Nature news: 未来40年，DNA测序将走向何方？

Nature news: 未来40年，DNA测序将走向何方？ 

2017-10-14 00:00

40年前,Sanger测序技术诞生,让DNA片段的测序成为现实.自此,DNA测序技术以惊人的速度发展,越过一座又一座的里程碑.那么,未来40年,DNA测序又将变成什么样?Eric Green、Edward Rubin和Maynard Olson三位科学家本周在《Nature》上发文,展望了这项技术的未来.

Coloured DNA bands.

测序的需求

技术的改进可能增加需求,也可能减少需求.作者认为,DNA测序将遵循计算和摄影的模式.随着测序变得更便宜、更方便,应用将会激增,需求将会上涨.当DNA测序突破科研市场,进入临床、消费者及其他领域,"供应越多意味着需求越多"的规则将愈加明显.

上世纪90年代,人类基因组测序的想法让人觉得不可思议.如今,遗传学家却希望对地球上的每个人、每种组织中的每个细胞进行测序.同时,考古学家也希望借助测序来了解祖先群体的基因流动.生态学家、进化生物学家也试图分析所有物种的基因组,甚至是整个生态系统.

当然,目前的瓶颈是分析和解释所有的DNA序列数据.作者预测,大量的DNA序列数据以及表型信息的结合,将让研究人员能够推断基因组序列所编码的生物学功能.更重要的是,解释数据所需的大部分基础知识已经准备好了.

杀手级应用

纵观其他技术,比如智能手机、互联网和数码摄影,真正的颠覆者都是应用,而不是新技术.作者确信,DNA测序将彻底改变的一个领域将是医学.

目前,DNA测序的突破性临床应用是产前检测.它通过检测在母体血液中循环的少量胎儿游离DNA,来检测染色体的数量异常.据估计,全世界每年大约有400-600万名孕妇在接受这一检测,十年内这个数字将超过1500万.从中也许能推断出未来应用的一些特征:非侵入性、易于开展、对核苷酸水平的准确性要求较低.

SOURCE: National Human genome research Institute

在肿瘤学方面,人们已经投入相当多的资金来开发液体活检.不难想象,基于序列的癌症检测将会成为常规的筛查工具,就像巴氏涂片和结肠镜检查.随着癌症治疗开始针对特定突变,而不是肿瘤类型,液体活检最终将指导治疗干预.

作者也设想了DNA测序在诊所之外的各种应用,特别是手持式DNA测序仪.流行病学家可以利用这种装置来检测空气、水、食品、动物和昆虫载体,当然还有人的咽部标本和体液.DNA测序技术的轻松获取可以促进"全球病毒组计划"这样的项目,以了解传播疾病的各种病毒.此外,这种仪器也可能成为刑侦上的工具.

最后,文章也提到了测序技术的绊脚石.DNA测序技术也许很快就能纳入常规的临床应用,以分析各种情况下获得的体液.不过,只有整合数百万人多年医疗史的数据,才能提供所需的元信息(meta-information),确定何时忽略这些数据,以及何时采取行动.这是一个挑战

原文标题：The future of DNA sequencing

Nature 550, 179–181 (12 October 2017) doi:10.1038/550179a

Forty years ago, two papers1, 2 described the first tractable methods for determining the order of the chemical bases in stretches of DNA. Before these 1977 publications, molecular biologists had been able to sequence only snippets.

The evolution of DNA sequencing from these nascent protocols to today's high-throughput technologies has occurred at a breathtaking pace3. Nearly 30 years of exponential growth in data generation have given way, in the past decade, to super-exponential growth. And the resultant data have spawned transformative applications in basic biology and beyond — from archaeology and criminal investigation to prenatal diagnostics.

What will the next 40 years bring?

Prognosticators are typically wrong about which technologies — or, more importantly, which applications — will be the most disruptive. In the early days of the Internet, few predicted that e-mail that would achieve staggering popularity. Similarly, traders on Wall Street and investors in Silicon Valley failed to foresee that games, online video streaming and social media would come to dominate the use of today's available processing power and network bandwidth.

We would probably fare no better in predicting the future of DNA sequencing. So instead, we offer a framework for thinking about it. Our central message is that trends in DNA sequencing will be driven by killer applications, not by killer technologies.

In demand

Improvements in a technology can either increase or decrease demand. Microsoft co-founder Bill Gates famously cited radial tyres as an example of the latter: because they were more durable than earlier designs, the need for tyres dropped and the tyre industry shrank.

We think that DNA sequencing will follow the pattern of computing and photography, not of tyres. As it becomes cheaper and more convenient, applications will proliferate, and demand will rise (see 'Better, cheaper, faster'). As DNA sequencing breaks out of the research market and into clinical, consumer and other domains, the rule of 'more supply means more demand' will hold ever more strongly.

Researchers have an insatiable appetite for DNA-sequence data. In the 1990s, the idea of sequencing a human genome seemed daunting. Now, geneticists would like to have DNA sequences for everyone on Earth, and from every cell in every tissue at every developmental stage (including epigenetic modifications), in health and in disease. They would also like to get comprehensive gene-expression patterns by sequencing the complementary DNA copies of messenger RNA molecules. Meanwhile, archaeologists are beginning to reconstruct the flow of genes through ancestral populations, just as they previously deduced the flow of languages, cultural practices and material objects. And taxonomists, ecologists, microbiologists and evolutionary biologists are seeking to analyse the genomes of all living (and extinct) species — and even whole ecosystems.

Obviously, a sustained demand for data would require that the vast cataloguing efforts proffer actual understanding. At present, the bottleneck is analysing and interpreting all the DNA-sequence data. But just as new informatics approaches and massive data sets have dramatically improved language translation and image recognition, we predict that massive DNA-sequence data sets coupled with phenotypic information will enable researchers to deduce the biological functions encoded within genome sequences.

What's more, much of the basic science needed to interpret the data is already in place for a growing repertoire of practical applications (such as high-quality reference sequences of bacterial genomes, or the rules by which certain gene networks operate in healthy people). These range from recognizing microbial DNA sequences in unbiased surveys of environmental or clinical samples to identifying genome changes associated with known biological consequences.

Killer applications

Over the years, the platforms for DNA sequencing have changed dramatically (see 'Many ways to sequence DNA'). Yet the trajectories of other technologies for which there is a seemingly insatiable demand — smartphones, the Internet, digital photography — suggest that the real disrupters will be the resulting applications, not the new technologies.

Many ways to sequence DNA

Over the past 40 years, the platforms for DNA sequencing have repeatedly been replaced.

By 1985, almost all DNA sequencing was performed with the Sanger or dideoxy chain-termination method2; reaction products were labelled with radionucleotides, separated on acrylamide slab gels, and detected with autoradiography (the use of X-ray or photographic film to detect radioactively labelled samples). By 2000, the four-colour-fluorescence method reigned supreme; reaction products were labelled with chain-terminating nucleotide analogues, separated electrophoretically in capillaries filled with a jelly-like media, and detected with energy-transfer fluorescent dyes. By 2010, the techniques had diversified. The dominant instruments were based on massively parallel analyses of DNA 'polonies' (clonal amplifications of a single DNA molecule) and on sequencing-by-synthesis chemistries (these rely on reversible chain-terminators).

From now on, the requirements for each DNA-sequencing platform will depend on what it is to be used for. In oncology and medical genetics, the goal will often be to identify every base correctly and to define every variant of genomic segments that exist in multiple copies. By contrast, when a yes or no 'match' is required — for instance, in species identification — the ability to run tests quickly and easily in the field may be more important than accuracy.

Another factor that will probably change is the relative need for centralized versus decentralized DNA sequencing. An epidemiologist trying to assess in real time what virus has affected a particular village in Sierra Leone might need cheap, portable devices. But for those generating massive data sets, it might be more efficient and cost effective to ship samples to centralized commercial operations, especially when the laboratories are required to meet exacting standards for quality control and sample tracking, as in clinical applications.

Today's 'breakout' clinical application of DNA sequencing — in terms of the sheer number of tests conducted — is prenatal testing for the presence of an abnormal number of chromosomes, such as trisomy 21, which causes Down's syndrome. This test now relies on detecting the small amount of cell-free fetal DNA that circulates in maternal blood. Not even imagined at the end of the Human Genome Project, it has been described as “the fastest growing genetic test in medical history”4. In fact, experts in the field estimate that some 4 million to 6 million pregnant women are now receiving this test each year worldwide, and that the number will surpass 15 million within a decade (D. Bianchi, D. Lo and D. Zhou, personal communication). Some of the hallmarks of the test seem likely to characterize many future applications of DNA sequencing in primary care: it is non-invasive, easy to perform and has low requirements for nucleotide-level accuracy (chromosomes can be counted without assessing sequence variation).

In high-income countries, genome sequencing is already used routinely to evaluate children with ill-defined congenital conditions. Analyses of the resulting sequences can reveal the disease-causing mutations in around 30% of such cases5, 6 — a figure that will only rise as the ability to interpret the data matures. In some instances, the resulting diagnoses have led to dramatic improvements in clinical management7,8. More typically, they benefit both families and physicians by ending a diagnostic odyssey and providing clinical clarity.

In oncology, considerable investments are being poured into the development of liquid biopsies9. It is easy to imagine such a sequence-based cancer test becoming a routine screening tool, used much like Pap smears and colonoscopies. With the advent of cancer treatments that target specific mutations, rather than tumour types10, liquid biopsies could ultimately guide therapeutic interventions even when tumours are known to exist only from DNA-sequence signatures present in blood samples.

Various applications can be envisioned outside the clinic, too, particularly for hand-held DNA sequencers. Epidemiologists and even caregivers working in rural areas could use such devices to test air, water, food, and animal and insect vectors, not to mention human throat swabs and body fluids. In fact, easy access to DNA-sequencing technologies in low- and middle-income countries is already facilitating projects such as the Global Virome Project. This aims to sequence numerous samples of wildlife DNA to identify a significant fraction of the viruses that can be transmitted into humans and cause disease.

Meanwhile, public-health specialists are starting to discuss how they might sequence the DNA of all the microorganisms in the waste-water outlets of entire cities to speed up the recognition of disease outbreaks. And marine biologists are exploring ways to monitor the health of the oceans through systematic metagenomic studies.

On the street, portable instruments could bring DNA analysis out of the crime lab and make it a front-line policing tool. Police might be able to 'read' people's DNA, much as they currently check car number plates or identification documents. In fact, the degree to which cheap and easy DNA sequencing opens up possibilities for mass surveillance has recently sparked concern among human-rights groups.

In the home, DNA-sequencing appliances could become the next 'smart' or 'connected' devices, after smoke alarms and thermostats. One commentator even identified the toilet as the ideal place to monitor family health through real-time DNA sequencing11.

Hitting limits

What are the stumbling blocks?

In a mere 40 years, the central goal of putting molecular data about cells to practical use has changed from an informational challenge to a meta-informational one.

“DNA-sequencing appliances could become the next 'smart' or 'connected' devices.”

Take clinical applications of genome-sequence data. It may soon be possible to use DNA sequencing routinely to analyse body fluids obtained for any clinical purpose. But only a vast amount of well-organized data about the multi-year medical histories of millions of people will provide the meta-information needed to establish when to ignore such data and when to act on them.

With respect to medicine, we echo the recommendations of advisory groups such as the US National Research Council's Precision Medicine Committee12 on the need to create a vast “information commons”. This would overlay molecular and clinical data onto the germ-line genome sequences of millions of individuals. Several such population-scale efforts are under way, including the UK Biobank resource and the US All of Us Research Program.

Here we have laid out our best guesses. Surprises are a certainty. In fact, it is possible that decades from now, much of the world's data (now residing on hard drives or in the cloud) will be stored in DNA, and that the main driver of DNA sequencing will be not our quest to tackle disease, but our insatiable appetite for data storage.

参考文献

1 Maxam, A. M. & Gilbert, W. Proc. Natl Acad. Sci. USA 74, 560–564 (1977).

2 Sanger, F., Nicklen, S. & Coulson, A. R. Proc. Natl Acad. Sci. USA 74, 5463–5467 (1977).

3 Shendure, J. et al. Nature http://dx.doi.org/10.1038/nature24286 (2017).

4 Paxton, A. CAP Today (March 2017); available at go.nature.com/2hoipsp.

5 Bick, D. et al. J. Pediatr. Genet. 6, 61–76 (2017).

6 Eldomery, M. K. et al. Genome Med. 9, 26 (2017).

7 Worthey, E. A. et al. Genet. Med. 13, 255–262 (2011).

8 Bainbridge, M. N. et al. Sci. Transl. Med. 3, 87re3 (2011).

9 Alix-Panabières, C. & Pantel, K. Cancer Discov. 6, 479–491 (2016).

10 Garber, K. Science 356, 1111–1112 (2017).

11 Erlich, Y. Genome Res. 25, 1411–1416 (2015).

12 National Research Council. Toward Precision Medicine: Building a Knowledge Network for Biomedical Research and a New Taxonomy of Disease (National Academies Press, 2011); available at go.nature.com/2fmz99