UBC Faculty Research and Publications

Comparison of genome-wide array genomic hybridization platforms for the detection of copy number variants… Tucker, Tracy; Montpetit, Alexandre; Chai, David; Chan, Susanna; Chénier, Sébastien; Coe, Bradley P; Delaney, Allen; Eydoux, Patrice; Lam, Wan L; Langlois, Sylvie; Lemyre, Emmanuelle; Marra, Marco; Qian, Hong; Rouleau, Guy A; Vincent, David; Michaud, Jacques L; Friedman, Jan M Mar 25, 2011

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

Item Metadata


52383-12920_2010_Article_220.pdf [ 317.69kB ]
JSON: 52383-1.0223648.json
JSON-LD: 52383-1.0223648-ld.json
RDF/XML (Pretty): 52383-1.0223648-rdf.xml
RDF/JSON: 52383-1.0223648-rdf.json
Turtle: 52383-1.0223648-turtle.txt
N-Triples: 52383-1.0223648-rdf-ntriples.txt
Original Record: 52383-1.0223648-source.json
Full Text

Full Text

RESEARCH ARTICLE Open AccessComparison of genome-wide array genomichybridization platforms for the detection of copynumber variants in idiopathic mental retardationTracy Tucker1*, Alexandre Montpetit2, David Chai3, Susanna Chan4, Sébastien Chénier5, Bradley P Coe6,Allen Delaney4, Patrice Eydoux3, Wan L Lam6, Sylvie Langlois3,7, Emmanuelle Lemyre5, Marco Marra4, Hong Qian4,Guy A Rouleau8,9, David Vincent2, Jacques L Michaud5,8 and Jan M Friedman1,7AbstractBackground: Clinical laboratories are adopting array genomic hybridization as a standard clinical test. A number ofwhole genome array genomic hybridization platforms are available, but little is known about their comparativeperformance in a clinical context.Methods: We studied 30 children with idiopathic MR and both unaffected parents of each child using Affymetrix500 K GeneChip SNP arrays, Agilent Human Genome 244 K oligonucleotide arrays and NimbleGen 385 K Whole-Genome oligonucleotide arrays. We also determined whether CNVs called on these platforms were detected byIllumina Hap550 beadchips or SMRT 32 K BAC whole genome tiling arrays and tested 15 of the 30 trios onAffymetrix 6.0 SNP arrays.Results: The Affymetrix 500 K, Agilent and NimbleGen platforms identified 3061 autosomal and 117 Xchromosomal CNVs in the 30 trios. 147 of these CNVs appeared to be de novo, but only 34 (22%) were found onmore than one platform. Performing genotype-phenotype correlations, we identified 7 most likely pathogenic and2 possibly pathogenic CNVs for MR. All 9 of these putatively pathogenic CNVs were detected by the Affymetrix500 K, Agilent, NimbleGen and the Illumina arrays, and 5 were found by the SMRT BAC array. Both putativelypathogenic CNVs identified in the 15 trios tested with the Affymetrix 6.0 were identified by this platform.Conclusions: Our findings demonstrate that different results are obtained with different platforms and illustratethe trade-off that exists between sensitivity and specificity. The large number of apparently false positive CNV callson each of the platforms supports the need for validating clinically important findings with a different technology.BackgroundChromosomal abnormalities, the most frequently diag-nosed cause of mental retardation (MR) [1], are routi-nely identified by cytogenetic analysis. Studies usingarray genomic hybridization (AGH) have found appar-ently-pathogenic gains or losses of genetic material in atleast 10% of children with MR and normal conventionalcytogenetic analysis [2-4]. These apparently pathogenicdeletions and duplications range in size from < 100 Kbto 15 Mb. Such submicroscopic chromosomal gains orlosses are collectively called pathogenic copy numbervariants (CNVs).However, most CNVs, despite producing genomicimbalance of many thousands of DNA base pairs, donot cause MR. In fact, CNVs are the greatest source ofgenetic variation in normal people; the mean number ofapparently benign CNVs observed ranges from 10’s-1000’s per person, depending on the technology used[5-9]. Distinguishing benign CNVs from those thatcause MR and other birth defects is the most seriouschallenge to the routine clinical use of AGH, especiallyfor prenatal diagnosis [4,10-13].A consensus has developed that AGH should beoffered routinely in the evaluation of children with MR* Correspondence: tbtucker@interchange.ubc.ca1Department of Medical Genetics, University of British Columbia, Vancouver,British Columbia, CanadaFull list of author information is available at the end of the articleTucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25© 2011 Tucker et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative CommonsAttribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction inany medium, provided the original work is properly cited.and other birth defects [2,3,11]. However, there is noagreement regarding the choice of AGH platform, reso-lution, or reference sample that is most appropriate forclinical use [3,14-17]. AGH for clinical diagnosis hasoften employed targeted arrays with probes in genomicregions known to be associated with microdeletion andmicroduplication syndromes, and more recent versionsof many targeted arrays have added additional probes(i.e., a ‘backbone’) to provide some degree of genome-wide coverage. As the density of probes in these back-bones increases, genome-wide and targeted platformsare converging, with both providing a survey of thewhole genome at relatively high resolution.Few studies have compared AGH whole genome tech-nologies [18-20], and most are retrospective and limited topathogenic CNVs. We performed AGH studies on 30 MRtrios (children with idiopathic MR and both of their unaf-fected parents) using three different high-resolution gen-ome-wide oligonucleotide platforms – Affymetrix 500 K,Agilent 244 K and NimbleGen 385 K – to assess their uti-lity for the identification of pathogenic CNVs in childrenwith MR. We determined whether the CNVs called onthese platforms were also detected by the Illumina Hap550Beadchip or the SMRT 32 K BAC whole genome tilingarray and tested samples from 15 of the MR trios on Affy-metrix 6.0 arrays. This large comparison of multiple AGHplatforms provides unique insights into both the powerand the limitations of current technology for detectingpathogenic genomic imbalance in children with MR.MethodsPatientsPatients with MR and at least one of the following addi-tional characteristics were selected for study: 1) growthretardation of pre- and/or post-natal onset; 2) microce-phaly or macrocephaly; 3) one or more major malforma-tions, and 4) more than two facial dysmorphic features.Characterization of this cohort with the checklist devel-oped by de Vries et al. [21] showed an average score of4.4 (range: 2-9). The cause of the MR in each child wasunknown despite full evaluation by a clinical geneticist,a karyotype at ≥500 band resolution and subtelomericFISH studies. This study was approved by the Universityof British Columbia Clinical Research Ethics Board andHopital Sainte-Justine Research Ethics Board, andinformed consent was obtained from each family.Array Genomic HybridizationGenomic DNA was extracted from blood samples usingthe Puregene DNA kit (Gentra System). DNA qualityfor each trio was assessed by electrophoresis in a 1%agarose gel, and DNA concentration was measured witha NanoDrop™ Spectrophotometer.AGH was performed on 30 children with idiopathicMR and on both normal parents of each child on Affy-metrix 500 K GeneChips, Agilent 244 K Oligonucleo-tide Arrays, and NimbleGen 385 K OligonucleotideArrays. In addition, DNA from the 30 children wasrun on Illumina Hap550 Beadchips using a set ofabout 100 HapMap samples as reference. DNAs fromthe 30 children were also run on Sub-Megabase Reso-lution Tiling-set (SMRT) human genomic BAC arraysusing one of the parents - the one of the same sex - asreference. In addition, DNA samples from 15 trioswere run on the Affymetrix Genome-Wide HumanSNP Array 6.0.All samples were handled according to the platformmanufacturer’s recommendations, and CNV detectionwas performed using the manufacturer’s recommendedsoftware with default settings (See Additional File fordetailed protocol and software settings).Identification of Autosomal de novo ChangesA CNV was considered to be de novo on Affymetrix500 K, Agilent 244 K, NimbleGen 385 K, or Affymetrix6.0 AGH if a set of probes identified by the platformalgorithm was called as a deletion in the child relativeto both parents or as a duplication in the child relativeto both parents on the same platform.Identification of X Chromosomal CNVsIn addition to looking for de novo CNVs as describedfor the autosomes above, we performed manual assess-ments for CNVs when there was a sex-mismatchbetween the child and parent because the NimbleGenand Affymetrix platforms are unable to account for sexmismatches between the test and reference DNAs.Identification of Pathogenic ChangesWe used de novo occurrence as a major criterion ofpathogenicity for autosomal CNVs; however, mostde novo changes found in these studies are unlikely tobe pathogenic for MR in the children studied. In addi-tion, all X chromosomal CNVs except those in the pseu-doautosomal or XY homology regions were assessed forpathogenicity. We used previously published criteria[4,12,13,22,23] to determine which CNVs are likely tobe pathogenic. We performed genotype-phenotype cor-relations only on CNVs that were greater than 50 Kb inlength and that met the criteria of pathogenicity citedabove.CNV ConfirmationCNVs were validated by FISH, MLPA, qPCR or PCR(for X chromosome deletions identified in males). SeeAdditional File 1 for detailed validation protocols.Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 2 of 10Statistical AnalysisA chi square analysis was performed to assess differencesbetween platforms in the proportion of singleton CNVcalls and the number of de novo CNVs. Differences insizes and types of CNVs called between platforms wereassessed by Mann-Whitney U tests. A p-value of 0.05was considered significant in all analyses.ResultsWe compared the ability of the Affymetrix 500 K, Agi-lent 244 K and NimbleGen 385 K platforms to detect denovo CNVs in 30 patients with idiopathic MR using thenormal parents of each child as reference.Overall, 1,492 autosomal deletions and 1,569 autoso-mal duplications were called in the 30 children with MRby one or more of the three main platforms (Affymetrix500 K, Agilent 244 K, and NimbleGen 385 K; AdditionalFile 2 and Additional File 1). Over 80% of the autosomalCNV calls were made by only one of the three majorplatforms. The proportion of singleton CNV calls madewas significantly different among the 3 platforms (p = <0.001). The NimbleGen platform identified about 40%more autosomal calls than the Agilent platform, whilethe higher density Affymetrix 500 K platform identifiedfewer than one third as many CNVs as the Agilent plat-form (Figure 1). However, 60% of the autosomal CNVsidentified only on the Agilent or NimbleGen platformare in genomic regions that had fewer than 5 probes onthe Affymetrix 500 K array, so recognition of suchCNVs would not be expected.Detection of autosomal de novo CNVsThere were 146 autosomal de novo CNVs identified onone or more of the three main platforms in the 30 MRprobands, an average of 4.9 de novo CNVs per patient(Additional File 3). Two patients had only one de novoCNV identified, 3 had 2 de novo CNVs, 5 had 3 de novoCNVs and 20 patients had 4 or more de novo CNVs.114 (78%) of the autosomal de novo calls were iden-tified on only one platform, 23 on 2 platforms and 9on all 3 platforms (Figure 2). Significantly fewer denovo calls were made with the Affymetrix 500 K sys-tem than with Agilent (p = < 0.001) or NimbleGenplatforms (p = < 0.001).A larger number of autosomal de novo deletions (96)than duplications (50) were called. The median size ofthe de novo deletions (113 Kb) was similar to that of thede novo duplications (98 Kb) (p = 0.348). The mediansize of the de novo CNVs detected by the NimbleGenplatform (183 Kb) was significantly larger than thoseidentified by the Agilent platform (82 Kb; p = 0.003) butnot significantly larger than those identified by the Affy-metrix 500 K platform (145 Kb; p = 0.083; AdditionalFile 1).Many autosomal de novo CNVs were observed in mul-tiple individuals in this small series. These recurrentcalls accounted for 45 (31%) of the 146 autosomalde novo CNVs called in the 30 trios, and all occurred inregions that contain polymorphic CNVs previouslyrecognized by oligonucleotide arrays or higher resolu-tion techniques (Database of Genomic Variants (DGV),http://projects.tcag.ca/variation/) (Additional File 3).Detection of autosomal de novo CNVs with the Affymetrix6.0 PlatformTo explore whether the higher density Affymetrix 6.0array improved CNV detection in comparison to thethree main platforms, we arbitrarily selected 15 of the30 MR trios for analysis using the Affymetrix 6.0 plat-form (see Additional File 1 for details of analysis).AGILENT 244KNIMBLEGEN 385KAFFYMETRIX 500K1291204840839 38140329191(47%)783(61%)1541(75%)Figure 1 Venn diagrams of CNV calls made by the 3 main AGHplatforms. The numbers under each platform name indicate thetotal number of CNV calls by that platform. The numbers in theintersecting regions indicate CNV calls made by multiple platforms.The numbers outside the intersecting regions are the number ofCNVs that were unique to that platform.AGILENT 244K NIMBLEGEN 385KAFFYMETRIX 500K74 80333 092021(64%)42(57%)51(64%)Figure 2 Venn diagrams of autosomal de novo CNV calls madeby the 3 main AGH platforms. The numbers under each platformname indicate the total number of de novo CNV calls by thatplatform. The numbers in the intersecting regions present CNV callsmade by multiple platforms. The numbers outside the intersectingregions are the number of CNVs that were unique to that platform.Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 3 of 10There was a total of 915 autosomal CNVs identifiedwith the Affymetrix 6.0 platform in the 15 probands (anaverage of 61 CNVs per person) (Additional File 4),compared to 281 autosomal CNVs (an average of 18.7per person) in these same 15 individuals on the Affyme-trix 500 K platform. 682 CNVs were called in these 15probands on the Agilent platform and 973 CNVs werecalled on the NimbleGen platform.41 of the 915 autosomal CNVs identified by Affyme-trix 6.0 AGH occurred de novo, and 29 of these 41(71%) CNVs were not found by any of the three mainplatforms. The Affymetrix 6.0 platform identified 9 of12 CNVs that were called as de novo on 2 or 3 of themain platforms in these patients (Additional File 3).Detection of autosomal de novo CNVs with the Illuminaand SMRT PlatformsIn more limited comparisons, we sought to determinewhether autosomal de novo CNVs identified by thethree main platforms studied are likely to be identifiedby the Illumina Hap550 Beadarray or the SMRT 32 KBAC tiling array.The Illumina platform identified 14 of the 146 autoso-mal de novo CNVs that had been called on one or moreof the 3 main platforms and 8 of the 9 de novo CNVsthat had been called on all 3 main platforms (AdditionalFile 3). All Illumina CNV calls are listed in AdditionalFile 5.The SMRT platform identified 9 of the 146 autosomalde novo CNVs called on one or more of the 3 mainplatforms, including 4 of the 9 de novo CNVs called onall 3 of the main platforms (size range: 333 Kb-9.8 Mb).All CNVs called by SMRT AGH are listed in AdditionalFile 6.Detection of X chromosomal CNVsThe X chromosome was analysed separately becausehalf of our hybridizations are sex-mismatched and theAffymetrix and NimbleGen CNV detection software areunable to correct for this. Therefore, identifying CNVsin sex-mismatched hybridizations on these platformsrequired manual assessment, a process that is inherentlymore subjective than the automated assessment used forthe other platforms.There were 117 X-chromosomal CNVs identified onone or more of the 3 main platforms in these 30 MRtrios (Additional File 7). 23 of these 117 CNVs wereidentified in 9 females. 101 (86%) of the CNVs wereidentified by only 1 platform (Figure 3).Fewer X-chromosomal CNV calls were made on theAgilent (38) and Affymetrix 500 K (26) platforms thanon the NimbleGen platform (73). A larger number ofdeletions (78) than duplications (39) were identified onthe X chromosome. The median size of the deletions(131 Kb) was similar to that of the duplications (97 Kb,p = 0.175). The median size of the X-chromosomalCNVs detected by the Agilent platform (32 Kb) was sig-nificantly smaller than that of the CNVs identified bythe Affymetrix 500 K (209 Kb, p < 0.001) or NimbleGen(168 Kb, p < 0.001) platforms.Four of the 117 X-chromosomal CNVs called on the 3main platforms were also identified by the Illumina plat-form, 6 by the Affymetrix 6.0 platform and 1 by theSMRT BAC platform (Additional File 7).Autosomal de novo CNVs with Potential ClinicalSignificanceIn a clinical laboratory, it is not usually possible to usemultiple platforms to determine which CNV calls arereal, and only a subset of the CNVs called with anytechnology is likely to be pathogenic. We used pre-viously published criteria to determine which CNVsidentified by one or more of the Agilent, NimbleGen,Affymetrix 500 K and Affymetrix 6.0 platforms are likelyto be pathogenic (Additional File 4).Only autosomal CNVs that occurred de novo wereassessed for pathogenicity. Although inherited CNVsmay cause MR [24-27], they are much less likely to doso than de novo CNVs, and there was no clinical reasonto suspect a pathogenic inherited CNV in any of thesechildren.Smaller CNVs are much more likely than large CNVsto be false positives and to be benign rather than patho-genic [23]. Therefore, we restricted the analysis for likelypathogenicity to de novo CNVs that were 50 Kb or lar-ger. This eliminated from consideration 51 CNVs (36deletions and 15 amplifications) that were < 50 Kb.We also eliminated from consideration 21 de novoCNVs (13 deletions and 8 duplications) that did notcontain validated open reading frames because suchAGILENT 244K NIMBLEGEN 385KAFFYMETRIX 500K38 73261 24919(73%)24(63%)58(79%)Figure 3 Venn diagram of X chromosome CNV calls made bythe 3 main AGH platforms. The numbers under each platformname indicate the total number of CNV calls by that platform. Thenumbers in the intersecting regions present CNV calls made bymultiple platforms. The numbers outside the intersecting regionsare the number of CNVs that were unique to that platform.Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 4 of 10CNVs cannot be interpreted as pathogenic unless theyinvolve a non-coding region known to be associatedwith MR. We eliminated a further 30 de novo CNVs (20deletions and 10 duplications) that occurred in regionsthat contain only genetically unstable highly repetitivegenes such as olfactory receptor genes or immunoglobingenes. In addition, given the small sample size, any denovo CNV that was identified in 4 or more probandswith differing phenotypes and that was not known to bepathogenic for MR was deemed unlikely to be patho-genic, and we eliminated an additional 19 CNVs (8 dele-tions and 11 duplications) for this reason.We subjected 21 of the remaining 25 de novo CNVsidentified in 14 individuals to FISH or MLPA validation.Eight of these de novo CNVs were called by all 3 of themain platforms, and all 8 were confirmed by FISH orMLPA. In contrast, the 11 de novo CNVs identified onjust one of the 3 main platforms and two de novo CNVsidentified by two of the 3 main platforms could not bevalidated by FISH or MLPA.X Chromosomal CNVs with Potential Clinical SignificanceWe considered 31 of the 117 X-chromosomal CNVsthat were less than 50 Kb and 21 that did not containopen reading frames unlikely to be pathogenic andeliminated them from further assessment. 32 other X-chromosomal CNVs were recurrent within this studypopulation and were also eliminated from furtheranalysis. One 7 Mb de novo amplification in a malewas validated by FISH (Patient 8960) and was identi-fied by all 3 of the main platforms. 17 other CNVswere tested with PCR or qPCR and could not be con-firmed (Additional File 1). These 17 CNVs wereeliminated from further consideration as beingpathogenic.Genotype-Phenotype CorrelationsOf the 28 remaining de novo CNVs, 12 were autosomaland 16, X-linked. We considered 4 autosomal de novoCNVs (Table 1) and 11 X chromosomal CNVs (Table 2)to be unlikely to cause MR because they contained noRefSeq genes that appeared to be reasonable candidatesfor pathogenicity. Two X-chromosomal CNVs result ina MR phenotype in males but not in females [28,29],and both were eliminated from further considerationbecause they occurred in a female (Patient 7093). One Xchromosomal deletion in a male (Patient 3094) waseliminated because it results in MR when deleted infemales but not in males [30]. Another maternally inher-ited duplication was eliminated in a female (Patient1815) because only deletions have been reported toresult in MR in females [31].Table 1 Summary of autosomal de novo CNVs identified on the three main AGH platforms selected for genotype-phenotype analysisTrioIDChr CNVTypeStart* Size* PlatformsIdentified CNV# of RefSeqGenesValidation Comment on Gene Function1815 3 DEL 196 904 149 54 518 Agilent 1 NT Mucin 20 - Expression pattern not consistentwith causing MR [40]4821 5 DEL 68 950 015 1 329 642 NimbleGen 7 NT Mutations in SMN1 associated with spinalmuscle atrophy [41]8960 5 DUP 180 309 941 55 922 NimbleGen 2 MLPA Pos Expression pattern not consistent with causingMR [42]1815 6 DEL 111 807 663 9 889 630 All 3 platforms 57 FISH Pos Likely pathogenic based on size7531 9 DEL 139 496 489 333 935 All 3 platforms 7 FISH Pos CNVs in region previously reported aspathogenic [32]1815 12 DEL 11 371 263 83 667 NimbleGen 1 MLPA Pos Expression pattern not consistent with causingMR [43]1056 13 DEL 107 190 506 2 206 948 All 3 platforms 5 FISH Pos Encompassed within de novo CNV in DECIPHERpatient with MR4821 16 DEL 3 862 993 78 891 All 3 platforms 1 MLPA Pos CNVs in region previously reported aspathogenic [35]3921 17 DEL 41 062 469 657 364 All 3 platforms 8 FISH Pos CNVs in region previously reported aspathogenic [33]9609 21 DEL 33 902 218 152 885 All 3 platforms 2 MLPA Pos Important in spinal development [37]9609 22 DEL 19 062 809 728 798 All 3 platforms 19 FISH Pos CNVs in region previously reported aspathogenic [34,44]8327 22 DUP 19 412 033 378 797 All 3 platforms 13 MLPA Pos Mutation has been reported in family withnormal phenotype [25]DEL = Deletion; DUP = Amplification, NT - Not tested; Pos = Positive; Neg = Negative; N/T = Not tested.*Start/end coordinates determined from largest region of overlap between any two platforms; size is the difference between these two coordinates (Build 36).Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 5 of 10Of the remaining 9 de novo CNVs, 7 are likely to bepathogenic. Table 3 summarizes the phenotypes of thesepatients. The CNVs that are likely to be pathogenicinclude deletions within 9q34.3 (Patient 7531), 17q21.31(Patient 3921) and 22q11.2 (Patient 9609) that havebeen previously reported to be pathogenic in otherpatients with similar phenotypes [32-34]. A 9.8 MBdeletion (Patient 1815) and 7 Mb duplication (Patient8960) are likely to be pathogenic based on their size andthe number of genes affected. Patient 4821 has a 78 Kbdeletion within the first exon and upstream sequence ofthe CREB binding protein (CREBBP), haploinsufficiencyof which causes the Rubinstein-Taybi syndrome [35].The phenotype of Patient 4821 is consistent with thisdiagnosis. The seventh patient has a previously unde-scribed 2.2 Mb deletion of chromosome 13q11 (Patient1056) that encompasses 5 genes, including myosin 16(MYO16), which codes for a protein that interacts withknown synaptic proteins that are important forcognition [36]. The deletion in this patient falls within a9.9 Mb de novo deletion in another patient with MRwho is listed in DECIPHER (DECIPHER patient ID4668).These 7 de novo CNVs that are likely to be pathogenicwere all identified on all three of the main AGH plat-forms studied - Agilent, NimbleGen and Affymetrix 500K - as well as on the Illumina platform. Five of these 7cases were also identified by SMRT AGH (AdditionalFiles 3 and 7), and the Affymetrix 6.0 platform detectedthe only CNV tested on that platform that is very likelyto be pathogenic (Patient 4821). None of the de novoCNV calls made only on the Affymetrix 6.0 platformoccurred in regions that are known to be pathogenic forMR.Two other validated de novo CNVs may be patho-genic. The first is a 152 Kb deletion of chromosome 21that involves intersectin 1, which regulates endocytosisand dendritic spine development [37]. This deletionTable 2 Summary of X chromosome CNVs identified on the three main AGH platforms selected for genotype-phenotype analysisTrioIDCNVTypeStart* Size* PlatformsIdentified CNVGenes Involved Validation Comment on Gene Function6428 DEL 148 264 112 156 992 Affy IDS NT Expression pattern not consistent withcausing MR2894 DEL 101 266 713 250 045 NimbleGen 5 RefSeq genes NT Expression pattern not consistent withcausing MR [45]3519 DEL 9 454 329 197 920 Affy TBL1X NT Expression pattern not consistent withcausing MR [46]8960 DUP 67 416 262 7 057 217 All 3 platforms 57 RefSeq genes FISH Pos FISH Pos9313 DEL 9 484 049 165 559 Affy TBL1X NT Expression pattern not consistent withcausing MR3921 DUP 74 811 330 208 698 NimbleGen TTC3L NT Expression pattern not consistent withcausing MR [47]1511 DEL 6 856 649 201 556 Affy HDHD1A NT Expression pattern not consistent withcausing MR [48]2714 DEL 76 534 899 67 182 NimbleGen FGF16 NT Expression pattern not consistent withcausing MR [49]5993 DEL 6 932 549 130 699 NimbleGen HDHD1A NT Expression pattern not consistent withcausing MR [48]4821 DEL 6 625 133 419 230 NimbleGen HDHD1A NT Expression pattern not consistent withcausing MR [48]3921 DUP 74 811330 208 698 NimbleGen MAGEE2 NT Expression pattern not consistent withcausing MR [47]8960 DEL 29 967 317 203 809 Affy MEGB2E NT Expression pattern not consistent withcausing MR [50]1815 DUP 67 767 923 2 019 581 NimbleGen 15 RefSeq genesincluding DLG3NT Amplification not reported to cause MR[31]7093 DUP 73 429 587 263 642 NimbleGen 3 RefSeq genes includingSLC16A2NT Females not affected [28]3094 DEL 99 293 227 205 443 Affy PCDH19 NT Males not affected [30]7093 DEL 6 687 308 906 505 Agilent &NimbleGenSTS NT Females not affected [29]DEL = Deletion; DUP = Amplification, NT - Not tested; Pos = Positive; Neg = Negative; N/T = Not tested.*Start/end coordinates determined from largest region of overlap between any two platforms; size is the difference between these two coordinates (Build 36).Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 6 of 10occurred in Patient 9609, who also has a pathogenic 728Kb deletion of chromosome 22. The second possiblypathogenic CNV is a 378 Kb duplication that involvesthe distal portion of the 22q11.2 DGS/VCFS region(Patient 8327). A similar duplication was previouslyreported in a child and father whose cognitive abilitywas not clearly described [25].DiscussionIn this study we compared the performance of variousAGH systems for the clinical detection of pathogenicCNVs in children with MR. Previous studies thatincluded more limited comparisons of AGH technolo-gies found substantial differences in the CNV detectionfrequency between platforms, but some of these studiescompared much lower resolution techniques andfocused on larger CNVs [18] or compared lower- tohigher-resolution arrays in an analysis that treated thehigher-resolution findings as correct when discrepancyoccurred [20]. In addition, most previous comparativestudies were retrospective, focusing on the ability of var-ious platforms to identify previously characterizedCNVs. These studies only report detection of pathogenicCNVs and do not discuss findings with respect to themore frequent apparently benign variants [18-20].Here we compared de novo CNVs identified on 3 plat-forms by analysing each child directly in relationship tohis/her parent. This approach was the most cost-effi-cient to distinguish de novo and inherited CNVs usingthe comparative AGH methodology. However, using theparents as reference means that half of the hybridiza-tions involve comparisons between samples from differ-ent sexes, and copy number estimates involving theX-chromosome(s) requires manual CNV identificationby analysis of raw log2 ratios with the NimbleGen andAffymetrix 500 K software when there is a sexmismatch.Each of the three main AGH platforms detected hun-dreds of autosomal CNVs in these 30 trios - an averageof 34 CNVs per trio on NimbleGen arrays, of 22 CNVson Aglient arrays, and 7 CNVs on Affymetrix 500 Karrays. What is most striking, however, is that 82% ofthe CNV calls were only made on one platform, sug-gesting a majority of false positive calls. As expected,most of the autosomal CNVs were inherited from oneof the parents. However, many of the CNVs called inthe child are probably not actually present in the childbut rather represent a copy number change in the oppo-site direction in the parent, e.g., a copy number losscalled in the child against the mother but not the fathercould actually be a copy number gain in the motherthat was not transmitted to the child.Overall, 146 autosomal de novo CNVs and 117X-chromosomal CNVs were called on the 3 main plat-forms. 48 (32 de novo autosome and 16 X chromosome)of these CNVs were found on more than one platform.10 de novo CNVs (9 autosome and 1 X chromosome)were called by all three main platforms. Genotype-phe-notype correlations identified 7 CNVs that are likely tobe pathogenic and 2 other CNVs that are goodTable 3 Summary of phenotypes in patients with validated pathogenic or possibly pathogenic CNVsID Chr CNV Start* Size* Pathogenicity Phenotype1815 6 DEL 111 807 663 9 889 630 Likely MR, microcephaly, epicanthic folds, small ears, hypoplastic lobes, micrognathia,brachycephaly, hypotonia7531 9 DEL 139 496 489 333 935 Likely Moderate global developmental delay, microcephaly, flat face, upslanting palpebralfissures, hypertelorism, synophrysm, anteverted nares, hypoplasia of theamygdalo-hippocampic complex1056 13 DEL 107 190 506 2 206 948 Likely Moderate MR, upslanting palpebral fissures, retrognathia4821 16 DEL 3 862 993 78 891 Likely Moderate MR, microcephaly, short stature, bilateral glaucoma, bilateralcolobomas of the optic nerves, neuro-sensorial deafness, large ASD, epicanthicfolds, low nasal septum, preauricular pits, low set ears, broad distal phalangesof all fingers and toes, cryptorchidy, hypotonia3921 17 DEL 41 062 469 657 364 Likely Mild MR, sagittal craniosynostosis, malar hypoplasia, mild retrognathia, shortand upslanting palpebral fissures, low set ears, high arched palate, broadproximal phalangeal joints of the hands, unilateral cryptorchidism9609 22 DEL 19 062 809 728 798 Likely Moderate MR, microcephaly, short stature, down-slanting palpebral fissures,low-set ears, wide nasal base, retrognathia, metopic craniosynostosis, cleftpalate, partial agenesis of the corpus callosum, tetralogy of Fallot8960 X DUP 67 416 262 7 057 217 Likely Moderate MR, brachycephaly, bilateral epicanthic folds, posteriorly rotatedears with hypoplastic helix and hypotnic9609 21 DEL 33 902 218 152 885 Possible See above pathogenic mutation8327 22 DUP 19 412 033 378 797 Possible Mild MR, small stature, Pierre Robin sequence with cleft palateDEL = Deletion; DUP = Amplification.*Start/end coordinates determined from largest region of overlap between any two platforms; size is the difference between these two coordinates (Build 36).Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 7 of 10candidates to cause MR. All 9 of the pathogenic or pos-sibly pathogenic CNVs were identified by each of thethree main platforms.Although we did not fully assess the Illumina Bead-chips, their sensitivity appears to be similar to that ofthe Affymetrix 500 K arrays. Only 15 trios were assessedon the Affymetrix 6.0 platform. However, it is clear thisplatform produces many more CNV calls, although itsdetection rate for pathogenic and possibly pathogenicCNVs appears to be similar to that of the other SNParrays. None of the additional de novo CNVs called onthe Affymetrix 6.0 platform appears to be pathogenic.The SMRT array only identified 5 of the 9 CNVs thatwere thought to be pathogenic or possibly pathogenic.One of the CNVs not identified was only 78 Kb in sizeand is probably below the resolution of the SMRT array.It is not clear why the other CNVs were not called bythe SMRT array.A number of differences exist among the platformsstudied that may have contributed to the differentresults. Given the large number of genomic segmentsthat were tested, there is a high probability that some ofthe de novo CNV calls are false positives in the probandand others are false negatives in a transmitting parent(i.e., the CNV is actually inherited, rather than de novo,in the proband). Differences in pre-processing, labelling,and hybridization protocols, which were performedaccording to the various manufacturers’ specifications(see Additional File 1), could contribute to the occur-rence of false negative and false positive calls. The low-est observed correlation between platforms was forsmaller CNVs (data not shown) which highlights theimportance of using probe number as a variable foridentifying CNVs. Nevertheless, the lack of concordancein every single two-way comparison and the fact thatthere were de novo CNVs that were identified by one ofthe platforms that were not identified by any of theothers make it very likely that neither optimization ofthe hybridization conditions nor optimization of thebioinformatic analysis parameters would produce perfectconcordance.Distinguishing pathogenic and benign CNVs is amajor part of clinical CNV analysis and goes wellbeyond the software analysis performed on the data. Weperformed genotype-phenotype correlations for eachde novo CNV using methods similar to those employedclinically, which have recently been discussed at length[4,12,13,22,23]. The size cut-off used in our study toassess pathogenicity (50 Kb) was arbitrary; however,almost all pathogenic CNVs detected by oligonucleotideAGH in recently reported studies of children with MRare much larger than 50 Kb [38,39].Although the oligonucleotide or SNP-based AGHtechnologies studied detected all of the pathogenic orpossibly pathogenic CNVs in the 30 MR trios studied,the need to manually assess CNVs on the sex chromo-somes when there is a sex-mismatch with the Nimble-Gen and Affymetrix 500 K software is an importantconsideration when testing for MR or other genomicdisorders in clinical service laboratories. Our resultsshow that the tiling BAC array is less sensitive than theoligonucleotide or SNP-based arrays studied. In anycase, the large number of apparently false positive CNVcalls obtained with each of the platforms studied sup-ports the need for validating all such calls with a differ-ent methodology before consideration of their possiblepathogenicity in a clinical setting.ConclusionsIt seems unlikely that any of the AGH platforms testedis completely right (or completely wrong) in its CNVcalls. Clinical use of any AGH platform to detect patho-genic CNVs in children with birth defects continues torequire considerable skill and experience.Additional materialAdditional file 1: Technical notes. Detailed protocols for AGH and CNVsize comparison between AGH platforms. In addition, there are detailedprotocols for CNV validation and brief discussion of the differencebetween each AGH platform.Additional file 2: Additional Table 5. Summary of inherited CNVsidentified in 30 MR trios identified by the Agilent 244 K, NimbleGen 385K and Affymetrix 500 K arrays and correlation with Illumina Hap550, 32 KSMRT BAC and Affymetrix 6.0 platforms.Additional file 3: Additional Table 6. Summary of de novo CNVsidentified in 30 MR trios identified by the Agilent 244 K, NimbleGen 385K and Affymetrix 500 K arrays and correlation with Illumina Hap550, 32 KSMRT BAC and Affymetrix 6.0 platforms.Additional file 4: Additional Table 7. Summary of all CNV calls madeby the Affymetrix 6.0 array in 15 MR trios studied.Additional file 5: Additional Table 8. Summary of all CNV calls madeby the Illumina Hap500 beadchip in 30 MR patients studied.Additional file 6: Additional Table 9. Summary of all CNV calls madeby the 32 K SMRT BAC tiling path array in 30 MR patients studies.Additional file 7: Additional Table 10. Summary of all X chromosomeCNV in 30 MR trios identified by the Agilent 244 K, NimbleGen 385 K andAffymetrix 500 K arrays and correlation with Illumina Hap550, 32 K SMRTBAC and Affymetrix 6.0 platforms.AcknowledgementsThe authors would like to acknowledge Dr Stéphane LeBihan and AnneHaegart at the Vancouver Prostate Centre Microarray Facility for theirassistance with the Agilent array protocol. This work was funded by grantsfrom the Canadian Institutes of Health Research to JMF, the Advocates forthee Rights of Citizens with Developmental Delay (ARC) of Washington Stateto PE, the Réseau de Génétique Médicale Appliquée du Fonds de laRecherche en Santé du Québec (JLM, EL), and by the Fondsd’encouragement à la recherche clinique du CHU Sainte-Justine (JLM, EL).JM is a Clinician Investigator of the Canadian Institutes of Health Research(Institute of Genetics).The data resulting from this project have been shared with the researchcommunity through publicly accessible Gene Expression Omnibus (GEO)Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 8 of 10under the accession ID GSE27367 http://www.ncbi.nlm.nih.gov/projects/geo/query/acc.cgi?acc=GSE27367.Author details1Department of Medical Genetics, University of British Columbia, Vancouver,British Columbia, Canada. 2McGill University and Genome Quebec InnovationCentre, Montréal, Quebec, Canada. 3Children’s & Women’s Hospital,Vancouver, British Columbia, Canada. 4Genome Sciences Centre, BC CancerAgency, Vancouver, British Columbia, Canada. 5CHU Sainte-Justine ResearchCenter, Montréal, Quebec, Canada. 6British Columbia Cancer Research Centre,Vancouver, British Columbia, Canada. 7Child & Family Research Institute,Vancouver, British Columbia, Canada. 8Center of Excellence in Neuromics ofUniversité de Montréal, Montréal, Quebec, Canada. 9CHUM Research Center,Montréal, Quebec, Canada.Authors’ contributionsTT, AM, and SC and carried out the microarray studies. TT, AM, SC, AD, HQ,DV and JLM analysed the microarray data. TT and AM compared microarrayplatform data and TT performed the statistical analysis. TT, AM and JMFdrafted the manuscript. DC, EM and TT performed molecular analysis tovalidated CNVs. BC and WLL provided array analysis software. GAR and JLMprovided patients. PE, SL, JLM, JFM conceived the study and participated instudy design.Competing interestsThe authors declare that they have no competing interests. All authors readand approved the final manuscript.Received: 17 June 2010 Accepted: 25 March 2011Published: 25 March 2011References1. Shaffer LG: American College of Medical Genetics guideline on thecytogenetic evaluation of the individual with developmental delay ormental retardation. Genet Med 2005, 7:650-654.2. Shaffer LG, Bejjani BA, Torchia B, Kirkpatrick S, Coppinger J, Ballif BC: Theidentification of microdeletion syndromes and other chromosomeabnormalities: cytogenetic methods of the past, new technologies forthe future. Am J Med Genet C Semin Med Genet 2007, 145C:335-345.3. Stankiewicz P, Beaudet AL: Use of array CGH in the evaluation ofdysmorphology, malformations, developmental delay, and idiopathicmental retardation. Curr Opin Genet Dev 2007, 17:182-192.4. Zahir F, Friedman JM: The impact of array genomic hybridization onmental retardation research: a review of current technologies and theirclinical utility. Clin Genet 2007, 72:271-287.5. Kidd JM, Cooper GM, Donahue WF, et al: Mapping and sequencing ofstructural variation from eight human genomes. Nature 2008, 453:56-64.6. Levy S, Sutton G, Ng PC, et al: The diploid genome sequence of anindividual human. PLoS Biol 2007, 5:e254.7. Redon R, Ishikawa S, Fitch KR, et al: Global variation in copy number inthe human genome. Nature 2006, 444:444-454.8. Wang J, Wang W, Li R, et al: The diploid genome sequence of an Asianindividual. Nature 2008, 456:60-65.9. Wheeler DA, Srinivasan M, Egholm M, et al: The complete genome of anindividual by massively parallel DNA sequencing. Nature 2008, 452:872-876.10. Friedman JM: High-resolution array genomic hybridization in prenataldiagnosis. Prenat Diagn 2009, 29:20-28.11. Manning M, Hudgins L: Use of array-based technology in the practice ofmedical genetics. Genet Med 2007, 9:650-653.12. Rodriguez-Revenga L, Mila M, Rosenberg C, Lamb A, Lee C: Structuralvariation in the human genome: the impact of copy number variants onclinical diagnosis. Genet Med 2007, 9:600-606.13. Vermeesch JR, Fiegler H, de Leeuw N, et al: Guidelines for molecularkaryotyping in constitutional genetic diagnosis. Eur J Hum Genet 2007,15:1105-1114.14. Aradhya S, Cherry AM: Array-based comparative genomic hybridization:clinical contexts for targeted and whole-genome designs. Genet Med2007, 9:553-559.15. Shaikh TH: Oligonucleotide arrays for high-resolution analysis of copynumber alteration in mental retardation/multiple congenital anomalies.Genet Med 2007, 9:617-625.16. Veltman JA, de Vries BB: Whole-genome array comparative genomehybridization: the preferred diagnostic choice in postnatal clinicalcytogenetics. J Mol Diagn 2007, 9:277.17. Zhang ZF, Ruivenkamp C, Staaf J, et al: Detection of submicroscopicconstitutional chromosome aberrations in clinical diagnostics: avalidation of the practical performance of different array platforms. Eur JHum Genet 2008, 16:786-792.18. Aston E, Whitby H, Maxwell T, et al: Comparison of targeted and wholegenome analysis of postnatal specimens using a commercially availablearray based comparative genomic hybridisation (aCGH) microarrayplatform. J Med Genet 2008, 45:268-274.19. Hehir-Kwa JY, Egmont-Petersen M, Janssen IM, Smeets D, van Kessel AG,Veltman JA: Genome-wide copy number profiling on high-densitybacterial artificial chromosomes, single-nucleotide polymorphisms, andoligonucleotide microarrays: a platform comparison based on statisticalpower analysis. DNA Res 2007, 14:1-11.20. Wicker N, Carles A, Mills IG, et al: A new look towards BAC-based arrayCGH through a comprehensive comparison with oligo-based array CGH.BMC Genomics 2007, 8:84.21. de Vries BB, White SM, Knight SJ, et al: Clinical studies onsubmicroscopic subtelomeric rearrangements: a checklist. J Med Genet2001, 38:145-150.22. Carter NP: Methods and strategies for analyzing copy number variationusing DNA microarrays. Nat Genet 2007, 39:S16-21.23. Lee C, Iafrate AJ, Brothman AR: Copy number variations and clinicalcytogenetic diagnosis of constitutional disorders. Nat Genet 2007, 39:S48-54.24. Brunetti-Pierri N, Berg JS, Scaglia F, et al: Recurrent reciprocal 1q21.1deletions and duplications associated with microcephaly ormacrocephaly and developmental and behavioral abnormalities. NatGenet 2008, 40:1466-1471.25. Ou Z, Berg JS, Yonath H, et al: Microduplications of 22q11.2 arefrequently inherited and are associated with variable phenotypes. GenetMed 2008, 10:267-277.26. Rosenberg C, Knijnenburg J, Bakker E, et al: Array-CGH detection of microrearrangements in mentally retarded individuals: clinical significance ofimbalances present both in affected children and normal parents. J MedGenet 2006, 43:180-186.27. van Bon BW, Mefford HC, Menten B, et al: Further delineation of the15q13 microdeletion and duplication syndromes: a clinical spectrumvarying from non-pathogenic to a severe outcome. J Med Genet 2009,46:511-523.28. Dumitrescu AM, Liao XH, Best TB, Brockmann K, Refetoff S: A novelsyndrome combining thyroid and neurological abnormalities isassociated with mutations in a monocarboxylate transporter gene. Am JHum Genet 2004, 74:168-175.29. Shapiro LJ, Yen P, Pomerantz D, Martin E, Rolewic L, Mohandas T:Molecular studies of deletions at the human steroid sulfatase locus. ProcNatl Acad Sci USA 1989, 86:8477-8481.30. Dibbens LM, Tarpey PS, Hynes K, et al: X-linked protocadherin 19mutations cause female-limited epilepsy and cognitive impairment. NatGenet 2008, 40:776-781.31. Tarpey P, Parnau J, Blow M, et al: Mutations in the DLG3 gene causenonsyndromic X-linked mental retardation. Am J Hum Genet 2004,75:318-324.32. Kleefstra T, Koolen DA, Nillesen WM, et al: Interstitial 2.2 Mb deletion at9q34 in a patient with mental retardation but without classical featuresof the 9q subtelomeric deletion syndrome. Am J Med Genet A 2006,140:618-623.33. Koolen DA, Sharp AJ, Hurst JA, et al: Clinical and molecular delineation ofthe 17q21.31 microdeletion syndrome. J Med Genet 2008, 45:710-720.34. Rauch A, Zink S, Zweier C, et al: Systematic assessment of atypicaldeletions reveals genotype-phenotype correlation in 22q11.2. J MedGenet 2005, 42:871-876.35. Schorry EK, Keddache M, Lanphear N, et al: Genotype-phenotypecorrelations in Rubinstein-Taybi syndrome. Am J Med Genet A 2008,146A:2512-2519.36. Patel KG, Liu C, Cameron PL, Cameron RS: Myr 8, a novel unconventionalmyosin expressed during brain development associates with the proteinphosphatase catalytic subunits 1alpha and 1gamma1. J Neurosci 2001,21:7954-7968.Tucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 9 of 1037. Thomas S, Ritter B, Verbich D, et al: Intersectin regulates dendritic spinedevelopment and somatodendritic endocytosis but not synaptic vesiclerecycling in hippocampal neurons. J Biol Chem 2009, 284:12410-12419.38. Bruno DL, Ganesamoorthy D, Schoumans J, et al: Detection of crypticpathogenic copy number variations and constitutional loss ofheterozygosity using high resolution SNP microarray analysis in 117patients referred for cytogenetic analysis and impact on clinical practice.J Med Genet 2009, 46:123-131.39. Fan YS, Jayakar P, Zhu H, et al: Detection of pathogenic gene copynumber variations in patients with mental retardation by genomewideoligonucleotide array comparative genomic hybridization. Hum Mutat2007, 28:1124-1132.40. Moehle C, Ackermann N, Langmann T, et al: Aberrant intestinal expressionand allelic variants of mucin genes associated with inflammatory boweldisease. J Mol Med 2006, 84:1055-1066.41. Ogino S, Wilson RB: Spinal muscular atrophy: molecular genetics anddiagnostics. Expert Rev Mol Diagn 2004, 4:15-29.42. Shibui A, Tsunoda T, Seki N, Suzuki Y, Sugano S, Sugane K: Cloning,expression analysis, and chromosomal localization of a novelbutyrophilin-like receptor. J Hum Genet 1999, 44:249-252.43. Azen EA, Maeda N: Molecular genetics of human salivary proteins andtheir polymorphisms. Adv Hum Genet 1988, 17:141-199.44. Garcia-Minaur S, Fantes J, Murray RS, et al: A novel atypical 22q11.2 distaldeletion in father and son. J Med Genet 2002, 39:E62.45. Alvarez E, Zhou W, Witta SE, Freed CR: Characterization of the Bex genefamily in humans, mice, and rats. Gene 2005, 357:18-28.46. Bassi MT, Ramesar RS, Caciotti B, et al: X-linked late-onset sensorineuraldeafness caused by a deletion involving OA1 and a novel genecontaining WD-40 repeats. Am J Hum Genet 1999, 64:1604-1616.47. Chomez P, De Backer O, Bertrand M, De Plaen E, Boon T, Lucas S: Anoverview of the MAGE gene family with the identification of all humanmembers of the family. Cancer Res 2001, 61:5544-5551.48. Yen PH, Ellison J, Salido EC, Mohandas T, Shapiro L: Isolation of a newgene from the distal short arm of the human X chromosome thatescapes X-inactivation. Hum Mol Genet 1992, 1:47-52.49. Miyake A, Konishi M, Martin FH, et al: Structure and expression of a novelmember, FGF-16, on the fibroblast growth factor family. Biochem BiophysRes Commun 1998, 243:148-152.50. Lurquin C, De Smet C, Brasseur F, et al: Two members of the humanMAGEB gene family located in Xp21.3 are expressed in tumors ofvarious histological origins. Genomics 1997, 46:397-408.Pre-publication historyThe pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1755-8794/4/25/prepubdoi:10.1186/1755-8794-4-25Cite this article as: Tucker et al.: Comparison of genome-wide arraygenomic hybridization platforms for the detection of copy numbervariants in idiopathic mental retardation. BMC Medical Genomics 20114:25.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 redistributionSubmit your manuscript at www.biomedcentral.com/submitTucker et al. BMC Medical Genomics 2011, 4:25http://www.biomedcentral.com/1755-8794/4/25Page 10 of 10


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