UBC Faculty Research and Publications

Spectrum of variations in dog-1/FANCJ and mdf-1/MAD1 defective Caenorhabditis elegans strains after long-term… Tarailo-Graovac, Maja; Wong, Tammy; Qin, Zhaozhao; Flibotte, Stephane; Taylor, Jon; Moerman, Donald G; Rose, Ann M; Chen, Nansheng Mar 18, 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-12864_2015_Article_1402.pdf [ 2.36MB ]
JSON: 52383-1.0223685.json
JSON-LD: 52383-1.0223685-ld.json
RDF/XML (Pretty): 52383-1.0223685-rdf.xml
RDF/JSON: 52383-1.0223685-rdf.json
Turtle: 52383-1.0223685-turtle.txt
N-Triples: 52383-1.0223685-rdf-ntriples.txt
Original Record: 52383-1.0223685-source.json
Full Text

Full Text

Spectrum of variations in dog-1/FANCJ andmdf-1/MAD1 defective Caenorhabditis elegansstrains after long-term propagationTarailo-Graovac et al.Tarailo-Graovac et al. BMC Genomics  (2015) 16:210 DOI 10.1186/s12864-015-1402-yRESEARCH ARTICLESpectrum of variations ineo1,Background (SAC) missing and this leads to accumulation of geneticTarailo-Graovac et al. BMC Genomics  (2015) 16:210 DOI 10.1186/s12864-015-1402-yture [7-9], which can pose a barrier to replication forkVancouver, BC, CanadaFull list of author information is available at the end of the articleGenome integrity is crucial for survival of all living organ-isms. Chromosomal instability (CIN), marked by whole orsegmental aneuploidy is hallmark of human tumors, andcan drive abnormal proliferation of cancer cells [1]. In Cae-norhabditis elegans, mdf-1(gk2) has an essential mdf-1/MAD-1 component of the spindle assembly checkpointerrors and ultimately death by the third generation [2].The checkpoint prevents CIN by inhibiting anaphase-promoting complex/cyclosome (APC/C), and delaying ana-phase onset until all the chromosomes have achievedproper attachment to the spindle [3].While MDF-1 prevents both loss and gain of wholechromosomes during mitosis, DOG-1 prevents seg-mental aneuploidies by ensuring proper replication ofguanine(G)-rich DNA [4-6]. G-rich DNA can adopt afour-stranded helical G-quadruplex (G4) DNA struc-* Correspondence: maja@cmmt.ubc.ca; chenn@sfu.ca1Department of Molecular Biology and Biochemistry, Simon Fraser University,V5A 1S6 Burnaby, BC, Canada3Department of Medical Genetics, University of British Columbia, V6T 1Z3AbstractBackground: Whole and partial chromosome losses or gains and structural chromosome changes are hallmarks ofhuman tumors. Guanine-rich DNA, which has a potential to form a G-quadruplex (G4) structure, is particularly vulnerableto changes. In Caenorhabditis elegans, faithful transmission of G-rich DNA is ensured by the DOG-1/FANCJ deadboxhelicase.Results: To identify a spectrum of mutations, after long-term propagation, we combined whole genome sequencing(WGS) and oligonucleotide array Comparative Genomic Hybridization (oaCGH) analysis of a C. elegans strain that waspropagated, in the absence of DOG-1 and MDF-1/MAD1, for a total of 470 generations, with samples taken for long termstorage (by freezing) in generations 170 and 270. We compared the genomes of F170 and F470 strains and identified 94substitutions, 17 InDels, 3 duplications, and 139 deletions larger than 20 bp. These homozygous variants were predictedto impact 101 protein-coding genes. Phenotypic analysis of this strain revealed remarkable fitness recovery indicating thatmutations, which have accumulated in the strain, are not only tolerated but also cooperate to achieve long-termpopulation survival in the absence of DOG-1 and MDF-1. Furthermore, deletions larger than 20 bp were the only variantsthat frequently occurred in G-rich DNA. We showed that 126 of the possible 954 predicted monoG/C tracts, larger than14 bp, were deleted in unc-46 mdf-1 such-4; dog-1 F470 (JNC170).Conclusions: Here, we identified variants that accumulated in C. elegans’ genome after long-term propagation in theabsence of DOG-1 and MDF-1. We showed that DNA sequences, with G4-forming potential, are vulnerable todeletion-formation in this genetic background.Keywords: C. elegans, Whole genome sequencing (WGS), oligonucleotide array Comparative Genomic Hybridization(oaCGH), Mutation accumulation (MA), Genomic variation (GV), Spindle assembly checkpoint (SAC), dog-1/FANCJ,G-quadruplex (G4) structuremdf-1/MAD1 defective Castrains after long-term prMaja Tarailo-Graovac1,3,4,5*, Tammy Wong1, Zhaozhao QinAnn M Rose3 and Nansheng Chen1*© 2015 Tarailo-Graovac et al.; licensee BioMedCreative Commons Attribution License (http:/distribution, and reproduction in any mediumDomain Dedication waiver (http://creativecomarticle, unless otherwise stated.Open Accessdog-1/FANCJ andnorhabditis eleganspagationStephane Flibotte2, Jon Taylor2, Donald G Moerman2,Central. This is an Open Access article distributed under the terms of the/creativecommons.org/licenses/by/4.0), which permits unrestricted use,, provided the original work is properly credited. The Creative Commons Publicmons.org/publicdomain/zero/1.0/) applies to the data made available in thisTarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 2 of 10progression if left unresolved. The ability to form G4structures makes the corresponding G-rich DNAsequences particularly vulnerable to chromosomal rear-rangements. Studies using C. elegans, as a model organ-ism, were the first to show the striking genomicinstability of G-rich DNA sequences when DOG-1, afunctional ortholog of the deadbox helicase FANCJ[10], was non-functional [4]. When DOG-1 is func-tional, G-rich DNA sequences are stable and deletionsaffecting these regions are not observed [4,5]. Genome-wide bioinformatic analysis of the human genome hadidentified more than 300,000 DNA sites with G4-formingpotential [11,12]. In humans, mutations in FANCJ/dog-1have been identified in Fanconi anemia (FA) complemen-tation group J patients [13-15], which is a severe, auto-somal recessive, disorder with increased spontaneous andDNA crosslink-induced CIN showing a wide range of clin-ical manifestations [13], and also in early onset breast can-cer patients [16,17].Knowledge of the mutational spectra is crucial fordeciphering the cause of heritable genetic disorders aswell as the progression of events relevant to cancers.Traditionally, analyses of mutation spectra and rateshave been based on a small portion of phenotypicallyand molecularly detectable loci. In C. elegans, the muta-tional spectrum of dog-1(gk10) (knockout allele of dog-1) strains, was analyzed using either PCR-based assays[4,10,18] or oligonucleotide array Comparative GenomicHybridization (oaCGH) [5,6,19]. The rapid advances in“next-generation” DNA sequencing technologies nowallows us to perform comprehensive genome-wide ana-lyses of mutational spectra by sequencing whole ge-nomes [20-23].Here we undertook a whole genome approach inorder to analyze mutational events in a C. elegans strainthat is defective for both mdf-1/MAD1 and dog-1/FANCJ. This strain was propagated for 470 generationsand samples were stored frozen at generations 170 and270. Phenotypic analysis of the strain revealed strikingfitness recovery, indicating that accumulated mutationscooperate to bypass the MDF-1 checkpoint require-ment and thus achieve long-term population survival.We performed whole-genome sequencing (WGS) andoaCGH analyses of the strain at three different genera-tions (170, 270 and 470). We identified substitutions,InDels, and copy number variants (CNVs) larger than20 bp, and compared their accumulation over the gen-erations. We showed that only deletions, which are lar-ger than 20 bp, frequently initiated in G-rich DNA(88% of all of the deletions). Consistent with the fitnessrecovery observed in this strain, rather than a declinein fitness, the mutation spectrum reported here reflectsvariants that are either advantageous or neutral in thisspecific genetic background.Results and discussionWhole genome analysisIn order to propagate mdf-1; dog-1 homozygous worms,it is first necessary to isolate a suppressor of mdf-1(gk2)sterility and lethality that occurs in the double mutant[24]. Previously, such-4(h2168) was isolated, whichallowed propagation of mdf-1; dog-1 beyond the thirdgeneration (Figure 1) [24]. The such-4 suppressor allowsfor an approximately five-fold increase in fertile herm-aphrodite progeny of mdf-1; dog-1 [24]. This increase infertility occurs in the generations immediately after iso-lation of the suppressor. We outcrossed one worm toobtain KR4233 [unc-46(e177) mdf-1(gk2) such-4(h2168)][24,25] (Figure 1). A second line was isolated and main-tained in the dog-1(gk10) background for 470 genera-tions, with storage by freezing at generations 170 and270 (Figure 1). To mark the presence of mdf-1(gk2), unc-46(e177) (a visible marker) was used, which is present inall of our strains (Figure 1). While propagating unc-46mdf-1 such-4; dog-1 homozygotes, we observed a furtherincrease in reproductive fitness. This increase was sig-nificant, 59% of F470 unc-46 mdf-1 such-4; dog-1 progenydevelop into fertile hermaphrodites, compared to only10% of the unc-46 mdf-1 such-4 progeny and 2% of unc-46 mdf-1 mutants. Detailed phenotypic analysis of thesestrains as well as genetic dissection of suppressors hasrecently been reported [26]. To identify the genomic var-iations (GVs) that had accumulated in mdf-1 such-4;dog-1 F470 worms after long-term propagation, the gen-ome was sequenced to a depth of 70× genome equiva-lents and aligned to the C. elegans reference genomeWS235 available at WormBase [27]. We also sequencedthe unc-46 mdf-1 such-4; dog-1 strains that were frozenat generations 170 and 270 and compared the progressof mutation accumulation (Additional file 1: Figure S1).Single Base Substitutions (SBSs) do not occur withinG4-DNAWe used the variant caller VarScan2; version 2.3.2 [28]to identify 525 homozygous SBSs that occurred with avariant frequency (VF) of 90% or higher in the mdf-1such-4; dog-1F470 genome (Additional file 2: Table S1A).All 42 tested substitutions were confirmed by Sanger re-sequencing (30 were randomly selected; 12 additionalsubstitutions were confirmed by re-sequencing as a re-sult of being adjacent to randomly selected SBSs or asbeing non-randomly selected as candidates in later ana-lyses), indicating a false positive rate of less than 5%. WGSanalysis of the unc-46 mdf-1 such-4; dog-1 strain, whichhad been frozen at generation 170 (Figure 1), revealed thatthe majority (431/525) of the SBSs present in generation470 are also present in generation 170 (Additional file 1:Figure S1A; Table 1). However, we did observe that 32additional substitutions had accumulated between F170Tarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 3 of 10and F270, and 62 more between F270 and F470 (Table 1).Large number of substitutions observed in F170 indicatethe possibility that the original unc-46 mdf-1 such-4; dog-1strain (Figure 1) had a large number of single nucleotidedifferences from the reference genome (WS235). One wayto test this possibility is to estimate the mutation ratesbased on available data. If we consider the 94 substitutionsthat had accumulated in 300 generations, between F170and F470, we estimate the rate of μbs = 3.1 × 10-9/base/generation. This estimate is similar to the previouslyFigure 1 A schematic representation of the long-term propagation ofWe isolated a suppressor (such-4) in F4 from the only plate that we set upthree generations. We generated a strain (KR4233) by crossing away dog-1(The worms were frozen at generations 170 (green), 270 (purple) and 470 (bmdf-1(gk2) in all of our strains.Table 1 Summary of the variants identified in the unc-46 mdfSBSs InDels ≤ 20 bpF170 431 133F270 32 8F470 62 9Total F170-F470 94 17*Two new duplications and amplification of the cyb-3 locus to three copies.reported spontaneous rate of base substitution in C. ele-gans, 2.7 ± 0.4 × 10−9 /base/generation [21] and othermodel organisms, 3.5 × 10−9 /base/generation in Drosoph-ila melanogaster [23] and 7.1 ± 0.7 × 10−9 /base/generationin Arabidopsis thaliana [22]. Furthermore, analysis of the94 substitutions on mutation bias revealed very similarmutation spectrum to the spontaneous mutation spectrain N2 (C. elegans wild-type) [21] (Figure 2), and our ana-lysis of transition bias (transition/transversion – Ts/Tv –base substitution ratios) revealed the Ts/Tv ratio = 0.54,unc-46 mdf-1 such-4; dog-1 homozygotes for 470 generations.that propagated in the absence of MDF-1 and DOG-1 for longer thangk10). A second line was propagated for the total of 470 generations.lue). Note that unc-46 (used as a visible marker) is tightly linked to-1 such-4; dog-1 strains at generations F170, F270 and F470Duplications Deletions >20 bp3 573* 45– 1 943 139SsTarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 4 of 10which is within the range observed for spontaneous muta-tions in multiple mutation accumulation lines (average0.45; range 0.19 – 0.79) [21]. Therefore, the similarity ofthe substitution rate over the last 300 generations to thepreviously reported μbs in N2 strongly implies that themajority of the 431 substitutions identified in the F170generation are variants originally present in the startingstrain; therefore, we focused our analysis on GVs thathad occurred between F170 and F470 (Additional file 2:Table S1B).Previous studies showed that G4 DNA secondary struc-ture is mutagenic in the absence of DOG-1 [4-6,19]. Usingthe G4 DNA signature (G3+N1-7G3+N1-7G3+N1-7G3+), weidentified 2,372 such sites in C. elegans’ genome (Additionalfile 2: Table S2). Next, we tested to see if any of the 94 sub-stitutions occurred within the G4 DNA signature sequenceand observed that none of them did (Additional file 2:Table S1B), indicating that the identified substitutions mostlikely arose spontaneously and were not due to lack offunctional DOG-1.Figure 2 Mutation rate estimates. The variants analyzed are the 94 SBSmall InDels do not occur within G4-DNAWe applied VarScan2 [28] to identify 150 homozygousInDels that were 20 bp or less in mdf-1 such-4; dog-1F470(Additional file 2: Table S3A). We randomly selected 25 ofthe InDels and confirmed all 25 by Sanger re-sequencing,indicating a false positive rate of less than 5% (Additionalfile 2: Table S3A). Analysis of InDels in the unc-46 mdf-1such-4; dog-1 strains propagated for 170 and 270 genera-tions (Additional file 1: Figure S1B) revealed that 88% ofthe F470 InDels (133/150) existed in the F170 generation(Table 1), which again indicates that the majority of InDelswere already present in our starting strain. In fact, we onlyobserve eight additional InDels in F270 (Table 1) and an-other nine InDels accumulated between F270 and F470(Table 1). Based on the last 300 generations of propaga-tion, we estimate a mutation rate for InDels to be 17/(1 ×108bases × 300generations) = 5.7 × 10-10/base/generation. Theratio of InDels to SBSs observed was 0.18 or one InDelper 5.5 substitutions, which is much lower than the 1.31ratio reported previously in C. elegans [29]. However, thelower number of InDels to SBSs that we observe in theunc-46 mdf-1 such-4; dog-1 background is comparable toanalyses in yeast [30], A. thaliana [22] and human [31].Namely, WGS analysis in yeast had revealed ratio ofInDels to SBSs of ∼ 0.03 [30], which was consistent withprevious findings of one InDel per 33 SBSs [20]. Further-more, analysis on spontaneous occurrence of InDels in A.thaliana revealed the ratio of 0.13 of InDels to SBSs [22].It may be possible that the small number of InDels, overthe last 300 generations, in our strain may be a result of amutation that was acquired by propagation; however, itmay also be that the spontaneous mutation rate of InDelsin C. elegans is comparable to that of other organisms.Next, we tested to see if any of the 17 InDels occurredwithin the G4 DNA signature sequence and observedthat none of them did (Additional file 2: Table S3B).Thus, we believe that these InDels arose spontaneouslythat were identified between generations 170 and 470.and are not due to a lack of functional DOG-1.Duplications do not occur within G4-DNAAnalysis of gene copy-number variant accumulation afterlong-term propagation in C. elegans using the oaCGH hasprovided evidence for a high rate of spontaneous gene du-plications in this multi-cellular eukaryote [32]. Previously,using the oaCGH we showed that the such-4 suppressorgenome contains a large tandem duplication on Chromo-some V (LGV) [6]. Here, we used both Pindel [33] andoaCGH [34], and identified four sites with copy number in-creases in the mdf-1 such-4; dog-1F470 genome (Additionalfile 2: Table S4A), including a previously identified largetandem duplication [6]. One of the duplications involvesa two-copy addition, making the final count of five du-plication events and four different duplication sites.Analysis of the CNVs in the F170 and F270 genomes cap-tured a dynamic property of duplications (AdditionalTarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 5 of 10file 1: Figure S1C). In F170, we observed three duplica-tions (Additional file 2: Table S4A). One is the large tan-dem duplication located on LGV that amplifies 62protein-coding genes, which we have described previ-ously [6]. In generation F270, we detected duplications oftwo new loci, as well as further amplification of the LGVregion to three copies (triplication) (Additional file 1:Figure S1C and Additional file 2: Table S4A). In F470, wedid not find any new duplications (Additional file 2: TableS4A), but did observe that the duplication on LGI waslost, resulting in a wild-type copy number for this region(Additional file 2: Table S4A). Thus, the LGV duplicationexemplifies the property of duplications to further amplify;while LGI duplication shows that a duplicated region canrevert back to a normal copy number.The gene duplication rate for C. elegans was recently esti-mated to be 3.4 × 10-7/gene/generation [32]. Our data,based on the last 300 generations (from F170 and F470), alsorevealed a comparably high rate of gene duplication (6.5 ×10-7/gene/generation), specific to the unc-46 mdf-1 such-4;dog-1 background, when a large duplication on LGV is ex-cluded from the analysis. The LGV duplicated region con-tains 62 protein-coding genes, including cyb-3 [6]. Weobserved a correlation between increased dosage of CYB-3and a striking fitness increase in our strains; thus, our ex-perimental protocol selected for LGV amplification (de-tailed experimental analysis on these findings has beenrecently published [26]). Importantly, like SBSs and InDels,none of the duplications occurred in the vicinity of the G4DNA signature sequences (Additional file 2: Table S4B),also indicating that these CNVs arose spontaneously andwere not due to a lack of functional DOG-1.Deletions frequently initiate at G4 DNA sitesThe major type of mutation observed, in the absence ofDOG-1, is a deletion of 300 bp or smaller, initiating at eitherthe 5′-end of C- or the 3′-end of G-tracts [4,5,10,19]. Usingthe unique alignments generated by Novoalign and Pindel[33], we identified 183 homozygous deletions larger than20 bp in the mdf-1 such-4; dog-1F470 genome (Additional file2: Table S5A), including the known mdf-1(gk2) and dog-1(gk10) deletions. We randomly selected 28 of the deletions,and confirmed all of them using PCR (Additional file 2:Table S5A). To identify deletions, which may have beenmissed, we also used oaCGH [34]. We confirmed all of thedeletions, predicted by Pindel, that were covered by oaCGHprobes and identified an additional 13 deletions not detectedby Pindel (Additional file 2: Table S5A), making the finalcount 196 homozygous deletions in the mdf-1 such-4; dog-1F470 genome (Additional file 2: Table S5A).We observed 57 deletions in F170, indicating that 139 de-letions had accumulated in 300 generations (between F170and F470) propagated in the absence of MDF-1 and DOG-1(Additional file 1: Figure S1D and Table 1). We found thatthe majority of the deletions (123 of 139) initiated in G4DNA (Additional file 2: Table S5A). Previous analysis in C.elegans [32] estimated the spontaneous rate of deletions tobe 2.2 × 10-7/gene/generation respectively. The 139 dele-tions accumulating between the generations F170 and F470affect 19 protein-coding genes which allowed us to calcu-late the mutation rate of deletions over the 300 generationsto be 19/(20,400protein-coding genes × 300generations) = 3.1 × 10-6/gene/generation in the unc-46 mdf-1 such-4; dog-1 strain,which is approximately 10-fold higher than the estimatedrate in N2 [32]. This is similar to the estimated forwardmutation frequency of eT1-balanced lethal mutations indog-1(gk10) background [6]. To determine if the elevatedmutation rate of deletions is due to DOG-1 deficiency, wecompared mutation rates in non-G4 sites versus mutationrate in the G4 sites. While the mutation rate of deletions af-fecting the non-G4 sites, 4.9 × 10-7/gene/generation or 5.0× 10-10/base/generation, is comparable to the previously re-ported spontaneous rate of deletions [32]; the mutation ratebased on G4 sites, 1.7 × 10-4 /G4 site/generation, illustratesthe striking vulnerability of these DNA regions whenDOG-1 is absent. Therefore, deletions larger than 20 bp arethe only variants in the unc-46 mdf-1 such-4; dog-1 strainthat frequently occurred in the G4-DNA sites and had sig-nificantly higher mutation rate than the spontaneous ratereported previously for the strains with normal DOG-1function.monoG/C tracts larger than 14 bp are frequently deletedwhen DOG-1 is not functionalIn recent years, the G4 DNA has been implicated in di-verse biological processes, such as gene expression [35]and DNA replication initiation [35]. Consistent with theestablished role of DOG-1, we found that the majority(114) of the homozygous deletions that we identified be-tween F170 and F470 (139) initiated in the previously pro-posed G4 DNA signature G3+N1-7G3+N1-7G3+N1-7G3+ [5](Additional file 2: Table S5B) where the G-tract was al-most completely removed together with 5′ flankingDNA sequence (Figure 3A and B). In agreement with aprevious study [5], we found that the majority of dele-tions initiate at monoG tracts larger than 14 bp (93deletions) (Figure 3A), while 11 deletions initiate atmonoG-like structures with no more than two nucleotidesthat interrupt the homopolymer (Figure 3B), and 10 dele-tions initiate at sequences that interrupt the homopolymersby three or more nucleotides (e.g., GGGtGGGGaagttatGGGaGGG) (Additional file 2: Table S5B). MonoG/C tractslarger than 14 bp have the highest potential of forming theG-quartet structure. In fact, we find here that the unc-46mdf-1 such-4; dog-1 F470 genome has 13.2% of all thepredicted monoG/C tracts larger than 14 bp deleted. Aninteresting question to be addressed with future researchwould be to determine how mutation rate changes withTarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 6 of 10decreasing numbers of available targets. Furthermore, itwould be also important to see how many of the G4-DNAsites could be deleted in a strain and still maintain viabilityof the animals.To investigate whether there may be additional se-quences, which are vulnerable in the absence of DOG-1,we analyzed the 25 deletions that do not initiate in G4DNA signature sequences to see if there are commonpatterns. We found that eight of these deletions initiateat G-rich sequences that correspond to the G2+N1-7G2+N1-7G2+N1-7G2+ signature (four stretches of two ormore guanines, alternated with one to seven nucleotidesof any type), while one had a G2+N1-7G2+N1-7G2+ signa-ture at the breakpoint (Figure 3C and Table 2). Al-though, it may be possible that our strain had gained amutation in an unknown gene important for genomestability, it is also possible that additional DNA se-quences may be vulnerable to rearrangements in the ab-sence of DOG-1.Previously, it was found that the deletion sizes in viablelines detected by PCR [4,18] or oaCGH methods [5,6]were predominately smaller than 300 bp. In this study, the123 deletions that occurred at G-rich sequences rangedFigure 3 The majority of large deletions initiates at G-rich DNA. Figurx-axis represents genome location, while y-axis represents number of readswhile orange depicts 10 or less; the gaps depict no coverage, which is indibases underlined. (A) A 99 bp deletion that initiates at G20 homopolymer.112 bp deletion that initiates at a G-rich sequence, G2N3G2N5G2N2G2. (D) Adistribution of homozygous deletions that occur at G-rich sequences (n = 1between 49 and 10,228 base-pairs with the majority of thedeletions (86%) removing less than 300 bp (Figure 3E).These findings are in agreement with the deletion distri-bution sizes revealed by the study of a 69 G-tract deletionset [5]. However, larger deletions, initiating at G-tracts,es (A) through (D) are Genome Browser snapshots of single deletions;that cover the region; blue reads depict coverage of more than 10,cative of deletions. The reference sequence is depicted with deleted(B) A 96 bp deletion that initiates at G14TG6AGAAG3 sequence. (C) A55 bp deletion that initiates at a non G-rich sequence. (E) Size23).Table 2 Schematic representation of the nine deletions thatinitiate at G-rich regions within sequences that deviate fromthe G4 DNA signature, G3+N1-7G3+N1-7G3+N1-7G3+LG Locus SignatureI 464950..465071 G2NG12I 815471..815611 G3NG3NG3N2G2I 3747905..3748027 G2NG2NG2NG2NG2NG2NG2NG2NG2I 1373716..1373827 G2N3G2N5G2N2G2II 12687460..12687546 G2NG14NG2III 460436..460942 G11IV 3298588..3298711 G14V 17778167..17778449 G3NG3NG3N5G2NG2NG3X 2802061..2802223* G2NG3N6G3*Analysis of the sequences revealed that nine match the G2+N1-7G2+N1-7G2+N1-7G2+ signature, while one (marked with *) matches G2+N1-7G2+N1-7G2+signature at the breakpoint.deletion-formations. Previous studies were dependentTarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 7 of 10have been recovered as lethal mutations [6], consistentwith these regions containing essential genes [6]. We alsofound that 16 deletions, which occurred at non-G-richsites, removed small regions of less than 300 bp in size(Additional file 2: Table S5B).101 protein-coding genes are affected by the 253 GVsTo determine an effect of all the identified variants onprotein-coding genes, after long-term propagation in theabsence of MDF-1 and DOG-1, we used CooVar [36], atool developed by our group for annotating variants. Manyof the G-rich DNA sites are located in close proximity toprotein-coding genes in C. elegans [19,37]. Using CooVar[36], we predicted that 19 genes would be affected by 18 ofthe deletions that had accumulated between generationsF170 and F470 (Additional file 2: Table S6). The majority ofthe deletions (14) represent the first knockout alleles forthose genes and thus provide a genetic resource for study-ing their functions (Additional file 2: Table S6). We foundthat the majority of the InDels are not located in protein-coding regions, and the one that affects a protein-codinggene is an in-frame deletion (Additional file 2: Table S6).We found that the majority of the detected SBSs are lo-cated within non-protein coding regions. However, thereare 15 SBSs that fall within protein-coding regions that arepredicted to result in missense mutations (Additional file 2:Table S6). In total, 66 protein-coding genes were affectedby the duplications (Additional file 2: Table S6). The major-ity of the genes (62) are affected by the large tandem dupli-cation on LGV (Additional file 2: Table S6). In addition tothe 253 GVs that had accumulated between generationsF170 and F470, we had also performed analysis of the effecton protein-coding genes of all the variants observed in themdf-1such-4; dog-1F470 genome (Additional file 2: Table S7).Mutations are a source of genetic variation that confersto an organism either advantage due to a beneficial change,disadvantage due to a deleterious alteration, or neither dueto a neutral change. Considering the ~30-fold fitness recov-ery in this strain [26], the identified GVs are expected to beeither advantageous or neutral in this genetic background.In fact we showed, in a parallel study, that three mutationalevents discovered in this strain cooperate to increase fitnesswhen MDF-1 absent [26].ConclusionIn this study we undertook a genomics approach toidentify variants that accumulate in the C. elegans gen-ome after long-term propagation in the absence of twogenome-guardians, DOG-1/FANCJ and MDF-1/MAD1.Combining WGS analysis with oaCGH analysis allowedus to comprehensively analyze both the small-scale vari-ants (SBSs and InDels of 20 bp or smaller) and large-scale variants (CNVs larger than 20 bp). Freezing thestrain, for long term storage, at three different timeonly on PCR or oaCGH analysis; the two approaches arelimited in their ability to detect small-scale variants likeSBSs and InDels. Furthermore, genome analysis of thestrains at F170, F270 and F470 allowed us to capture andvisualize an intriguing property of CNVs (Additional file1: Figure S1). We did not find any reversions of dele-tions, substitutions, or InDels. As expected, once fixed,these types of mutations are propagated indefinitely(Additional file 1: Figure S1). However, we captured adynamic property of duplications: amplification of a re-gion on LGV from one-to-two-to-three copies and re-version of the LGI duplication back to a normal copynumber (Additional file 1: Figure S1C).This is the first extensive analysis of a strain that hadbeen propagated in the absence of DOG-1 helicase forhundreds of generations. We identified 954 monoG/Ctracts larger than 14 bp in the C. elegans genomeWS325; the polyG/C DNA sequence is the sequencewith the highest potential of forming G-quartet. Weshowed that 13% of these 954 sites are deleted in themdf-1 such-4; dog-1F470 genome. This finding raises animportant question on the changes in mutation ratewhen number of mutagenic targets is decreased. An-other important question is regarding to the role of G4DNA in normal development. Recently, G4 DNA hasbeen implicated in a variety of biological processes in-cluding telomere maintenance, gene expression, epigen-etic regulation, and DNA replication [38]. One questionto consider is how many G4 DNA sites could be re-moved from a genome yet still maintain viability of theanimals.MethodsC. elegans strainsThe following mutant alleles were used in this work: unc-46(e177), mdf-1(gk2), dog-1(gk10), and such-4(h2168). Thefollowing strains were used in this work: KR4233 [unc-46(e177) mdf-1(gk2) such-4(h2168)] and KR3627 [unc-46(e177) mdf-1(gk2) V/nT1[let-?(m435)])]; VC13 [dog-1(gk10)]. Additional strains used in this work were generatedin this study. Strains were maintained using standard proto-points, F170, F270 and F470, allowed us to compare andvisualize mutation accumulation over the generations.We were able to estimate the mutation rates in thisstrain for different types of mutations over the 300 gen-erations (F170 to F470). We observed a significantly ele-vated rate of deletions, larger than 20 bp, that initiatedat G/C-rich DNA sequence. Our approach had allowedus to show that, in the absence of DOG-1, DNA se-quences with G4-forming potential are vulnerable tocol on nematode growth media (NGM) plates seeded withOP50 bacteria [39]. The strains were maintained at 20°C.tutions were confirmed by re-sequencing as a result ofTarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 8 of 10Mutation accumulation procedure and phenotypicanalysisThe first suppressor, such-4, was isolated as previouslydescribed [24]. Briefly, VC13 was backcrossed to N2 tentimes to remove any mutations present in VC13. Thenthe outcrossed dog-1(gk10) males were used to constructunc-46(e177) mdf-1(gk2) +/+ + nT1[let-X]; dog-1(gk10)/dog-1(gk10). Note that unc-46(e177) is linked mdf-1(gk2)and used as a visible marker to track mdf-1(gk2). F1 unc-46 mdf-1 homozygotes (n = 40) were picked and platedindividually and a single worm, from a plate containingfertile worms in the third generation, was isolated as asuppressor candidate (such-4) [24]. We outcrossed oneworm from this strain to establish KR4233 mdf-1(gk2)such-4(h2168) [6,24], while we maintained a second line at20°C for 470 generations (strain JNC170). Each generation5 L4 hermaphrodites were transferred to a fresh plate. Wealso froze the worms at generations 170 (strain JNC168)and 270 (strain JNC169). The phenotypic analysis was per-formed as previously described [24].Whole genome sequencing and computational analysisof the unc-46 mdf-1 such-4; dog-1F470 genomeGenomic DNA was prepared from JNC170 followinga standard protocol (http://genetics.wustl.edu/tslab/protocols/genomic-stuff/worm-genomic-dna-prep/) ori-ginally set up by the Andy Fire Lab. The library was pre-pared with average insert size of 300 bp and the genomewas then sequenced using Illumina Solexa sequencing (atCanada’s Michael Smith Genome Sciences Centre) and92,282,948 reads of 101 bp in length were obtained. Thereads were then aligned to the C. elegans reference gen-ome WS235 (hosted at WormBase) [27] in paired endmanner (46,141,474 pairs) using the Novoalign alignmenttool. 73,482,133 (79.63%) of the total reads were of basequality 30 or more and were uniquely mapped, generating70-fold coverage of the genome.SBSs were detected using the uniquely mapped reads andpileup2snp function of the variant caller VarScan2; version2.3.2 [28]. We used SAMtools [40] to generate the mpileupfile necessary as input for VarScan. We filtered out the SBSsthat did not meet the following criteria: depth of coverage> 5 and ≤ 200, variant frequency ≥ 0.9 and base quality ≥ 30.After these filtering steps, we identified 776 substitutions inthe mdf-1 such-4; dog-1F470 genome. However, comparisonwith the sequenced N2 strains from CGC and Horvitz lab(sequencing reads were kind gift from Dr. Bob Waterston)revealed 525 homozygous substitutions that are unique tomdf-1 such-4; dog-1F470 genome. The 251 SBSs were notincluded in our analysis as they did not accumulate duringthe course of our experiment. To determine rate of falsepositives in our set, we randomly selected 30 SBSs(Additional file 2: Table S1A), designed primers flankingthe predicted substitution sites, amplified the fragmentsbeing adjacent to randomly selected SBSs or as being non-randomly selected as candidates in later analyses.InDels were detected using the uniquely mappedreads and pileup2indel function of the variant callerVarScan2; version 2.3.2 [28]. We used SAMtools [40] togenerate the mpileup file necessary as input for VarS-can2. We filtered out the InDels that did not meet thefollowing criteria: depth of coverage > 5 and ≤ 200, vari-ant frequency ≥ 0.9 and base quality ≥ 30. Additionally,we re-evaluated and filtered the output so that each co-ordinate corresponds to one variant. After these filter-ing steps, we identified 556 InDels in the mdf-1 such-4;dog-1F470 genome. r, comparison with the sequenced N2strains from CGC and Horvitz lab (sequencing readswere kind gift from Dr. Bob Waterston) revealed 150homozygous InDels that are unique to mdf-1 such-4;dog-1F470 genome. The 406 InDels were not included inour analysis as they did not accumulate during thecourse of our experiment. To determine rate of falsepositives in our set, we have randomly selected 25InDels (Additional file 2: Table S3A), designed primersflanking the predicted InDel sites, amplified the frag-ments and re-sequenced using Sanger sequencingmethod at Genewiz, Inc. All of the 25 InDels were con-firmed by Sanger re-sequencing, suggesting a false posi-tive rate of less than 5% (Additional file 2: Table S3A).These unique-mapping reads were then used, togetherwith Pindel [33], to identify deletions. The final set offixed/homozygous deletions was selected based on thefollowing criteria: a = the number of unique reads sup-porting the breakpoints of the deletion; b = the numberof reads within the deleted region; select the predicteddeletion if aaþb is larger than 0.5 and the size of the dele-tion is larger than 20 bp. From the candidate deletions,28 were randomly selected and the regions were PCR-amplified using the same genomic DNA that was usedfor the WGS and primers designed in the flanking re-gions of the computationally identified deletions. All ofthe 28 randomly selected predicted deletion sizes wereconfirmed using DNA electrophoresis gels, suggestingthe rate of false positives of less than 5%.Whole genome sequencing and computational analysisof the unc-46 mdf-1 such-4; dog-1F170 and mdf-1 such-4;dog-1F270 strainsThe genomic DNA was prepared from JNC168 and JNC169and re-sequenced using Sanger sequencing method at Gen-ewiz, Inc. All 30 substitutions were confirmed by Sangerre-sequencing, suggesting a false positive rate of less than5% (Additional file 2: Table S1A). Also, 12 additional substi-following a standard protocol (http://genetics.wustl.edu/tslab/protocols/genomic-stuff/worm-genomic-dna-prep/).to disrupt protein-coding genes in the mdf-1; dog-1; such-4 F470 genome.Hybridization; G4: G-quartet; SBS: Single base substitution; GV: GenomeThe authors declare that they have no competing interests.AMR and NC led the study. MTG wrote the manuscript. All authors read,edited and approved the final manuscript.Intellectual Disability Endeavour in British Columbia, Vancouver, BC, Canada.Tarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 9 of 10The genomic DNA was sheared to generate 500 bp frag-ment and the library was prepared using the NEBNext®Ultra™ DNA Library Prep Kit for Illumina®. The librarywas then sequenced using Illumina Solexa sequencing (atSimon Fraser University) and 9,950,748 (F170) and6,225,488 (F270) reads of 250 bp length were obtained. Thereads were then aligned to the C. elegans reference gen-ome WS235 (hosted at WormBase) [27] in paired endmanner using the Novoalign alignment tool as describedabove to achieve 20-fold (F170) and 14-fold (F270) coverage.For both strains, we used Pindel [33] to identify deletionsand duplications, and VarScan2; version 2.3.2 [28] to iden-tify small InDels and substitutions, as described above.We also used sequenced N2 strains from CGC and Hor-vitz lab to remove variants existing in the N2 strain. Then,we compared the variants present in the F170 and F270 ge-nomes with the ones identified in the F470 genome.oaCGH analysis of the unc-46 mdf-1 such-4; dog-1 strainsTo perform oaCGH analysis, we used the same genomicDNA from the mdf-1 such-4; dog-1 lines (F170, F270 andF470) that were used for the WGS and the reference N2DNA that was prepared following a standard protocol.oaCGH analysis was performed as described by Maydanand colleagues [41] with a newly designed microarray.The 3-plex microarray with design name 120618_Cel-e_WS230_JK_CGH was manufactured by Roche Nim-bleGen Inc. with each individual sub-array comprising720 k 50-mer oligonucleotide probes. The filters used toselect the probes primarily followed Maydan and col-leagues [41] without focusing on coding regions inorder to provide a more uniform coverage of the gen-ome (WormBase release WS230). In regions whereunique probes could not be designed selection filterswere slightly relaxed in order to allow the inclusion ofprobes with possible cross-hybridization to at most oneother location in the genome.Availability of supporting dataWhole Genome Sequencing (WGS) data, fastq files, for thethree strains: JNC168, JNC169 and JNC170 are available inthe NCBI Sequence Read Archive (SRA) (http://www.ncbi.nlm.nih.gov/sra) under the BioProject accession numberSRP053517 (PRJNA275156) with the fastq files: (170_R1.fastq, 170_R2.fastq, 270_R1.fastq, 270_R2.fastq, 470_R1.fastq and 470_R2.fastq) under the SRR1797354 accessionnumber.Additional filesAdditional file 1: Figure S1. Mutation accumulation in unc-46 mdf-1such-4; dog-1 strains. Accumulation of mutations was visualized usingCircos [42]. Note that due to the limited resolution, data points thatoccur close in the genome may appear as a single line or a single link.AcknowledgementsWe thank the C. elegans Gene Knockout Consortium for generating thedeletion mutants, the Caenorhabditis Genetics Center (CGC) for providingthe strains, and the Michael Smith Genome Sciences Centre for Illuminasequencing. We also thank Bob Waterston for the sequencing reads of N2strains; Robert Johnsen, Steven Jones, Harald Hutter and Nancy Hawkins forcomments. This work was supported by the Canadian Institutes for HealthResearch (CIHR) and Fanconi Anemia Fellowship to MTG, and the DiscoveryGrant from the Natural Science and Engineering Research Council (NSERC) toNC. NC is a Michael Smith Foundation for Health Research (MSFHR) Scholarand a CIHR New Investigator. Work in the laboratory of DGM is supported bya grant from CIHR. DGM is a Senior Fellow of the Canadian Institute forAdvanced Research (CIFAR). Work in the laboratory of AMR is supported bygrants from CIHR and NSERC.Author details1Department of Molecular Biology and Biochemistry, Simon Fraser University,V5A 1S6 Burnaby, BC, Canada. 2Department of Zoology, University of BritishColumbia, V6T 1Z4 Vancouver, BC, Canada. 3Department of Medical Genetics,University of British Columbia, V6T 1Z3 Vancouver, BC, Canada. 4Currentaffiliation: Centre for Molecular Medicine and Therapeutics; Child and FamilyResearch Institute, Vancouver, BC, Canada. 5Current affiliation: TreatableAuthors’ contributionsMTG made the strains; MTG and TW wrote Perl scripts; MTG and ZQprepared DNA for WGS and oaCGH. SF and JT generated oaCGH data. MTGanalyzed the WGS data. MTG and SF analyzed the oaCGH data. MTG, DGM,variant.Competing interestsAbbreviationsSAC: Spindle assembly checkpoint; APC/C: Anaphase promoting complex orcyclosome; CIN: Chromosome instability; WGS: Whole genome sequencing;FA: Fanconi anemia; oaCGH: oligonucleotide array Comparative GenomicF170 chromosomes and variants originating at F170 are depicted in green;F270 chromosomes and variants originating at F270 are depicted in purple,while F470 chromosomes and variants unique to F470 are depicted inblue. The outer circle is a plot of all the variants present at a specificgeneration, while inner links are depicting propagation of the variantsfrom one generation time-point to the next. The following variants areshown: (A) SBSs, (B) InDels, (C) Duplications, (D) Deletions.Additional file 2: Table S1A. Homozygous SBSs identified in the F170;F270 and F470 genomes. Table S1B. Homozygous SBSs identified afterpropagating mdf-1; dog-1; such-4 for 300 generations. Table S2. Locationof the G3-5 N1-7G3-5 N1-7G3-5 N1-7G3-5 sequences in the C. elegans genome.Table S3A. Homozygous InDels smaller than or equal to 20 bp identifiedin the F170; F270 and F470 genomes. Table S3B. Homozygous InDelssmaller than or equal to 20 bp after propagating mdf-1; dog-1; such-4 for300 generations. Table S4A. Duplications identified in the F170; F270 andF470 genomes. Table S4B. Duplications identified after propagatingmdf-1; dog-1; such-4 for 300 generations. Table S5A. Homozygousdeletions larger than 20 bp identified in the F170; F270 and F470 genomes.Table S5B. Homozygous deletions larger than 20 bp identified afterpropagating mdf-1; dog-1; such-4 for 300 generations. Table S6.Accumulation of mutations predicted to disrupt protein-coding genesafter 300 generations of propagation. Table S7. All mutations predictedReceived: 3 September 2014 Accepted: 24 February 2015and spectrum in yeast. Proc Natl Acad Sci U S A. 2014;111:E2310–8.31. Chen J-Q, Wu Y, Yang H, Bergelson J, Kreitman M, Tian D. Variation in theratio of nucleotide substitution and indel rates across genomes in mammalsand bacteria. Mol Biol Evol. 2009;26:1523–31.32. Lipinski KJ, Farslow JC, Fitzpatrick KA, Lynch M, Katju V, Bergthorsson U.High spontaneous rate of gene duplication in Caenorhabditis elegans. CurrBiol. 2011;21:306–10.33. Ye K, Schulz MH, Long Q, Apweiler R, Ning Z. Pindel: a pattern growthapproach to detect break points of large deletions and medium sizedinsertions from paired-end short reads. Bioinformatics. 2009;25:2865–71.Tarailo-Graovac et al. BMC Genomics  (2015) 16:210 Page 10 of 10References1. Burrell RA, McGranahan N, Bartek J, Swanton C. The causes andconsequences of genetic heterogeneity in cancer evolution. Nature.2013;501:338–45.2. Kitagawa R, Rose AM. Components of the spindle-assembly checkpoint areessential in Caenorhabditis elegans. Nat Cell Biol. 1999;1:514–21.3. Musacchio A, Salmon ED. The spindle-assembly checkpoint in space andtime. Nat Rev Mol Cell Biol. 2007;8:379–93.4. Cheung I, Schertzer M, Rose A, Lansdorp PM. Disruption of dog-1 inCaenorhabditis elegans triggers deletions upstream of guanine-rich DNA.Nat Genet. 2002;31:405–9.5. Kruisselbrink E, Guryev V, Brouwer K, Pontier DB, Cuppen E, Tijsterman M.Mutagenic capacity of endogenous G4 DNA underlies genome instability inFANCJ-defective C. elegans. Curr Biol. 2008;18:900–5.6. Zhao Y, Tarailo-Graovac M, O’Neil NJ, Rose AM. Spectrum of mutationalevents in the absence of DOG-1/FANCJ in Caenorhabditis elegans. DNARepair. 2008;7:1846–54.7. Sen D, Gilbert W. Formation of parallel four-stranded complexes byguanine-rich motifs in DNA and its implications for meiosis. Nature.1988;334:364–6.8. Sen D, Gilbert W. A sodium-potassium switch in the formation of four-stranded G4-DNA. Nature. 1990;344:410–4.9. Gellert M, Lipsett MN, Davies DR. Helix formation by guanylic acid. Proc NatlAcad Sci U S A. 1962;48:2013–8.10. Youds JL, Barber LJ, Ward JD, Collis SJ, O’Neil NJ, Boulton SJ, et al. DOG-1 isthe Caenorhabditis elegans BRIP1/FANCJ homologue and functions ininterstrand cross-link repair. Mol Cell Biol. 2008;28:1470–9.11. Huppert JL, Balasubramanian S. Prevalence of quadruplexes in the humangenome. Nucleic Acids Res. 2005;33:2908–16.12. Todd AK, Johnston M, Neidle S. Highly prevalent putative quadruplexsequence motifs in human DNA. Nucleic Acids Res. 2005;33:2901–7.13. Levitus M, Waisfisz Q, Godthelp BC, de Vries Y, Hussain S, Wiegant WW, et al.The DNA helicase BRIP1 is defective in Fanconi anemia complementationgroup. J Nat Genet. 2005;37:934–5.14. Levran O, Attwooll C, Henry RT, Milton KL, Neveling K, Rio P, et al. TheBRCA1-interacting helicase BRIP1 is deficient in Fanconi anemia. Nat Genet.2005;37:931–3.15. Litman R, Peng M, Jin Z, Zhang F, Zhang J, Powell S, et al. BACH1 is criticalfor homologous recombination and appears to be the Fanconi anemiagene product FANCJ. Cancer Cell. 2005;8:255–65.16. Seal S, Thompson D, Renwick A, Elliott A, Kelly P, Barfoot R, et al. Truncatingmutations in the Fanconi anemia J gene BRIP1 are low-penetrance breastcancer susceptibility alleles. Nat Genet. 2006;38:1239–41.17. Cantor SB, Bell DW, Ganesan S, Kass EM, Drapkin R, Grossman S, et al.BACH1, a novel helicase-like protein, interacts directly with BRCA1 andcontributes to its DNA repair function. Cell. 2001;105:149–60.18. Youds JL, O’Neil NJ, Rose AM. Homologous recombination is required forgenome stability in the absence of DOG-1 in Caenorhabditis elegans.Genetics. 2006;173:697–708.19. Pontier DB, Kruisselbrink E, Guryev V, Tijsterman M. Isolation of deletionalleles by G4 DNA-induced mutagenesis. Nat Methods. 2009;6:655–7.20. Lynch M, Sung W, Morris K, Coffey N, Landry CR, Dopman EB, et al. Agenome-wide view of the spectrum of spontaneous mutations in yeast.Proc Natl Acad Sci U S A. 2008;105:9272–7.21. Denver DR, Dolan PC, Wilhelm LJ, Sung W, Lucas-Lledó JI, Howe DK, et al. Agenome-wide view of Caenorhabditis elegans base-substitution mutationprocesses. Proc Natl Acad Sci U S A. 2009;106:16310–4.22. Ossowski S, Schneeberger K, Lucas-Lledó JI, Warthmann N, Clark RM, ShawRG, et al. The rate and molecular spectrum of spontaneous mutations inArabidopsis thaliana. Science. 2010;327:92–4.23. Keightley PD, Trivedi U, Thomson M, Oliver F, Kumar S, Blaxter ML. Analysisof the genome sequences of three Drosophila melanogaster spontaneousmutation accumulation lines. Genome Res. 2009;19:1195–201.24. Tarailo M, Kitagawa R, Rose AM. Suppressors of spindle checkpoint defect(such) mutants identify new mdf-1/MAD1 interactors in Caenorhabditiselegans. Genetics. 2007;175:1665–79.25. Tarailo-Graovac M, Wang J, Tu D, Baillie DL, Rose AM, Chen N. Duplicationof cyb-3 (cyclin B3) suppresses sterility in the absence of mdf-1/MAD1spindle assembly checkpoint component in Caenorhabditis elegans. CellCycle Georget Tex. 2010;9:4858–65.34. Kallioniemi A, Kallioniemi OP, Sudar D, Rutovitz D, Gray JW, Waldman F,et al. Comparative genomic hybridization for molecular cytogenetic analysisof solid tumors. Science. 1992;258:818–21.35. Verma A, Yadav VK, Basundra R, Kumar A, Chowdhury S. Evidence ofgenome-wide G4 DNA-mediated gene expression in human cancer cells.Nucleic Acids Res. 2009;37:4194–204.36. Vergara IA, Frech C, Chen N. CooVar: co-occurring variant analyzer. BMC ResNotes. 2012;5:615.37. Zhao Y, O’Neil NJ, Rose AM. Poly-G/poly-C tracts in the genomes ofCaenorhabditis. BMC Genomics. 2007;8:403.38. Tarsounas M, Tijsterman M. Genomes and G-quadruplexes: for better or forworse. J Mol Biol. 2013;425:4782–9.39. Brenner S. The genetics of Caenorhabditis elegans. Genetics. 1974;77:71–94.40. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. TheSequence Alignment/Map format and SAMtools. Bioinformatics.2009;25:2078–9.41. Maydan JS, Flibotte S, Edgley ML, Lau J, Selzer RR, Richmond TA, et al.Efficient high-resolution deletion discovery in Caenorhabditis elegans byarray comparative genomic hybridization. Genome Res. 2007;17:337–47.42. Krzywinski M, Schein J, Birol I, Connors J, Gascoyne R, Horsman D, et al.Circos: an information aesthetic for comparative genomics. Genome Res.2009;19:1639–45.Submit your next manuscript to BioMed Centraland take full advantage of: • Convenient online submission• Thorough peer review• No space constraints or color figure charges• Immediate publication on acceptance• Inclusion in PubMed, CAS, Scopus and Google Scholar• Research which is freely available for redistribution26. Tarailo-Graovac M, Wong T, Qin Z, Flibotte S, Tylor J, Moerman DG, et al.Cyclin B3 and dynein heavy chain cooperate to increase fitness in theabsence of mdf-1/MAD1 in Caenorhabditis elegans. Cell Cycle. 2014;13:1–11.27. Yook K, Harris TW, Bieri T, Cabunoc A, Chan J, Chen WJ, et al. WormBase2012: more genomes, more data, new website. Nucleic Acids Res. 2012;40(Database issue):D735–41.28. Koboldt DC, Zhang Q, Larson DE, Shen D, McLellan MD, Lin L, et al. VarScan2: somatic mutation and copy number alteration discovery in cancer byexome sequencing. Genome Res. 2012;22:568–76.29. Denver DR, Morris K, Lynch M, Thomas WK. High mutation rate andpredominance of insertions in the Caenorhabditis elegans nuclear genome.Nature. 2004;430:679–82.30. Zhu YO, Siegal ML, Hall DW, Petrov DA. Precise estimates of mutation rateSubmit your manuscript at www.biomedcentral.com/submit


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items