Genome-wide copy number variant discovery in dogs using the CanineHD genotyping array
© Molin et al.; licensee BioMed Central Ltd. 2014
Received: 20 February 2014
Accepted: 13 March 2014
Published: 19 March 2014
Substantial contribution to phenotypic diversity is accounted for by copy number variants (CNV). In human, as well as other species, the effect of CNVs range from benign to directly disease-causing which motivates the continued investigations of CNVs. Previous canine genome-wide screenings for CNVs have been performed using high-resolution comparative genomic hybridisation arrays which have contributed with a detailed catalogue of CNVs. Here, we present the first CNV investigation in dogs based on the recently reported CanineHD 170 K genotyping array. The hitherto largest dataset in canine CNV discovery was assessed, 351 dogs from 30 different breeds, enabling identification of novel CNVs and a thorough characterisation of breed-specific CNVs.
A stringent procedure identified 72 CNV regions with the smallest size of 38 kb and of the 72 CNV regions, 38 overlapped 148 annotated genes. A total of 29 novel CNV regions were found containing 44 genes. Furthermore, 15 breed specific CNV regions were identified of which 14 were novel and some of them overlapped putative disease susceptibility genes. In addition, the human ortholog of 23 canine copy number variable genes identified herein has been previously suggested to be dosage-sensitive in human.
The present study evaluated the performance of the CanineHD in detecting CNVs and extends the current catalogue of canine CNV regions with several dozens of novel CNV regions. These novel CNV regions, which harbour candidate genes that possibly contribute to phenotypic variation in dogs or to disease-susceptibility, are a rich resource for future investigations.
KeywordsCopy number variation CNV SNP genotyping array Dog genome Deletion Duplication CanineHD
About a decade ago the first large catalogs of copy number variants (CNVs) in the human genome were presented [1, 2]. Numerous studies later, CNVs are known to contribute to the genomic variation to a larger extent than single nucleotide polymorphisms (SNPs) in terms of number of nucleotide differences [3, 4]. CNVs are defined as DNA segments of variable length, up to several megabases (Mb), that varies in copy numbers in comparison to a reference genome . The different types of CNVs include deletions and duplications, while consequences of CNVs include e.g. altered gene dosage and regulation, changed gene structure and unmasking of recessive alleles . The effect of CNVs varies from being benign or neutral, to having subtle effects on disease predisposition or directly causing disease. The contribution and importance of CNVs for phenotypic diversity and disease susceptibility has been repeatedly shown in human and several complex diseases such as psoriasis, Crohn’s disease, type 2 diabetes, coronary heart disease among others have been associated with CNVs [3, 6–9]. Also for the group of patients with intellectual disability and/or multiple congenital anomalies, CNVs have been shown to be of major importance and explain about 10-20% of these patients .
Today, genome-wide screens for CNVs have been conducted in many different species in addition to human, including mouse, pig, cattle, goat and horse [11–22]. These screens, both in human and other species, were enabled by methods to perform high-throughput genome scans. Several methods exist to search genomes for the presence of CNVs, e.g. array-based methods such as array comparative genomic hybridization (aCGH) and whole-genome SNP-genotyping arrays (SNP arrays). The higher the resolution of the arrays the smaller CNVs can be detected with more accuracy and precision. The advantage with SNP arrays is that CNV detection can be performed in conjunction with genome-wide association studies (GWAS) and specific algorithms have been developed to identify CNVs using SNP arrays [23–25]. More recently the next-generation sequencing technologies have emerged to detect CNVs. The necessity for cataloguing CNVs also in animals is evident considering the numerous examples of CNVs involvement in shaping the phenotype in animals . For example, the pigmentation phenotype dermal hyperpigmentation/Fibromelanosis in chicken is caused by a complex rearrangement of the EDN3 locus that involves two duplications . In dogs, the dorsal hair ridge in Rhodesian and Thai Ridgeback is caused by a duplication encompassing three genes, FGF3, FGF4 and FGF19, and this CNV predisposes to dermoid sinus . Also, the wrinkled skin in Chinese Shar-Pei is explained by a CNV, namely a duplication upstream of the gene encoding Hyaluronic acid synthase 2 . Similarly as in ridgebacks, the CNV in Shar-Pei predisposes the breed to a disease, namely periodic fever syndrome . One example of a CNV that has been suggested to confer an advantage in early dogs during domestication is a 8 kb duplication encompassing the amylase gene AMY2B . This duplication was one of the identified selection signals involving genes in starch digestion and it was shown to increase amylase activity suggesting that this duplication contributed to the ability of dogs to thrive on a starch-rich diet in comparison to wolves that lack the duplication .
In dogs, four CNV screens have previously been performed with the aim of identifying canine CNVs and all were based on the aCGH method [31–34]. The analyses by Chen et al. of nine dogs from different breeds identified 60 CNV regions (CNVRs, i.e. region of the genome where CNVs have been identified) using an array with an average probe spacing of 4,675 bp . The contemporary study by Nicholas et al. was directed towards segmentally duplicated regions of the genome and 137 Mb of sequence was studied at a very high resolution (200 bp probe spacing) . This study identified 678 CNVRs using 17 dogs (all from different breeds) and a gray wolf . Recently two studies were reported using the same aCGH that has the highest resolution of existing canine arrays up to date with 2.1 million probes (1 kb probe spacing) [31, 33]. One of these studies used nine dogs and a gray wolf and identified 403 CNVRs  and the hitherto most comprehensive CNV screen in dogs was based on 50 dogs from 17 breeds and 3 wolves and identified 430 CNVRs . In 2011, the development of a high-density canine SNP array (the Illumina CanineHD array) containing 174,943 SNPs with an average probe spacing of 13 kb was reported . This array enables CNV detection simultaneously with SNP genotyping, although the CanineHD array was not evaluated with respect to CNVs in that study . In the present study, the first CNV screen based on the CanineHD array is reported. Using 351 dogs from 30 different breeds, the present study includes the largest number of dogs screened for CNVs hitherto. This investigation illustrates the usefulness of the CanineHD array in CNV detection and demonstrates the size-range of CNVs that are detectable with the CanineHD array. Furthermore, this study identified novel CNVRs including some breed-specific CNVRs.
Results and discussion
CNV discovery and genotyping
Genome-wide CNV analysis was conducted on 359 dogs from 30 different breeds (Additional file 1: Table S1) using the Illumina 170 K CanineHD SNP array [35, 36]. A total of 26 breeds included 7 or more samples, enabling a thorough identification of single breed CNVs, i.e. CNVs present in one breed exclusively. Importantly, 13 breeds had never been analyzed genome-wide for the presence of CNVs before and an additional five breeds had only been genotyped for previously identified CNVs in the study by Nicholas et al. . Therefore, the present study enabled also the identification of novel CNVs.
Briefly, the initial CNV discovery was performed using two different algorithms, PennCNV and QuantiSNP [23, 24, 37, 38]. After quality filtering, 351 dogs remained and the relaxed QuantiSNP result file (Log Bayes factor; LBF ≥ 10) contained 1748 CNVs, the stringent QuantiSNP result file (LBF ≥ 30) contained 493 CNVs and the PennCNV file contained 1455 CNVs. In the quality-filtering step, three large regions of a total size of 9.2 Mb containing 59 genes (Additional file 2: Figure S1 and Additional file 1: Table S2) were excluded due to extensive segmentation of the data creating many overlapping CNV calls. Thorough investigations of these regions were beyond the scope of this study but this observation is of importance in future CNV studies based on the CanineHD array.
The identification of CNV regions (CNVRs) was performed by using a stringent procedure. Overlapping CNVs were merged from the two programs separately, i.e. the stringent list of 493 CNVs identified by QuantiSNP were merged into 116 CNVRs and the 1455 CNVs found by PennCNV yielded 261 CNVRs. The breakpoints of each of these CNVRs were defined by the outer boundaries of the individual CNVs. Finally, the CNVRs from QuantiSNP that did not overlap with CNVRs from PennCNV, with at least 10 kb overlap were excluded. This identified a stringent list of CNVRs where the breakpoints of each of these final CNVRs were obtained from the QuantiSNP CNVRs. The strategy where the final list of stringent CNVRs was found by both algorithms has been recommended previously in order to reduce the false discovery rate .
Altogether 110 CNVRs, distributed on the 38 autosomes, were identified following this process (Additional file 1: Table S3). For these stringent CNVRs, each of the 351 individuals were genotyped using the relaxed QuantiSNP file with 1748 CNVs, hence a less strict threshold was used in CNV genotyping than in CNV discovery (LBF ≥ 10 vs. ≥ 30). Of these 110 CNVRs, 38 were found in only one single sample and thereby denoted singletons (Additional file 1: Table S3). The frequency of singletons in the present study was 0.11 (38/351) singletons per sample which confirm the observation by Berglund et al. of a low number of singletons per sample as compared to the conflicting results reported in Nicholas et al. [31, 33]. Since singletons are more likely to be false-positives than CNVRs that are detected in at least two independent individuals, a conservative approach was selected where singletons were excluded from further analysis. However, future CNVR screens of additional individuals could reveal if some of these singletons (Additional file 1: Table S3) in fact represent rare CNVRs instead.
Summary of identified CNVR
Number of CNVs
No. of genes
No. of CNVs with genes
Number of CNVs
No. of genes
No. of CNVs with genes
Comparison with published CNVRs and validation of CNVRs
The currently identified 72 CNVRs were compared to CNVRs defined in two previously published datasets to look for overlap between the dataset. As threshold for overlap, a minimum of at least 10 kb overlap between the currently identified CNVRs and a previously published CNVR was required [31, 33].
Comparison with previously identified CNVRs using a minimum of 10 kb overlap
Number of CNVRs
10 kb overlap*
CNVRs with frequency >5%
CNVRs with frequency >1%
CNVRs with frequency <1%
The high proportion of CNVRs with a frequency of at least 1% in the population that overlap with previously found CNVRs suggests a high accuracy in the currently described CNV discovery procedure. In fact, an overlap of 73% is higher than the observed reproducibility of <70% for most platforms when the same sample has been analyzed in replicate experiments .
Given that a majority of CNVRs identified herein were supported by previously identified CNVR datasets, a subset of CNVRs was selected for validation only among the novel CNVRs/single breed CNVRs (see "CNVRs in single breeds" and "Dosage sensitive genes and novel CNVRs"). Five deletions and four duplications with a size range from 52 to 343 kb were investigated by quantitative real-time PCR (qRT-PCR) in 28 samples from three breeds (Additional file 1: Table S4). Eight of the nine CNVRs were validated yielding a validation rate of 89% (Additional file 1: Table S4). The unconfirmed CNVR, no. 45, was a deletion present in only 2/351 individuals and only one sample was available for validation. Based on visual inspection of the Log R ratio and the B allele frequency plots, the unavailable sample showed a more convincing deletion than the qRT-PCR-investigated sample (Additional file 3: Figure S2) suggesting that CNVR no. 45 may still be a true CNVR. The genotype concordance between the CanineHD array data and the qRT-PCR validation was 98%. Overall 106 sample-locus combinations were tested, where 20 of these displayed a deletion or duplication in the CanineHD CNV genotype. The results showed that two sample-locus combinations did not match (one false positive and one false negative, Additional file 1: Table S4). In summary, both the validation rate of CNVRs of 89% and the genotyping concordance rate between the two methods of 98% suggest that the described CNV discovery and genotyping procedure is of high accuracy and that the performance of the CanineHD array in detecting CNVs is high.
CNVRs in single breeds
The stringent list of 15 single breed CNVRs
No. of genes
No. of affected
No. of screened
Freq. in breed
DSCC1, TNFRSF11B, ENPP2, COLEC10, NOV, TAF2, MRPL13, DEPTOR, COL14A1, MTBP, SNTB1
C4ORF33, SCLT1, PHF17
Single breed CNVRs are of particular interest if it is fixed in the breed since it could be involved in breed specific characteristics. In the study by Berglund et al. no fixed single breed CNVRs were identified in the two 10-sample breeds included in the study and they suggested that breed specific CNVRs are most often not involved in breed specific characteristics . In the stringent list of 15 single breeds CNVRs identified herein (Table 3), the frequencies of the CNVRs were in the range of 17-42%, hence our results confirm previous suggestions . The frequency of CNVRs in a breed is however not an exact measurement since it can depend on e.g. the discovery procedure, filtering criteria and whether the size of the CNVR is close to the detection limit of the array.
CNVRs and the effect on genes
Examining the CNVRs for the presence of genes, the list of 24,660 genes from Ensembl CanFam2 was filtered for human orthologous genes leaving 17,239 genes . Furthermore, genes located on chromosome X, unknown or mitochondria were first removed and subsequently all non-unique dog entries and all non-unique human entries leaving a total gene list of 16,305 1:1 human-dog orthologous genes. The start and end point of the CNVRs were analyzed for overlap with the start and endpoint of the genes with a minimum of 1 bp overlap. A 1 bp overlap was chosen due to the rough estimates of CNVR breakpoints and consequently some of the resulting overlapping genes located in the vicinity of the breakpoints could in fact be non-overlapping but in close proximity or alternatively be completely within the CNVR.
Analysis of the 72 CNVRs revealed that 38 CNVRs overlapped 148 genes (Additional file 1: Table S6) while the remaining 34 CNVRs contained no protein coding genes. Duplications overlapping genes outnumber deletions (8 deletions vs. 21 duplications) (Table 1), which is a similar observation as previously reported .
The 148 1:1 orthologous genes affected by CNVRs (Additional file 1: Table S6) were analyzed with respect to gene enrichment using the g:GOSt tool from g:Profiler [41, 42]. In the analysis, the human gene IDs corresponding to the canine genes were analyzed against all 16,305 1:1 human-dog orthologous genes that were used as background. The results were corrected for multiple testing using g:GOSt native method g:SCS. The most significant results in the domain molecular function was "olfactory receptor activity" (p = 1.0 × 10-25) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enriched genes belonged to the "Olfactory transduction" pathway (p = 1.1 × 10-23) (Additional file 1: Table S7). This enrichment of olfactory receptor activity genes in CNVRs in dogs has been previously reported . When all olfactory receptor genes were removed (totally 34/148 genes), no significant enriched category was obtained and in the domain molecular function the highest enriched gene category was "riboflavin transporter activity" with p = 2.34 × 10-1.
Of the stringent list of 15 single breed CNVRs, 7 CNVRs overlapped in total 20 genes of which 18 genes were in duplications and two genes were in deletions (Table 3). A gene enrichment analysis using the stringent list of 15 single breed CNVRs showed no statistically significant result neither when all 20 genes were analyzed nor when the 18 genes present in single breed duplications were analyzed separately. Of the two deleted genes, PDE3B was within the estimated CNVR boundary, and 14 of the 18 duplicated genes were within the estimated CNVR boundaries (Table 3, Additional file 1: Table S6). The CNVR, no. 78, encompassing PDE3B was found to affect 4/12 Finnish Spitz and in all individuals a heterozygous deletion was detected. This finding could be of importance for the occurrence of diabetes mellitus in this breed since knockout of the gene in mice has been shown to affect energy homeostasis, including altered insulin secretion and regulation of lipogenesis and lipolysis . One interesting gene residing completely within the duplication CNVR no. 48, is NOV which is heterozygously duplicated along with additional ten genes in two dogs of the Standard Poodle breed (Table 3). NOV has been shown to have a selective pro-apoptotic activity towards adrenocortical cells in human and reduced expression and protein levels has been observed in childhood adrenocortical tumours . These findings could have implications for Standard Poodles since this breed has a genetic predisposition towards Addison’s disease, which is characterized by insufficient production of hormones by the adrenal glands.
Dosage sensitive genes and novel CNVRs
Human-defined dosage sensitive genes within presently identified canine CNVRs
No of affected individuals/total individuals
Golden Retr.Labrador Retr., Flat-coated Retr.
1/12 GR, 1/14 LR, 2/2 FR
Bernese Mountain dog
totally 87 affected
Brittany Spaniels, English Setter
2/12 BS, 7/12 ES
Chinese Shar-Pei, Eurasian
10/11 SP, 4/11 E
In addition to the list of canine genes that have previously been seen to be dosage-sensitive in human , the present study identified 60 novel CNVRs where 31 were singletons and were therefore not considered in detail. The remaining 29 novel CNVRs (Additional file 1: Table S8) were not found in more than three breeds and 14 of these novel CNVRs were single breed CNVRs. There were 44 genes present in 15 of the 29 novel CNVRs and seven of these genes were suggested to be dosage sensitive in human, e.g. SGPP2. The list of novel CNVRs constitutes a resource for future studies.
The present investigation is the first screen for canine CNVs using the CanineHD array and thereby this study demonstrates the performance of the CanineHD array in CNV detection in addition to SNP genotyping. The procedure described herein identified 72 CNVRs with high stringency where the smallest CNVR was 38 kb. A total of 29 novel CNVRs were identified that contribute to the catalogue of known canine CNVs and this also illustrates the importance of continuing the identification of canine CNVRs and the necessity of including a large number of samples and previously unscreened breeds in order to obtain a comprehensive catalog of canine CNVRs. Furthermore, a list of 15 confidently defined single breed CNVRs were found of which as many as 14 were novel CNVRs. These single breed CNVRs could be of interest for further genotype-phenotype analysis although the number of fixed single breed CNVRs are rare as previously shown . Hence, breed-specific characteristics are most likely not hidden among hitherto unknown CNVs or alternatively are smaller than the CNV detection limit of approximately 50 kb on the CanineHD array. Nevertheless, there is still the possibility of having disease-predisposing CNVs among the growing list of defined CNVRs, which of course are much less likely to be fixed in a breed. A final contribution herein to future studies were the identification of 23 copy number variable canine genes that have been suggested as being dosage-sensitive in human and as such are remarkably interesting for future phenotype association analysis in dogs.
Samples and genotyping
A total of 359 dogs from 30 breeds were included in the study (Additional file 1: Table S1). This dataset was a subset of the "Gentrain" dataset that was used to initially evaluate the performance of the CanineHD 170 K SNP array with respect to SNP genotyping [35, 36]. Blood sampling was performed by trained veterinarians according to relevant international guidelines and to the approval of the Swedish Animal Ethical Committee (no. C139/9 and no. C2/12). Genomic DNA was extracted and genotyped using the array according to manufacturer’s instructions. The initial step of the analysis, normalization of the total signal intensity and calculations of Log R ratio and B allele frequencies for the SNPs were performed in the GenomeStudio V2010.3 software package according to the manual and Peiffer et al. . Furthermore, the Log R ratios were GC-corrected, adjusted for GC content surrounding the SNP as recommended to reduce waviness . Subsequently, SNPs with a quality score of zero were excluded, likewise the intensity-only SNPs were removed since the analysis that followed could not manage these SNPs on other arrays than predefined human SNP arrays. This filtering left a total number of 172,115 SNPs for the CNV analysis (available through GEO Series accession number GSE55134) and at ).
CNV discovery and quality filtering
The CNV discovery was performed using two programs; QuantiSNP and PennCNV. Both have been developed for CNV analysis on SNP array data and hence use both the Log R ratio and B allele frequency for each SNP in the CNV calling [23, 24]. In QuantiSNP, default settings were used, i.e. L = 2 M, expectation-maximization iterations = 10, and the parameter file levels-hd.dat as recommended for high-density SNP arrays . In PennCNV, a file containing the population frequency of B alleles was initially created using the samples within the present study. Thereafter the script "detect_cnv.pl" was used with the Hidden Markov Model file hhall.hmm .
The discovered CNVs were filtered based on quality information from PennCNV . As recommended, a standard deviation of the Log R ratio below 0.25 and a waviness factor of >0.04 or < -0.04 were used to filter out samples with bad quality . In addition, samples with 40 CNVs or more were also removed. This quality criteria was based on the observation that the number of CNVs per sample in the PennCNV file were ≤ 30 in 98% of the samples. Totally 8/359 samples did not fulfill the quality criteria and were removed from both the PennCNV and the QuantiSNP resulting CNV files. Subsequently, additional filtering was performed on the results file from both QuantiSNP and PennCNV, i.e. CNVs having less than 5 SNPs/CNV were excluded, as were CNVs on chromosome X. Also, three chromosomal regions were removed based on visual inspection of plotted Log R ratio due to extensive fragmented and noisy CNVs; chr1:13–15.7 Mb, chr15:10.2-13.5 Mb, chr36:16–19.2 Mb (Additional file 2: Figure S1, Additional file 1: Table S2). The QuantiSNP result file was finally filtered using a score, Log Bayes factor (LBF), calculated by the program that represents the support for the existence of the CNV. Two thresholds were used; CNVs with LBF < 10 were removed in the result file as recommended to give a maximum of 10% false positive calls, a threshold with LBF ≥ 10 was used to obtain a relaxed list of CNVs and a threshold with LBF ≥ 30 was used to obtain a stringent list of CNVs [23, 37].
Subsequently, individual CNVs were merged into overlapping CNV regions (CNVR). Merging was performed separately on the result files from PennCNV and from the stringent list from QuantiSNP containing CNVs with LBF ≥ 30. The tool "Operate on genomic intervals" on the Galaxy platform was used for merging . A CNVR was defined as in Redon et al., a region in the genome that can vary in copy number among individuals and the breakpoints of the CNVRs are the outermost boundaries of all the individual CNVs constituting each CNVR . Thereafter, the QuantiSNP result file containing individual CNVs with a relaxed threshold of LBF ≥ 10 was used to genotype each sample for CNVs residing within the defined CNVRs. Hence, a single individual can have a different size of the CNV than the CNVR but it is located within the CNVR boundaries. The CNVRs were defined against CanFam 2.0.
Comparison with published CNVRs and validation of CNVRs
To compare the performance of the CanineHD array in CNV detection and the described procedure to detect CNVRs, the results from two previously published CNV screens were used. The list of the herein identified CNVRs were analyzed for overlapping CNVRs from the two published datasets, using the tool "Operate on genomic intervals" on the Galaxy platform . The two published datasets contain 394 CNVRs  and 1235 CNVRs  after removal of CNVRs on chromosome X and unknown.
Available DNA from 28 samples from three breeds allowed validation of nine CNVRs by quantitative real-time PCR (qRT-PCR) (Additional file 1: Table S4). The CNVRs were tested in breed/breeds where at least one individual per breed displayed a CNV, yielding a total of 106 sample-locus combinations (Additional file 1: Table S4). For each CNVR, one primer pair was designed in the centre of the CNVR based on the estimated boundaries obtained from the CanineHD CNVR discovery (primer sequences, see Additional file 1: Table S4). The qRT-PCR was performed in quadruplicates in a 10 ul volume using SYBR Green PCR Master Mix (Applied Biosystems) in a 7900HT Real Time PCR system (Applied Biosystems) according to manufacturer’s instructions. Final primer concentrations were 800nM and each primer pair was initially evaluated for amplification efficiency using a 5-point serial dilution curve with a 5X dilution factor. A dissociation curve analysis was performed after each run to assess PCR specificity. The 2-∆∆Ct method was used for relative quantification of the copy number where the copy number assignment was obtained by the formula 2 x 2-∆∆Ct . The ∆Ct is calculated for each individual by the average Ct (threshold cycle) of the tested CNVR – average Ct of a reference gene. The ∆∆Ct is given by the ∆Cttested individual - ∆Ctreference individual. The previously used C7orf28B was used as a reference gene (F: 5′-3′ CAACACAGGTTGACCAAGGA and R: 5′-3′ TTGTGCAGGATCAGAGCATC) . The reference individual was chosen among the 351 samples and it displayed neither deletions nor duplications in the CanineHD CNV genotyping in any of the nine investigated CNVRs,
Availability of supporting data
Copy number variant
Copy number variant region
Array comparative genomic hybridization
Genome-wide association study
Single nucleotide polymorphism
Log Bayes factor
Quantitative real-time PCR.
We thank all dog owners who contributed with samples for this project and the LUPA consortium for providing samples used in this project. We thank Gerli Rosengren Pielberg, Merete Fredholm and Jennifer R. S. Meadows for valuable discussions concerning the samples and certain breeds, Göran Andersson for helpful discussions about CNVs and Jeremy Johnson for help with making data accessible at the Broad homepage. We also thank Illumina, especially Mark Hansen and Jim Gordon for support. For financial support, we thank the Swedish Research Council, the European Commission FP7 project LUPA-GA201370 and the European Science Foundation EURYI award to KLT.
- Iafrate AJ, Feuk L, Rivera MN, Listewnik ML, Donahoe PK, Qi Y, Scherer SW, Lee C: Detection of large-scale variation in the human genome. Nat Genet. 2004, 36: 949-951. 10.1038/ng1416.PubMedView ArticleGoogle Scholar
- Sebat J, Lakshmi B, Troge J, Alexander J, Young J, Lundin P, Maner S, Massa H, Walker M, Chi M, Navin N, Lucito R, Healy J, Hicks J, Ye K, Reiner A, Gilliam TC, Trask B, Patterson N, Zetterberg A, Wigler M: Large-scale copy number polymorphism in the human genome. Science. 2004, 305: 525-528. 10.1126/science.1098918.PubMedView ArticleGoogle Scholar
- Girirajan S, Campbell CD, Eichler EE: Human copy number variation and complex genetic disease. Annu Rev Genet. 2011, 45: 203-226. 10.1146/annurev-genet-102209-163544.PubMedView ArticleGoogle Scholar
- Redon R, Ishikawa S, Fitch KR, Feuk L, Perry GH, Andrews TD, Fiegler H, Shapero MH, Carson AR, Chen W, Cho EK, Dallaire S, Freeman JL, Gonzalez JR, Gratacos M, Huang J, Kalaitzopoulos D, Komura D, MacDonald JR, Marshall CR, Mei R, Montgomery L, Nishimura K, Okamura K, Shen F, Somerville MJ, Tchinda J, Valsesia A, Woodwark C, Yang F: Global variation in copy number in the human genome. Nature. 2006, 444: 444-454. 10.1038/nature05329.PubMed CentralPubMedView ArticleGoogle Scholar
- Feuk L, Carson AR, Scherer SW: Structural variation in the human genome. Nat Rev Genet. 2006, 7: 85-97.PubMedView ArticleGoogle Scholar
- Craddock N, Hurles ME, Cardin N, Pearson RD, Plagnol V, Robson S, Vukcevic D, Barnes C, Conrad DF, Giannoulatou E, Holmes C, Marchini JL, Stirrups K, Tobin MD, Wain LV, Yau C, Aerts J, Ahmad T, Andrews TD, Arbury H, Attwood A, Auton A, Ball SG, Balmforth AJ, Barrett JC, Barroso I, Barton A, Bennett AJ, Bhaskar S, Blaszczyk K: Genome-wide association study of CNVs in 16,000 cases of eight common diseases and 3,000 shared controls. Nature. 2010, 464: 713-720. 10.1038/nature08979.PubMedView ArticleGoogle Scholar
- Hollox EJ, Huffmeier U, Zeeuwen PL, Palla R, Lascorz J, Rodijk-Olthuis D, van de Kerkhof PC, Traupe H, de Jongh G, den Heijer M, Reis A, Armour JA, Schalkwijk J: Psoriasis is associated with increased beta-defensin genomic copy number. Nat Genet. 2008, 40: 23-25. 10.1038/ng.2007.48.PubMed CentralPubMedView ArticleGoogle Scholar
- Kraft HG, Lingenhel A, Kochl S, Hoppichler F, Kronenberg F, Abe A, Muhlberger V, Schonitzer D, Utermann G: Apolipoprotein(a) kringle IV repeat number predicts risk for coronary heart disease. Arterioscler Thromb Vasc Biol. 1996, 16: 713-719. 10.1161/01.ATV.16.6.713.PubMedView ArticleGoogle Scholar
- McCarroll SA, Huett A, Kuballa P, Chilewski SD, Landry A, Goyette P, Zody MC, Hall JL, Brant SR, Cho JH, Duerr RH, Silverberg MS, Taylor KD, Rioux JD, Altshuler D, Daly MJ, Xavier RJ: Deletion polymorphism upstream of IRGM associated with altered IRGM expression and Crohn’s disease. Nat Genet. 2008, 40: 1107-1112. 10.1038/ng.215.PubMed CentralPubMedView ArticleGoogle Scholar
- Hehir-Kwa JY, Pfundt R, Veltman JA, de Leeuw N: Pathogenic or Not? assessing the clinical relevance of copy number variants. Clin Genet. 2013, 84: 415-421. 10.1111/cge.12242.PubMedView ArticleGoogle Scholar
- Chen C, Qiao R, Wei R, Guo Y, Ai H, Ma J, Ren J, Huang L: A comprehensive survey of copy number variation in 18 diverse pig populations and identification of candidate copy number variable genes associated with complex traits. BMC Genomics. 2012, 13: 733-10.1186/1471-2164-13-733.PubMed CentralPubMedView ArticleGoogle Scholar
- Cutler G, Marshall LA, Chin N, Baribault H, Kassner PD: Significant gene content variation characterizes the genomes of inbred mouse strains. Genome Res. 2007, 17: 1743-1754. 10.1101/gr.6754607.PubMed CentralPubMedView ArticleGoogle Scholar
- Doan R, Cohen N, Harrington J, Veazy K, Juras R, Cothran G, McCue ME, Skow L, Dindot SV: Identification of copy number variants in horses. Genome Res. 2012, 22: 899-907. 10.1101/gr.128991.111.PubMed CentralPubMedView ArticleGoogle Scholar
- Fadista J, Nygaard M, Holm LE, Thomsen B, Bendixen C: A snapshot of CNVs in the pig genome. PLoS One. 2008, 3: e3916-10.1371/journal.pone.0003916.PubMed CentralPubMedView ArticleGoogle Scholar
- Fontanesi L, Martelli PL, Beretti F, Riggio V, Dall’Olio S, Colombo M, Casadio R, Russo V, Portolano B: An initial comparative map of copy number variations in the goat (Capra hircus) genome. BMC Genomics. 2010, 11: 639-10.1186/1471-2164-11-639.PubMed CentralPubMedView ArticleGoogle Scholar
- Graubert TA, Cahan P, Edwin D, Selzer RR, Richmond TA, Eis PS, Shannon WD, Li X, McLeod HL, Cheverud JM, Ley TJ: A high-resolution map of segmental DNA copy number variation in the mouse genome. PLoS Genet. 2007, 3: e3-10.1371/journal.pgen.0030003.PubMed CentralPubMedView ArticleGoogle Scholar
- Li J, Jiang T, Mao JH, Balmain A, Peterson L, Harris C, Rao PH, Havlak P, Gibbs R, Cai WW: Genomic segmental polymorphisms in inbred mouse strains. Nat Genet. 2004, 36: 952-954. 10.1038/ng1417.PubMedView ArticleGoogle Scholar
- Li Y, Mei S, Zhang X, Peng X, Liu G, Tao H, Wu H, Jiang S, Xiong Y, Li F: Identification of genome-wide copy number variations among diverse pig breeds by array CGH. BMC Genomics. 2012, 13: 725-10.1186/1471-2164-13-725.PubMed CentralPubMedView ArticleGoogle Scholar
- Liu GE, Hou Y, Zhu B, Cardone MF, Jiang L, Cellamare A, Mitra A, Alexander LJ, Coutinho LL, Dell’Aquila ME, Gasbarre LC, Lacalandra G, Li RW, Matukumalli LK, Nonneman D, Regitano LC, Smith TP, Song J, Sonstegard TS, Van Tassell CP, Ventura M, Eichler EE, McDaneld TG, Keele JW: Analysis of copy number variations among diverse cattle breeds. Genome Res. 2010, 20: 693-703. 10.1101/gr.105403.110.PubMed CentralPubMedView ArticleGoogle Scholar
- Ramayo-Caldas Y, Castello A, Pena RN, Alves E, Mercade A, Souza CA, Fernandez AI, Perez-Enciso M, Folch JM: Copy number variation in the porcine genome inferred from a 60 k SNP BeadChip. BMC Genomics. 2010, 11: 593-10.1186/1471-2164-11-593.PubMed CentralPubMedView ArticleGoogle Scholar
- Snijders AM, Nowak NJ, Huey B, Fridlyand J, Law S, Conroy J, Tokuyasu T, Demir K, Chiu R, Mao JH, Jain AN, Jones SJ, Balmain A, Pinkel D, Albertson DG: Mapping segmental and sequence variations among laboratory mice using BAC array CGH. Genome Res. 2005, 15: 302-311. 10.1101/gr.2902505.PubMed CentralPubMedView ArticleGoogle Scholar
- Wang J, Jiang J, Fu W, Jiang L, Ding X, Liu JF, Zhang Q: A genome-wide detection of copy number variations using SNP genotyping arrays in swine. BMC Genomics. 2012, 13: 273-10.1186/1471-2164-13-273.PubMed CentralPubMedView ArticleGoogle Scholar
- Colella S, Yau C, Taylor JM, Mirza G, Butler H, Clouston P, Bassett AS, Seller A, Holmes CC, Ragoussis J: QuantiSNP: an Objective Bayes Hidden-Markov Model to detect and accurately map copy number variation using SNP genotyping data. Nucleic Acids Res. 2007, 35: 2013-2025. 10.1093/nar/gkm076.PubMed CentralPubMedView ArticleGoogle Scholar
- Wang K, Li M, Hadley D, Liu R, Glessner J, Grant SF, Hakonarson H, Bucan M: PennCNV: an integrated hidden Markov model designed for high-resolution copy number variation detection in whole-genome SNP genotyping data. Genome Res. 2007, 17: 1665-1674. 10.1101/gr.6861907.PubMed CentralPubMedView ArticleGoogle Scholar
- Winchester L, Yau C, Ragoussis J: Comparing CNV detection methods for SNP arrays. Brief Funct Genomic Proteomic. 2009, 8: 353-366. 10.1093/bfgp/elp017.PubMedView ArticleGoogle Scholar
- Andersson L: Molecular consequences of animal breeding. Curr Opin Genet Dev. 2013, 23: 295-301. 10.1016/j.gde.2013.02.014.PubMedView ArticleGoogle Scholar
- Dorshorst B, Molin AM, Rubin CJ, Johansson AM, Stromstedt L, Pham MH, Chen CF, Hallbook F, Ashwell C, Andersson L: A complex genomic rearrangement involving the endothelin 3 locus causes dermal hyperpigmentation in the chicken. PLoS Genet. 2011, 7: e1002412-10.1371/journal.pgen.1002412.PubMed CentralPubMedView ArticleGoogle Scholar
- Salmon Hillbertz NH, Isaksson M, Karlsson EK, Hellmen E, Pielberg GR, Savolainen P, Wade CM, von Euler H, Gustafson U, Hedhammar A, Nilsson M, Lindblad-Toh K, Andersson L, Andersson G: Duplication of FGF3, FGF4, FGF19 and ORAOV1 causes hair ridge and predisposition to dermoid sinus in Ridgeback dogs. Nat Genet. 2007, 39: 1318-1320. 10.1038/ng.2007.4.PubMedView ArticleGoogle Scholar
- Olsson M, Meadows JR, Truve K, Rosengren Pielberg G, Puppo F, Mauceli E, Quilez J, Tonomura N, Zanna G, Docampo MJ, Bassols A, Avery AC, Karlsson EK, Thomas A, Kastner DL, Bongcam-Rudloff E, Webster MT, Sanchez A, Hedhammar A, Remmers EF, Andersson L, Ferrer L, Tintle L, Lindblad-Toh K: A novel unstable duplication upstream of HAS2 predisposes to a breed-defining skin phenotype and a periodic fever syndrome in Chinese Shar-Pei dogs. PLoS Genet. 2011, 7: e1001332-10.1371/journal.pgen.1001332.PubMed CentralPubMedView ArticleGoogle Scholar
- Axelsson E, Ratnakumar A, Arendt ML, Maqbool K, Webster MT, Perloski M, Liberg O, Arnemo JM, Hedhammar A, Lindblad-Toh K: The genomic signature of dog domestication reveals adaptation to a starch-rich diet. Nature. 2013, 495: 360-364. 10.1038/nature11837.PubMedView ArticleGoogle Scholar
- Berglund J, Nevalainen EM, Molin AM, Perloski M, Andre C, Zody MC, Sharpe T, Hitte C, Lindblad-Toh K, Lohi H, Webster MT: Novel origins of copy number variation in the dog genome. Genome Biol. 2012, 13: R73-10.1186/gb-2012-13-8-r73.PubMed CentralPubMedView ArticleGoogle Scholar
- Chen WK, Swartz JD, Rush LJ, Alvarez CE: Mapping DNA structural variation in dogs. Genome Res. 2009, 19: 500-509.PubMed CentralPubMedView ArticleGoogle Scholar
- Nicholas TJ, Baker C, Eichler EE, Akey JM: A high-resolution integrated map of copy number polymorphisms within and between breeds of the modern domesticated dog. BMC Genomics. 2011, 12: 414-10.1186/1471-2164-12-414.PubMed CentralPubMedView ArticleGoogle Scholar
- Nicholas TJ, Cheng Z, Ventura M, Mealey K, Eichler EE, Akey JM: The genomic architecture of segmental duplications and associated copy number variants in dogs. Genome Res. 2009, 19: 491-499.PubMed CentralPubMedView ArticleGoogle Scholar
- Vaysse A, Ratnakumar A, Derrien T, Axelsson E, Rosengren Pielberg G, Sigurdsson S, Fall T, Seppala EH, Hansen MS, Lawley CT, Karlsson EK, Bannasch D, Vila C, Lohi H, Galibert F, Fredholm M, Haggstrom J, Hedhammar A, Andre C, Lindblad-Toh K, Hitte C, Webster MT: Identification of genomic regions associated with phenotypic variation between dog breeds using selection mapping. PLoS Genet. 2011, 7: e1002316-10.1371/journal.pgen.1002316.PubMed CentralPubMedView ArticleGoogle Scholar
- Illumina CanineHD. [http://res.illumina.com/documents/products/datasheets/datasheet_caninehd.pdf],
- QuantiSNP. [https://sites.google.com/site/quantisnp/howto],
- PennCNV. [http://www.openbioinformatics.org/penncnv/],
- Pinto D, Darvishi K, Shi X, Rajan D, Rigler D, Fitzgerald T, Lionel AC, Thiruvahindrapuram B, Macdonald JR, Mills R, Prasad A, Noonan K, Gribble S, Prigmore E, Donahoe PK, Smith RS, Park JH, Hurles ME, Carter NP, Lee C, Scherer SW, Feuk L: Comprehensive assessment of array-based platforms and calling algorithms for detection of copy number variants. Nat Biotechnol. 2011, 29: 512-520. 10.1038/nbt.1852.PubMed CentralPubMedView ArticleGoogle Scholar
- Ensembl biomart. [http://may2012.archive.ensembl.org/biomart/martview/4fbfd6936ed52a8cf483dfa44f1b9a9c],
- g:Profiler - functional profiling of gene lists from large-scale experiments. [http://biit.cs.ut.ee/gprofiler/],
- Reimand J, Arak T, Vilo J: g:Profiler–a web server for functional interpretation of gene lists (2011 update). Nucleic Acids Res. 2011, 39: W307-W315. 10.1093/nar/gkr378.PubMed CentralPubMedView ArticleGoogle Scholar
- Choi YH, Park S, Hockman S, Zmuda-Trzebiatowska E, Svennelid F, Haluzik M, Gavrilova O, Ahmad F, Pepin L, Napolitano M, Taira M, Sundler F, Stenson Holst L, Degerman E, Manganiello VC: Alterations in regulation of energy homeostasis in cyclic nucleotide phosphodiesterase 3B-null mice. J Clin Invest. 2006, 116: 3240-3251. 10.1172/JCI24867.PubMed CentralPubMedView ArticleGoogle Scholar
- Doghman M, Arhatte M, Thibout H, Rodrigues G, De Moura J, Grosso S, West AN, Laurent M, Mas JC, Bongain A, Zambetti GP, Figueiredo BC, Auberger P, Martinerie C, Lalli E: Nephroblastoma overexpressed/cysteine-rich protein 61/connective tissue growth factor/nephroblastoma overexpressed gene-3 (NOV/CCN3), a selective adrenocortical cell proapoptotic factor, is down-regulated in childhood adrenocortical tumors. J Clin Endocrinol Metab. 2007, 92: 3253-3260. 10.1210/jc.2007-0342.PubMedView ArticleGoogle Scholar
- Johansson AC, Feuk L: Characterization of copy number-stable regions in the human genome. Hum Mutat. 2011, 32: 947-955. 10.1002/humu.21524.PubMedView ArticleGoogle Scholar
- Mechtcheriakova D, Wlachos A, Sobanov J, Kopp T, Reuschel R, Bornancin F, Cai R, Zemann B, Urtz N, Stingl G, Zlabinger G, Woisetschlager M, Baumruker T, Billich A: Sphingosine 1-phosphate phosphatase 2 is induced during inflammatory responses. Cell Signal. 2007, 19: 748-760. 10.1016/j.cellsig.2006.09.004.PubMedView ArticleGoogle Scholar
- Peiffer DA, Le JM, Steemers FJ, Chang W, Jenniges T, Garcia F, Haden K, Li J, Shaw CA, Belmont J, Cheung SW, Shen RM, Barker DL, Gunderson KL: High-resolution genomic profiling of chromosomal aberrations using Infinium whole-genome genotyping. Genome Res. 2006, 16: 1136-1148. 10.1101/gr.5402306.PubMed CentralPubMedView ArticleGoogle Scholar
- Diskin SJ, Li M, Hou C, Yang S, Glessner J, Hakonarson H, Bucan M, Maris JM, Wang K: Adjustment of genomic waves in signal intensities from whole-genome SNP genotyping platforms. Nucleic Acids Res. 2008, 36: e126-10.1093/nar/gkn556.PubMed CentralPubMedView ArticleGoogle Scholar
- CNV_BMCGenomics_2014. [http://www.broadinstitute.org/ftp/pub/vgb/dog/CNV_BMCGenomics_2014/],
- Galaxy. [https://main.g2.bx.psu.edu/],
- Livak KJ, Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001, 25: 402-408. 10.1006/meth.2001.1262.PubMedView ArticleGoogle Scholar
- Edgar R, Domrachev M, Lash AE: Gene expression omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res. 2002, 30: 207-210. 10.1093/nar/30.1.207.PubMed CentralPubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.