in rice genetics and breeding. Plant Mol ditis elegans. J Cell Biol 1991, 114:465-479. Biol 1997, 35:89-99. 20. Knapik EW, Goodman A, Ekker M, Chevrette M, Delgado J, Neuhauss S, Shimoda N, Driever W, Fishman MC, Jacob HJ: A microsatellite genetic linkage map for zebrafish (Danio rerio). Nat Genet 1998, 18:338-343.

Genome Biology 2005, 6:R19