The Veterinary Journal 232 (2018) 16–19
Contents lists available at ScienceDirect
The Veterinary Journal journal homepage: www.elsevier.com/locate/tvjl
Evaluation of genes associated with human myxomatous mitral valve disease in dogs with familial myxomatous mitral valve degeneration K.M. Meursa,* , S.G. Friedenbergb , B. Williamsa , B.W. Keenea , C.E. Atkinsa , D. Adina , B. Aonaa , T. DeFrancescoa , S. Toua , T. Mackayc a b c
Department of Veterinary Clinical Sciences, North Carolina State University, Raleigh, NC 27607, USA Department of Veterinary Clinical Sciences, University of Minnesota, Saint Paul, MN 55108, USA Department of Biological Sciences, Genetics Program and Comparative Medicine Institute, North Carolina State University, Raleigh, NC 27607, USA
A R T I C L E I N F O
Article history: Accepted 2 December 2017 Keywords: Canine Cavalier King Charles spaniel Dachshund Genetic Mitral valve Myxomatous valve
A B S T R A C T
Myxomatous mitral valve disease (MMVD) is the most common heart disease in the dog. It is believed to be heritable in Cavalier King Charles spaniels (CKCS) and Dachshunds. Myxomatous mitral valve disease is a familial disease in human beings as well and genetic mutations have been associated with its development. We hypothesized that a genetic mutation associated with the development of the human form of MMVD was associated with the development of canine MMVD. DNA was isolated from blood samples from 10 CKCS and 10 Dachshunds diagnosed with MMVD, and whole genome sequences from each animal were obtained. Variant calling from whole genome sequencing data was performed using a standardized bioinformatics pipeline for all samples. After ﬁltering, the canine genes orthologous to the human genes known to be associated with MMVD were identiﬁed and variants were assessed for likely pathogenic implications. No variant was found in any of the genes evaluated that was present in least eight of 10 affected CKCS or Dachshunds. Although mitral valve disease in the CKCS and Dachshund is a familial disease, we did not identify genetic cause in the genes responsible for the human disease in the dogs studied here. © 2017 Elsevier Ltd. All rights reserved.
Introduction Myxomatous mitral valve degeneration (MMVD) is the most common heart disease in the dog (Gordon et al., 2017). Affected dogs can live with subclinical disease for years. However, many eventually develop clinical signs consistent with congestive heart failure and are at risk of sudden cardiac death. No medical cure for the disease exists. Instead, medical treatment is directed primarily at hemodynamic management of the clinical signs that result from severe mitral regurgitation with subsequent left atrial enlargement or congestive heart failure and in some cases, pulmonary hypertension (Gordon et al., 2017). While surgical repair or replacement of the valve is possible, the cost and expertise required to successfully implement these treatment strategies severely limit their applicability to the general pet population. The lack of knowledge about the factors that contribute to the development of this common and important canine disease has
* Corresponding author. E-mail address: [email protected]
(K.M. Meurs). https://doi.org/10.1016/j.tvjl.2017.12.002 1090-0233/© 2017 Elsevier Ltd. All rights reserved.
prevented the development of effective long-term clinical management plans. Myxomatous degeneration of the mitral valve is believed to be heritable in at least some breeds, including the Cavalier King Charles spaniel (CKCS) and the Dachshund (Swenson et al., 1996; Olsen et al., 1999; Lewis et al., 2011). However, the mode of inheritance has not been determined and genetic mutation(s) responsible for the disease have not yet been identiﬁed. In the Whippet with mitral valve disease, the severity of the disease has been linked to a region on chromosome 15 (Stern et al., 2015). Myxomatous mitral valve disease is a familial disease in human beings that exists either as part of a larger multisystem syndrome, or as a singular disease, and genetic mutations have been associated with its development (Padang et al., 2012; LaHaye et al., 2014). Several of the genes associated with the development of MMVD are associated with the tumor growth factor-b (Tgf-b) superfamily. Mitral valve abnormalities have been associated with mutations in the Fibrillin 1 (FBN1) gene and Fibrillin 1 contributes to the regulation of transforming growth factor beta (Tgf-b) by targeting and concentrating Tgf-b at speciﬁc locations. Fibrillin 1 deﬁciency in mice has been associated with enhanced Tgf-b
K.M. Meurs et al. / The Veterinary Journal 232 (2018) 16–19
signaling and development of a myxomatous and prolapsing mitral valve (Ng et al., 2004). Loeys-Dietz syndrome, characterized by cardiac changes including mitral valve prolapse is associated with mutations in Tgf-b receptors 1 and 2 (Loeys et al., 2005). A homozygous mutation was also identiﬁed in the LTBP3 gene, a gene that produces a protein that increases the bioavailability of Tgf-b in a family with mitral valve disease as well as short stature and oligodontia (Dugan et al., 2015). These ﬁndings have led to the development of a hypothesis that MMVD may be associated with a dysregulation of Tgf-b signaling (Hulin et al., 2013). Familial MMVD in human beings is also observed as an isolated disease, associated with variants in the urokinase-plasminogen activator (PLAU) gene, Dachsous Cadherin-Related 1 (DCHS1) gene and several of the collagen genes (Chou et al., 2004a,b; Lardeux et al., 2011; Padang et al., 2012). PLAU encodes for a protein that has been suggested to be a trigger to activate the matrix metalloproteinase pathway (Chou et al., 2004a,b). DCHS1 encodes for a protein thought to be associated with atrioventricular canal development and cellular stability of the mitral valve (Durst et al., 2015). Finally, abnormalities of collagen have been identiﬁed in other syndromic forms of MMVD including Ehler-Danos, Stickler and Osteogenesis imperfecta (Padang et al., 2012). Similarities in the disease have been noted between human and canine MMVD (Pedersen and Haggstrom, 2000; Aupperle and Disatian, 2012). In both species, familial MMVD starts in adulthood and increases in prevalence with increasing age. The clinical course is generally one of fairly slow progression and mitral valve prolapse (MVP) is a common ﬁnding such that the disease is commonly referred to as MVP in human beings. Pathologic ﬁndings include an accumulation of glycosaminoglycans, and fragmentation of collagen bundles and elastic ﬁbers within the valve leaﬂets (Pedersen and Haggstrom, 2000; Aupperle and Disatian, 2012). Given the similarities between the human and canine forms of the disease, we hypothesized that a genetic mutation associated with the development of the human form of MMVD might be associated with the development of the canine form of the disease. The objective of this study was to evaluate genes known to be associated with human form of MMVD in two dog breeds in which the disease is known to be heritable, the CKCS and the Dachshund. Materials and methods This study was conducted in accordance with the guidelines of the North Carolina State University Institutional Animal Care and Use Committee (Approval No. IACUC 13-103-0; Approval date 4 November 2012). DNA sequencing DNA was isolated from blood samples from 10 Cavalier King Charles spaniels and 10 Dachshunds diagnosed with MMVD by the presence of a systolic left apical heart murmur and echocardiographic ﬁndings consistent with mitral valve degeneration (valve thickening, prolapse, mitral regurgitation), as identiﬁed by board-certiﬁed veterinary cardiologists (Borgarelli and Haggstrom, 2010). Approximately 3 mg of DNA from each animal was submitted for library preparation and whole genome sequencing.1 All sequencing experiments were designed as 150 bp paired-end reads and samples were run in one lane of an Illumina HiSeq 4000 high-throughput sequencing system. Variant calling and ﬁltering Variant calling from whole genome sequencing data was performed using a standardized bioinformatics pipeline for all samples as described previously (Friedenberg and Meurs, 2016). Brieﬂy, sequence reads were trimmed using Trimmomatic 0.32 to a minimum phred-scaled base quality score of 30 at the start and end of each read with a minimum read length of 70 bp (Bolger et al., 2014). Sequences were then aligned to the canFam3 reference sequence using BWA 0.7.13 (Li and Durbin, 2009; Lindblad-Toh et al., 2005). Aligned reads were prepared for
Table 1 Gene names and chromosomal coordinates. Gene name FBN1 LTBP3 PLAU DCHS1 FLNA LTBP2 TGFBR1 TGFBR2 COL1A1 COL1A2 COL2A1 COL3A1 COL5A1 COL5A2 COL5A3 COL11A1 COL11A2
30 18 4 21 X 8 11 23 9 14 27 36 9 36 20 6 12
14,638,844 51,653,800 24,327,934 29,932,714 122,061,570 47,659,413 56,192,030 13,886,405 26,183,852 19,883,733 6,756,994 30,487,822 50,741,552 30,537,639 51,037,378 47,425,307 2,626,822
End 14,864,146 51,668,012 24,334,907 29,949,159 122,0883,141 47,758,444 56,223,557 13,945,343 26,199,927 19,920,718 6,787,846 30,526,563 50,856,744 30,677,658 51,074,394 47,622,225 2,655,698
Variant evaluation Variants were then evaluated for likely pathogenic implications using the Standards and Guidelines for the Interpretation of Sequence Variants (Richards et al., 2015). First, identiﬁed variants were evaluated to determine any that encoded for a premature stop codon, frameshift, altered splice site, altered start codon, and/ or a single or multi-exon deletion. Next, any missense variants were evaluated using the in-silico programs Polyphen,3 Sift,4 and Provean5 to assess the predicted impact of any amino acid change. Variants that were predicted to be deleterious in at least two of the three programs were selected for further evaluation. Variants identiﬁed within noncoding regions (upstream, downstream, intronic) were evaluated for evidence of conservation with additional in-silico programs (GERP, PhyloP and Phastcons) as described (Nalpathamkalam et al., 2014; Pollard et al., 2010). Intronic variants were also evaluated with three prediction programs to determine if the
See: https://www.genewiz.com/en (Accessed 2 December 2017).
analysis using Picard Tools 2.52 and GATK 3.7 following best practices for base quality score recalibration and indel realignment as speciﬁed by the Broad Institute, Cambridge, MA (McKenna et al., 2010; DePristo et al., 2011; Van Der Auwera et al., 2014). Variant calls were made using GATK’s HaplotyeCaller walker, and variant quality score recalibration (VQSR) was performed using sites from dbSNP 146 and the Illumina 174K CanineHD BeadChip as training resources. A VQSR tranche sensitivity cutoff was applied to SNPs at 99.9% and 99% to indels for use in downstream analyses; genotype calls with a phred-scaled quality score < 20 were ﬂagged but not removed from the variant callset. Variants present in affected CKCS and Dachshunds were selected and ﬁltered against a database of variants derived from whole genome sequencing of 98 medium to large breed dogs. Selected variants had to be present in at least eight of 10 of the affected dogs from each breed to allow for possible areas of poor coverage. The whole genome sequencing database included 13 different dog breeds not known to have an increased prevalence of MMVD including: American staffordshire terrier, Boxer, Doberman pinscher, German shepherd, Goldendoodle, Golden retriever, Great dane, Labradoodle, Portuguese water dog, Rhodesian ridgeback, Rottweiler, Scottish deerhound and Standard poodle. Variants were then annotated using Variant Effect Predictor 89 (Mclaren et al. 2016). Two ﬁles of variants were developed. The ﬁrst ﬁle (no pipeline) contained variants found in at least eight of 10 dogs of each breed and not in any of the medium- and large-breed dogs. The second ﬁle (rare pipeline) contained variants found in at least eight of 10 dogs of each breed, but allowed variants present at an allele frequency of up to 5% within the population of medium- and large-breed dogs. After ﬁltering, the orthologous canine genes to the human genes known to be associated with mitral valve degeneration (either as a syndromic or single trait) were identiﬁed. The following genes were evaluated: Collagen Type I Alpha 1 (COL1A1), Collagen Type 1 Alpha 2 (COL1A2), Collagen Type II Alpha 1 (COL2A1), Collagen Type III Alpha 1 (COL3A1), Collagen Type V Alpha 1 (COL5A1), Collagen Type 5 Alpha 2 (COL5A2), Collagen Type 5 Alpha 3 (Col5A3), Collagen Type XI Alpha 1 (COL11A1), Collagen Type 11 Alpha (COL11A2), Dachsous Cadherin-Related 1 (DCHS1), Fibrillin 1 (FBN1), Filamin A (FLNA), Latent Transforming Growth Factor Beta Binding Protein 2 and 3 (LTBP2, LTBP3), Plasminogen Activator Urokinase (PLAU), Transforming Growth Factor Beta Receptor 1 and 2 (TGFBR1, TGFBR2) (Chou et al., 2004a,b; Padang et al., 2012; LaHaye et al., 2014; Dugan et al., 2015; Durst et al., 2015) (Table 1).
See: See: See: See:
http://broadinstitute.github.io/picard (Accessed 2 December 2017). http://genetics.bwh.harvard.edu/pph2/ (Accessed 2 December 2017). http://sift.jcvi.org/ (Accessed 2 December 2017). http://provean.jcvi.org/index.php (Accessed 2 December 2017).
K.M. Meurs et al. / The Veterinary Journal 232 (2018) 16–19
Table 2 Variants identiﬁed in nine of 10 Cavalier King Charles spaniels with myxomatous mitral valve disease. Gene COL5A1 COL5A1 COL5A1 FBN1
9:50759457 9:50762902 9:50771721 30:14652100–14652101
Missense Intronic Intronic Intronic
variant could create a cryptic slice site. The following programs were used: Human Splice Site Finder,6 NNSPLICE,7 and SPL.8 Splice site creation was interpreted positively if it was predicted in all three programs (Vreeswijk et al., 2009). Finally, the data was visually inspected for multiallelic variations including larger insertions and deletions using the GenomeBrowse 2.1.2 (Golden Helix9) visualization tool.
Predicted to alter splice site No No No
No variant present in at least eight of 10 affected CKCS and absent in our control population of medium- and large-breed dogs was found in any of the genes of interest. A single coding variant was found in the gene encoding for collagen type V, alpha 1 (COL5A1) in nine of 10 affected CKCS and in the rare pipeline. The variant was a missense mutation that was predicted to be benign (Table 2). Two intronic variants in COL5A1 and two in FBN1 were present in nine of 10 affected CKCS and in the rare pipeline (Table 2). None of these variants met our criteria for conservation or were predicted to create or change a splice site.
the disease (Padang et al., 2012; LaHaye et al., 2014). Similarly, it is very likely that there are multiple different genes responsible for the canine disease, both within and across breeds. In this study, we limited our evaluation to two breeds of a known familial etiology. The absence of the identiﬁcation of important variants in the genes of these two breeds does not mean that the genes we evaluated here will not be important in other breeds affected with MMVD. This study has several limitations. First, we only ﬁltered for variants that were present in at least eight of 10 dogs as our cutoff for ﬁltering evaluation. This was based on an assumption that in these pure breed dogs, there would more likely be a single mutation per breed and that all dogs within a breed will share a common mutation. However, this might not, in fact, turn out to be the case, given the genetic heterogeneity observed in humans. Additionally, we made an assumption that variants likely responsible for the disease would not exist, or would only exist to a minimal extent (less than 0.05 of the alleles) in our control pipeline population of medium and large breed dogs. This is a somewhat arbitrary cut off, based on what would generally be considered to be a rare variant. However, if some of the dogs in our control population had an allele in one of these genes at a higher frequency than this, we would have ﬁltered out the variant by our approach. This would be unlikely given the low prevalence of the disease within medium to large breed dogs but not impossible.
No variant was found in any of our candidate genes that was present in least eight of 10 Dachshunds and in the no pipeline or rare pipeline of our population of medium- and large-breed dogs.
Although mitral valve disease in the CKCS and Dachshund is a familial disease as it is in human beings, we did not identify a causative mutation in the genes known to be responsible for the human form of the disease in the dogs studied here. Further evaluation of the regulatory regions and genes associated with valve development, integrity and function is warranted to identify additional possible genetic candidates. The lack of knowledge about the factors that contribute to the development and progression of this extremely common and important canine disease has prevented the development of effective long-term clinical management or prevention plans. Advances in the understanding of the genetic factors responsible for MMVD may lead to both new medical treatment and/or prevention options.
Results Across the 14 genes and 20 samples we examined, median coverage was 38.4 (range 10.3–79.6); a median of 99.9% of bases were covered to a depth of at least 1 (range 99.7–99.9%) and 99.6% of bases to a depth of at least 5 (range 99.0–99.9%). Cavalier King Charles spaniels
Discussion The similarities between canine and human MMVD could suggest a similar genetic cause; therefore, evaluation of the genes associated with the human form of the disease were warranted in the dog with familial MMVD (Pedersen and Haggstrom, 2000; Aupperle and Disatian, 2012). Although the canine form of MMVD appears to be an isolated disease rather than a syndrome, we evaluated all of the well-documented genes associated with the human form of the disease. However, in the two breeds of dogs with familial MMVD evaluated in this study, neither breed had a likely causative mutation in a gene associated with the human disease (Swenson et al., 1996; Olsen et al., 1999; Lewis et al., 2011). It is still possible that human and canine forms of MMVD share disruption of a similar biological pathway, perhaps even in a Tgf-b pathway, that results in the development of this valvular disease. Mitral valve degenerative disease in human beings is a disease of genetic heterogeneity with multiple different genes leading to
Conﬂict of interest statement None of the authors has any ﬁnancial or personal relationships that could inappropriately inﬂuence or bias the content of the paper. Acknowledgement This study was funded by the Mark L. Morris Jr. Investigator Award, D16CA509, Morris Animal Foundation.
See: http://www.umd.be/HSF3/ (Accessed 2 December 2017). See: http://www.fruitﬂy.org/seq_tools/splice.html (Accessed 2 December 2017). 8 See: http://www.softberry.com/berry.phtml?topic=spl&group=programs&subgroup=gﬁnd (Accessed 2 December 2017). 9 See: http://goldenhelix.com/products/GenomeBrowse/index.html (Accessed 2 December 2017). 7
References Aupperle, H., Disatian, S., 2012. Pathology, protein expression and signaling in myxomatous mitral valve degeneration: comparison of dogs and humans. Journal of Veterinary Cardiology 14, 59–71.
K.M. Meurs et al. / The Veterinary Journal 232 (2018) 16–19 Bolger, A.M., Lohse, M., Usadel, B., 2014. Trimmomatic: a ﬂexible trimmer for Illumina sequence data. Bioinformatics 30, 2114–2120. Borgarelli, M., Haggstrom, J., 2010. Canine degenerative myxomatous mitral valve disease: natural history, clinical presentation and therapy. Veterinary Clinics of North America: Small Animal Practice 40, 651–663. Chou, H., Chen, Y., Wu, J., Tsai, F., 2004a. Association between urokinase-plasminogen activator gene T4065C polymorphism and risk of mitral valve prolapse. International Journal of Cardiology 96, 165–170. Chou, H.T., Hung, J.S., Chen, Y.T., Wu, J.Y., Tsai, F.J., 2004b. Association between COL3A1 collagen gene exon 31 polymorphism and risk of ﬂoppy mitral valve/mitral valve prolapse. International Journal of Cardiology 95, 299–305. DePristo, M.A., Banks, E., Poplin, R.E., Garimella, K.V., Maguire, J.R., Hartl, C., Phillappakis, A.A., Del Angel, G., Rivas, A.M., Hanna, M., et al., 2011. A framework for variation discovery and genotyping using next-generation DNA sequencing data. Nature Genetics 43, 491–498. Dugan, S.L., Temme, R.T., Olson, R.A., Mikhailov, A., Law, R., Mahmood, H., Noor, A., Vincent, J.B., 2015. New recessive truncating mutation in LTBP3 in a family with oligodontia, short stature, and mitral valve prolapse. American Journal of Medical Genetics Part A 167, 1396–1399. Durst, R., Sauls, K., Peal, D., deVlaming, A., Toomer, K., Leyne, M., Salani, M., Talkowski, M.E., Brand, H., Perrocheau, M., et al., 2015. Mutations in DCHS1 causes mitral valve prolapse. Nature 34, 355–368. Friedenberg, S.G., Meurs, K.M., 2016. Genotype imputation in the domestic dog. Mammalian Genome 27, 485–494. Gordon, S.G., Saunders, A.B., Wesselowski, S.R., 2017. Asymptomatic canine degenerative valve disease. Current and future therapies. Veterinary Clinics of North America — Small Animal Practice 47, 955–975. Hulin, A., Deroanne, C., Lambert, C., Defraigne, J.O., Nusgens, B., Radermecker, M., Colige, A., 2013. Emerging pathogenic mechanisms in human myxomatous mitral valve: lessons from past and novel data. Cardiovascular Pathology 22, 245–250. LaHaye, S., Lincoln, J., Garg, V., 2014. Genetics of valvular heart disease. Current Cardiology Reports 16, 487–490. Lardeux, A., Kyndt, F., Lecointe, S., Marec, H.L., Merot, J., Schott, J.J., Tourneau, T., Le Probst, V., 2011. Filamin-a-related myxomatous mitral valve dystrophy: genetic, echocardiographic and functional aspects. Journal of Cardiovascular Translational Research 4, 748–756. Lewis, T., Swift, S., Woolliams, J.A., Blott, S., 2011. Heritability of premature mitral valve disease in Cavalier King Charles Spaniels. Veterinary Journal 188, 73–76. Li, H., Durbin, R., 2009. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinformatics 25, 1754–1760. Lindblad-Toh, K., Wade, C.M., Mikkelsen, T.S., Karlsson, E.K., Jaffe, D.B., Kamal, M., Clamp, M., Chang, J.L., Kulbokas, E.J., Zody, M.C., et al., 2005. Genomic sequence, comparative analysis and haplotype structure of the domestic dog. Nature 438, 803–819.
Loeys, B.L., Chen, J., Neptune, E.R., Judge, D.P., Podowski, M., Holm, T., Meyers, J., Leitch, C., Katsanis, N., Shariﬁ, N., et al., 2005. A syndrome of altered cardiovascular, craniofacial, neurocognitive and skeletal development caused by mutations in TGFBR1 or TGFBR2. Nature Genetics 37, 275–281. McKenna, A., Hanna, M., Banks, E., Sivachenko, A., Cibulskis, K., Kerytsky, A., Garimella, K., Altshuler, D., Gabriel, S., Daly, M., et al., 2010. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Research 1297–1303. Mclaren, W., Gil, L., Hunt, S.E., Riat, H.S., Ritchie, G.R.S., Thormann, A., Flicek, P., Cunningham, F., 2016. The Ensembl Variant Effect Predictor. Genome Biology 17, 1–14. Nalpathamkalam, T., Derkach, A., Paterson, A.D., Merico, D., 2014. Genetic Analysis Workshop 18 single-nucleotide variant prioritization based on protein impact, sequence conservation, and gene annotation. BMC Proceedings 8, 11. Ng, C.M., Cheng, A., Myers, L.A., Martinez-Murillo, F., Jie, C., Bedja, D., Gabrielson, K. L., Hausladen, M.W.J., Mecham, R.P., Judge, D.P., et al., 2004. TGF-b-dependent pathogenesis of mitral valve prolapse in a mouse model of Marfan syndrome. The Journal of Clinical Investigation 114, 1586–1592. Olsen, L.H., Fredholm, M., Pedersen, H.D., 1999. Epidemiology and inheritance of mitral valve prolapse in Dachshunds. Journal of Veterinary Internal Medicine 13, 448–456. Padang, R., Bagnall, R.D., Semsarian, C., 2012. Genetic basis of familial valvular heart disease. Circulation: Cardiovascular Genetics 5, 569–580. Pedersen, H.D., Haggstrom, J., 2000. Mitral valve prolapse in the dog: a model of mitral valve prolapse in man. Cardiovascular Research 47, 234–243. Pollard, K.S., Hubisz, M.J., Rosenbloom, K.R., Siepel, A., 2010. Detection of nonneutral substitution rates on mammalian phylogenies. Genome Research 20, 110–121. Richards, S., Aziz, N., Bale, S., Bick, D., Das, S., Gastier-Foster, J., 2015. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genetics in Medicine 17, 405–424. Stern, J.A., Hsue, W., Song, K.H., Ontiveros, E.S., Fuentes, V.L., Stepien, R.L., 2015. Severity of mitral valve degeneration is associated with chromosome 15 loci in whippet dogs. PLoS One 10, 1–11. Swenson, L., Haggstrom, J., Kvart, C., Juneja, R., 1996. Relationship between parental cardiac status in Cavalier King Charles spaniels and prevalence and severity of chronic valvular disease in offspring. Journal of the American Veterinary Association 208, 2009–2012. Van Der Auwera, G., Carneiro, M.O., Hartl, C., Poplin, R., Levy-Moonshine, A., Jordan, T., Shakir, K., Roazen, D., Thibault, J., Banks, E., et al., 2014. From FastQ data to high conﬁdence varant calls: the Genonme Analysis Toolkit best practices pipeline. Current Protocols in Bioinformatics 11, 1–33. Vreeswijk, M.P.G., Kraan, J.N., Van Der Klift, H.M., Vink, G.R., Cornelisse, C.J., Wijnen, J.T., Bakker, E., Van Asperen, C.J., Devilee, P., 2009. Intronic variants in BRCA1 and BRCA2 that affect RNA splicing can be reliably selected by splice-site prediction programs. Human Mutation 30, 107–114.