Over the last couple of years, whole-exome sequencing (WES) has emerged as a powerful research tool with which to identify novel germline mutations underlying a variety of Mendelian disorders with hitherto unknown etiologies Citation[1–3]. WES is a sequencing strategy that focuses on the protein coding regions of the human genome and is performed by high-throughput sequencing after exome enrichment steps. The introduction of multiple commercial exome enrichment kits, coupled with next-generation sequencing (NGS) technologies, has made this approach technically feasible, albeit with some limitations Citation[4]. In parallel, the application of WES in a molecular diagnostic context has also been increasingly evident Citation[5–8]. The transformative power of WES in both research discovery and diagnostic settings is obvious. This notwithstanding, it is widely believed that WES will soon become obsolete and will be replaced by whole-genome sequencing (WGS). In this editorial, we discuss whether or not WES is likely to be a transient technology in the context of molecular diagnostics for Mendelian disorders.
WES as a molecular diagnostic tool
One of the most immediate applications of WES is to facilitate accurate molecular diagnosis for patients with a clinically diagnosed/suspected Mendelian disorder whose etiology has not yet been narrowed down to a specific gene. In this context, WES has amply demonstrated its advantages over the traditional PCR–Sanger sequencing approach to: correct an initially incorrect clinical diagnosis for congenital chloride-losing diarrhea Citation[5]; unravel the genetic bases of patients’ conditions with broad and previously unsuspected genetic/phenotypic heterogeneity Citation[6]; and untangle the genetic basis of two distinct clinical phenotypes manifested by the same patient Citation[7]. Taken together, the power of WES lies in its ability to sequence beyond the known causal genes already potentially implicated in the etiology of the disease in question. The impact of WES on clinical management is also evidenced by the case of the patient, diagnosed with an intractable inflammatory bowel disease through the identification of a missense mutation in the XIAP gene, in whom allogeneic hematopoietic progenitor cell transplantation was subsequently performed Citation[8].
WES also represents an efficient screening and diagnostic tool for genetically heterogeneous conditions, such as Charcot–Marie–Tooth disease Citation[9] and the neuroacanthocytosis disorders Citation[10]. However, others have deployed a targeted NGS approach that harnesses the technological advances inherent in PCR-based enrichment (e.g., Fluidigm and RainDance technologies) and custom-designed hybrid selection (Agilent and Nimblegen platforms) to capture all the known causal genes for a particular group of diseases, such as the congenital disorders of glycosylation Citation[11] and the inherited cardiomyopathies Citation[12], followed by massively parallel sequencing. These disorders are not amenable to analysis by traditional PCR–Sanger methodology because they could be due to mutation in any one of a large number of different genes (e.g., there are >30 genes for Charcot–Marie–Tooth disease and for the inherited cardiomyopathies) and predicting the most likely causal gene is fairly challenging. Although WES and targeted NGS approaches are promising in molecular diagnostics, WES has a logistical advantage in diagnostic laboratories since it obviates the need to develop a diagnostic test specific to each disorder. WES has the advantage of being less biased, unlike targeted sequencing, which presupposes that the abnormalities of clinical relevance are confined within the panel of genes to be tested. These NGS approaches should be validated in a heavily regulated clinical setting if either platform is to be used to make a therapeutic recommendation.
The potential of WES as a diagnostic tool is becoming increasingly evident. However, several technical challenges remain and WES therefore requires further optimization and refinement Citation[4,13]. In addition to its role in confirming a clinical diagnosis, WES may also be used as a genetic screening test to determine the carrier status of given individuals with respect to disease variants underlying an autosomal recessive disorder Citation[14]. In addition to these technical considerations, there are complex ethical issues surrounding the disclosure of any findings that might be considered incidental or unrelated to the original purpose of diagnosis. A further complication is that the classification of detected variants into different categories of clinical importance is constantly evolving and will change continuously as new discoveries are made Citation[15].
Likely future coexistence of multiple NGS strategies in molecular diagnostics
The diagnostic application of WES has been amply demonstrated by numerous examples. However, will it endure as a diagnostic tool for many years to come, or will it be supplanted by WGS in the relatively near future? We believe that WES is likely, at least for the time being, to coexist with WGS even though the total cost of WGS will soon become much more affordable and, technically and analytically, more feasible. Nevertheless, the gap between WES and WGS, in terms of the relative technical challenges they represent, is unlikely to diminish in size any time soon; WES is always going to be the less challenging approach. The three NGS-based strategies (i.e., targeted, WES and WGS) are essentially mutually exclusive and therefore one has to select the best tool for the particular application under consideration, based upon available resources. Obviously, one approach does not fit all and all three NGS-based strategies may be expected to coexist in molecular diagnostics.
Further advances in sequencing technologies and bioinformatic tools, and the development of the means to interpret the plethora of variant data that will be generated, should allow WGS to be applied in molecular diagnostics in the not too distant future. However, we do not foresee that WGS will render the two other NGS-based strategies completely redundant or obsolete. Indeed, the relative suitability of each approach is likely to be context-dependent. For example, since most cases of familial adenomatous polyposis coli and Lynch syndrome are caused by mutations in the APC and DNA mismatch repair genes, respectively, a targeted NGS approach to the analysis of these genes should be quite sufficient for diagnostic purposes Citation[16]. Similarly, the diagnosis of X-linked mental retardation and other X-linked disorders would require an NGS approach targeted to genes on the X chromosome Citation[17]. By contrast, for diseases characterized by marked genetic heterogeneity, a targeted NGS approach or WES could be applied Citation[9–12]. In addition to single nucleotide variants and small indels, both NGS strategies are able to accurately identify larger copy number variants (CNVs), based on the depth of coverage approach of the sequence reads, which could then be further validated by multiplex ligation-dependent probe amplification Citation[18,19]. More recently, WES has been found to be capable of detecting structural variation and indels from 1 bp to 1 Mb using a new algorithm based on a ‘split-read’ approach Citation[20]. Taken together, NGS-based strategies represent a very promising approach to molecular diagnostics; however, WES is preferable for logistical reasons because it is a ‘common’ tool for all diseases caused by coding region variants. Although targeted NGS of the ‘Mendelianome’ (i.e., the coding regions for all genes implicated in Mendelian disorders) has been proposed as an alternative to WES, the gene target list of any Mendelianome analytical test would need to be updated periodically as new discoveries are made Citation[21].
On the other hand, there are diseases (e.g., mental retardation and congenital malformations) for which the underlying genetic causes are as likely to be due to CNVs/structural rearrangements, and where a WGS approach might be the preferred option. However, deep sequencing of the entire genome to 30–50× is not required; rather, sequencing of paired-end or large spanning mate-pair libraries with adequate physical coverage would be quite sufficient to identify CNVs/structural rearrangements. Indeed, CNVs are routinely detected using a variety of microarray platforms targeting the entire human genome Citation[22], which could be replaced by paired-end or mate-pair sequencing approaches that would also detect balanced chromosomal rearrangements. Recently, WGS has been used to diagnose a cryptic fusion oncogene in acute promyelocytic leukemia Citation[23]. Furthermore, WGS would be the only realistic option in a situation where WES had failed to identify the causative genetic defect having screened the exomes of different patients multiple times; this would suggest a possible extragenic location for the pathological lesions for example, mutations deep within introns or in remote regulatory elements Citation[24], and imply that WGS would be the screening technique of choice.
Our discussion has so far assumed that the diagnostic application of these NGS strategies would be confined to known disease genes. Hence, any ‘additional mutation data’ generated by WES (and WGS) is likely to be of relatively minor importance. However, if the additional data were to be communicated to patients or individuals undergoing the diagnostic/genetic screening testing, other issues would arise.
From diagnostic & genetic screening to personalized medicine
There is still a lack of any consensus view or general guidelines with respect to the disclosure of incidental findings to patients undergoing a WES- or NGS-based diagnostic test for those diseases where effective clinical intervention or management is available to prevent, delay the onset of, or treat the disease. If the incidental finding were not to be disclosed to the patients, then WES would remain a ‘convenient’ diagnostic tool as compared with the targeted NGS approach, and the sequencing data derived from genes other than the disease gene(s) underlying the disorder in question should not be subjected to further analysis. In this scenario, the outcome of the WES and targeted NGS approaches would be essentially similar. By contrast, if it were deemed appropriate for the incidental findings to be transmitted to the patients (or individuals who have undergone a genetic screening test) after proper consultation by medical geneticists, then this would potentiate reanalysis of the sequencing data with a view to retrospectively identifying new pathological mutations in the future as they came to be recognized. Indeed, variants of unknown significance may well acquire importance in a pathological context as new discoveries are made. If this reanalysis were to be deemed desirable (or even necessary) for reasons of good clinical practice, many questions would then arise, such as the timing of any follow-up analysis and the relative costs and benefits of recontacting and reconsulting subjects with respect to any new findings. Data reanalysis might also become possible with the development of more powerful analytical tools (such as alignment, assembly and variant calling tools), which could further optimize the sensitivity and specificity to detect variants missed by earlier analyses.
Although our discussion of incidental findings has so far referred exclusively to causal mutations for Mendelian or familial disorders, a further complication for WES/WGS is the generation of a substantial set of variants that could influence susceptibility to various other complex diseases. It will be some time before this ‘personalized genomic medicine information’ will be adequately defined in terms of its extent and likely clinical impact Citation[25]. In the meantime, a growing list of disease-associated and pharmacogenetic-relevant variants is being generated by genome-wide association studies; through deep sequencing of genome-wide association studies identified loci for rare variants, and in-depth postgenome-wide association studies functional characterization of specific variants (A Catalog of Published Genome-Wide Association Studies). The generation of such genomic information will be essential if personalized medicine is to realize its promise in the context of calibrating risk and assessing therapeutic options. Furthermore, in parallel to the goal of personalized medicine, this personalized genomic information should be made available to the subjects who have undergone the NGS diagnostic/screening test. However, it is noteworthy that, in contrast to the high penetrance mutations underlying most Mendelian disorders, the majority of variants known to be associated with complex disease or pharmacogenetic traits are not clinically actionable in terms of intervention or management. In addition, most of these variants have small effect sizes individually (odds ratios <1.5) and, for this reason, personalized risk assessment has been particularly challenging Citation[26,27].
Although WGS is preferable to WES in the context of personalized genomic medicine because it identifies all of the variants with potential clinical significance in the genome (and not just those variants residing within the coding regions), the scale of the task of data interpretation (and subsequent storage) and the cost of the whole process are currently (and in the foreseeable future) quite substantial. In the future, periodic reanalysis of WGS data once generated, in the context of new personalized genomic medicine information becoming available, is almost certainly an inevitable progression.
Conclusion
In our view, all available NGS approaches are likely to be used in different molecular diagnostic settings, and no particular technique will become predominant in the near future. Currently, there is still a complete lack of any consensus as to whether or not incidental findings generated by WES or WGS in a diagnostic setting should be disclosed to the patients tested. Given this uncertainty, our discussion of the diagnostic application of these NGS strategies assumes two scenarios. In the first, where the data analysis and interpretation are focused on variants in known disease genes, a targeted NGS approach would be the method of choice. However, WES might also be selected for logistical reasons, as it would serve as a diagnostic test for all diseases caused by coding region variants.
By contrast, if incidental findings and additional personalized genomic medical information were deemed to be useful and/or important to the subjects being tested, then WGS might be preferable. This then would be a one-time genetic test that would provide the basis for life-long follow-up. This would inevitably be costly and would require professionals from multiple disciplines to analyze and interpret the data and communicate their findings, potentially on an ongoing basis. The challenges involved in deploying WGS would not only be financial, technical and analytical but would also be interpretational and ethical. Therefore, in either scenario, we believe that WES is most unlikely to be merely a transient technology that will soon be abandoned. However, WGS will become more relevant as we move closer to the goal of personalizeds medicine. This notwithstanding, we are still very far from being able to interpret an entire genome in clinical terms. Although the goal of the US$1000 WGS is fast approaching, the costs of analysis, data storage and life-long follow-up for personalized medicine will not be trivial. Taken together, we anticipate that WES is likely to remain attractive and will be the preferred technique for certain diagnostic applications for some time to come.
Acknowledgements
The authors would like to thank Brendan Pang, National University Health System, Singapore and Richie Soong Cancer Science Institute of Singapore, Singapore.
Financial & competing interests disclosure
The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.
No writing assistance was utilized in the production of this manuscript.
References
- Ng SB, Nickerson DA, Bamshad MJ, Shendure J. Massively parallel sequencing and rare disease. Hum. Mol. Genet.19(R2), R119–R124 (2010).
- Gilissen C, Hoischen A, Brunner HG, Veltman JA. Unlocking Mendelian disease using exome sequencing. Genome Biol.12(9), 228 (2011).
- Bamshad MJ, Ng SB, Bigham AW et al. Exome sequencing as a tool for Mendelian disease gene discovery. Nat. Rev. Genet.12(11), 745–755 (2011).
- Clark MJ, Chen R, Lam HY et al. Performance comparison of exome DNA sequencing technologies. Nat. Biotechnol.29(10), 908–914 (2011).
- Choi M, Scholl UI, Ji W et al. Genetic diagnosis by whole exome capture and massively parallel DNA sequencing. Proc. Natl Acad. Sci. USA106(45), 19096–19101 (2009).
- Majewski J, Wang Z, Lopez I et al. A new ocular phenotype associated with an unexpected but known systemic disorder and mutation: novel use of genomic diagnostics and exome sequencing. J. Med. Genet.48(9), 593–596 (2011).
- Cullinane AR, Vilboux T, O’Brien K et al. Homozygosity mapping and whole-exome sequencing to detect SLC45A2 and G6PC3 mutations in a single patient with oculocutaneous albinism and neutropenia. J. Invest. Dermatol.131(10), 2017–2025 (2011).
- Worthey EA, Mayer AN, Syverson GD et al. Making a definitive diagnosis: successful clinical application of whole exome sequencing in a child with intractable inflammatory bowel disease. Genet. Med.13(3), 255–262 (2011).
- Montenegro G, Powell E, Huang J et al. Exome sequencing allows for rapid gene identification in a Charcot–Marie–Tooth family. Ann. Neurol.69(3), 464–470 (2011).
- Walker RH, Schulz VP, Tikhonova IR et al. Genetic diagnosis of neuroacanthocytosis disorders using exome sequencing. Mov. Disord. doi:10.1002/mds.24020 (2011) (Epub ahead of print).
- Jones MA, Bhide S, Chin E et al. Targeted polymerase chain reaction-based enrichment and next generation sequencing for diagnostic testing of congenital disorders of glycosylation. Genet. Med.13(11), 921–932 (2011).
- Meder B, Haas J, Keller A et al. Targeted next-generation sequencing for the molecular genetic diagnostics of cardiomyopathies. Circ. Cardiovasc. Genet.4(2), 110–122 (2011).
- Koboldt DC, Ding L, Mardis ER, Wilson RK. Challenges of sequencing human genomes. Brief Bioinform.11(5), 484–498 (2010).
- Kingsmore SF, Dinwiddie DL, Miller NA, Soden SE, Saunders CJ. Adopting orphans: comprehensive genetic testing of Mendelian diseases of childhood by next-generation sequencing. Expert Rev. Mol. Diagn.11(8), 855–868 (2011).
- Berg JS, Khoury MJ, Evans JP. Deploying whole genome sequencing in clinical practice and public health: meeting the challenge one bin at a time. Genet. Med.13(6), 499–504 (2011).
- Jasperson KW, Tuohy TM, Neklason DW, Burt RW. Hereditary and familial colon cancer. Gastroenterology138(6), 2044–2058 (2010).
- Tarpey PS, Smith R, Pleasance E et al. A systematic, large-scale resequencing screen of X-chromosome coding exons in mental retardation. Nat. Genet.41(5), 535–543 (2009).
- Walsh T, Lee MK, Casadei S et al. Detection of inherited mutations for breast and ovarian cancer using genomic capture and massively parallel sequencing. Proc. Natl Acad. Sci. USA107(28), 12629–12633 (2010).
- Lim BC, Lee S, Shin JY et al. Genetic diagnosis of Duchenne and Becker muscular dystrophy using next-generation sequencing technology: comprehensive mutational search in a single platform. J. Med. Genet.48(11), 731–736 (2011).
- Karakoc E, Alkan C, O’Roak BJ et al. Detection of structural variants and indels within exome data. Nat. Methods.9(2), 176–178 (2011).
- Kingsmore SF, Saunders CJ. Deep sequencing of patient genomes for disease diagnosis: when will it become routine? Sci. Transl. Med.3(87), 87ps23 (2011).
- Vissers LE, de Vries BB, Veltman JA. Genomic microarrays in mental retardation: from copy number variation to gene, from research to diagnosis. J. Med. Genet.47(5), 289–297 (2010).
- Welch JS, Westervelt P, Ding L et al. Use of whole-genome sequencing to diagnose a cryptic fusion oncogene. JAMA305(15), 1577–1584 (2011).
- Cooper DN, Chen JM, Ball EV et al. Genes, mutations, and human inherited disease at the dawn of the age of personalized genomics. Hum. Mutat.31(6), 631–655 (2010).
- Chan IS, Ginsburg GS. Personalized medicine: progress and promise. Annu. Rev. Genomics Hum. Genet.12, 217–244 (2011).
- Jostins L, Barrett JC. Genetic risk prediction in complex disease. Hum. Mol. Genet.20(R2), R182–R188 (2011).
- Ashley EA, Butte AJ, Wheeler MT et al. Clinical assessment incorporating a personal genome. Lancet375(9725), 1525–1535 (2010).