当前位置: X-MOL 学术Environ. MicroBiol. Rep › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Space, time and microdiversity: towards a resolution revolution in microbiomics
Environmental Microbiology Reports ( IF 3.3 ) Pub Date : 2020-10-15 , DOI: 10.1111/1758-2229.12897
Lucas Paoli 1 , Shinichi Sunagawa 1
Affiliation  

Predicting the future of our field can be an exciting and valuable thought experiment. For this exercise to be more than just wishful thinking, we set out to discuss some areas in environmental microbiology where we think current evidence suggests a particularly rapid change is taking place. To this end, through three imagined conversations, we explore some anticipated impacts on the current state of the field by what we refer to as a resolution revolution in microbiomics. We then conclude with a discussion of what in our view will help overcoming critical limitations in advancing our understanding of microbiome structure, function and dynamics in light of technical, analytical and conceptual developments.

Ten years after graduating, a former PhD student and their advisor are reunited once more at a physical conference, an experience necessarily rare owing to environmental impacts, and sorely missed during recurring pandemic events. After catching up on years of personal news and gossip over dinner, the conversation inevitably turns to microbiome science…

Advisor: So, thinking back, what were the main challenges that your research faced during those years in my lab?

Former student: Difficult to say really, but besides the constant struggle to keep up with the avalanche of microbiome papers being published every week… I would probably argue that most of the data available before the resolution revolution in microbiomics only gave a very coarse picture of the natural systems we were looking at. The limited spatial resolution available blurred our view of local heterogeneity, which was inevitably pooled or undersampled. Plus the limited temporal resolution only provided snapshots of the dynamic changes going on. Phylogenomic resolution was also a pain as it usually collapsed natural variation within and across microbial populations, restricting our understanding of strain‐level microdiversity. In the end, do you think that the field has actually managed to fully resolve all this, if you'll pardon the pun?

Advisor: I remember us speculating in a lab meeting once that a combination of robotics, microfluidics, single‐cell, gap‐free replicons sequencing would fix all of our problems. Actually, I think that the field has made huge leaps in most of these areas, although there is always room for improving resolution even further.

Former student: Definitely. I think the advances in data generation have changed the way we look at the sequence space, literally. We've moved away from unidimensional representations like FASTA files, towards the richer, graph‐based visualization of genomic variation. I'm actually quite happy about this: the graphs are a far better way to capture within‐species diversity and genome plasticity. I also suspect these changes in computational microbiomics were crucial in mainstreaming microdiversity.

Advisor: Absolutely! It's also been nice to see that this shift in data representation was accompanied by changes in lab‐based experiments. As we predicted back in the day, the shift in focus, back from descriptive, hypotheses‐generating data correlations to experimental hypothesis‐testing has genuinely begun to improve our mechanistic understanding of these systems. Still, even back then it was clear that tackling genes of unknown function and strain‐level differences would remain a challenge for quite some time.

Former student: I agree. But what do you think prompted this change?

Advisor: Well… arguably it all goes back to Baas Becking's tenet ‘everything is everywhere, but, the environment selects' and the many ways we have been able to use it as a working hypothesis. By adjusting the focus wheel, we would first define: What actually is a ‘thing’? Is it bacterial life? a microbial species? an eco‐evolutionary cohesive unit of genomic variants? an individual strain? a gene? an adaptive mutation? With haplotype‐resolved analyses, I feel we have made significant progress on some long‐standing questions. For instance: are conspecific strain population dynamics driven by growth and extinction of a standing stock of genomic variation, by the generation of new diversity, or by migration? How does the migration (or controlled addition/removal) of individual strains impact interactions within and functional properties of the whole community? I would say this promoted the value of microdiversity‐focused ‘N plus 1’ and ‘N minus 1’ types of experiments. In just these past few years, such experiments have shown how instrumental microdiversity is to understand microbiomes. For instance, the impact of ‘1’ on ‘N' and vice‐versa depends on the strain‐specific gene content of ‘1’ and ‘N', just as much as their species‐level composition. Of course, this has put a few noses out of joint by challenging the portability of results based on single lab‐grown strains to natural communities, since the model strains that many people worked with are often not the ones we find in nature, and natural communities are composed of diverse rather than clonal populations.

Former student: Very true. Along these lines, I think computational and experimental microbiomics is actually now progressing hand in hand pretty well. You cannot peer‐proof a computational analysis without functional validation anymore, the field just will not have it. Which is great for science, but… I feel this is complicating the process of me getting tenured!

Advisor: Right, because you think we had it easy?

Former student: Well actually… never mind. But coming back to the resolution revolution, I expected that it would have allowed us to better understand community functions, to predict how these would change in response to environmental and biotic changes, consequently to manipulate them, and eventually to inform decision makers. But it seems that there is still quite some way to go! We do have a few specific examples with a good enough mechanistic understanding to make robust predictions, but they remain the exception. Of course, the plethora of empiric data coupled with new AI tools now allow everyone to predict associations between microbiome features and phenotypes, particularly in the biomedical area. But knowing that it works is still seen by too many as more important than knowing how it works, so efforts to unravel the molecular mechanisms are sadly neglected. This does leave behind loads of black boxes for basic researchers to crack open and illuminate, of course. And on the upside, the microbiomics resolution revolution has paved the way to establish congruent ecological and evolutionary microbial units, which should allow the field to address how these affect microbiome structure, functions and dynamics, and this way further unite ecology and evolution.

Advisor: One thing that has not changed for sure though, is your focus on the conceptual and academic advances. But do not forget when completing that tenure application to describe how these advances in high‐resolution microbiomics are now impacting everyday lives!

In a hospital, a patient suffering from what appears to be a gastrointestinal infection carefully listens to their doctor, who, based on results of tests conducted earlier the same day, is suggesting a new type of treatment…

Doctor: If you recall, this morning I said we would need a current profile of your gut microbiome. To that end, we obtained a sample… upon induction of bowel movements… well, you know what I mean. Based on analysis of that sample, we now know the exact composition of your gut microbiome, not only what species and in what amount, but more specifically which variants within each species, and importantly the one causing the disease. This microbiome profile will be added to your unified medical records, where we already have your genome, immune profile and previous time points of your microbiome. It will allow us to take the most appropriate therapeutic decision for your specific situation.

Patient: Okay Doc. I read somewhere that a drug can lose its effect if you have certain bugs in your gut. Is that why you needed this information before deciding on a treatment?

Doctor: Correct. You can have drug‐bug or bug‐drug interactions, where either a drug will have a deleterious effect on your microbiome or your microbiome will prevent the drug from working properly. This is why much of modern medicine has moved away from a ‘one‐size‐fits‐all’ model. We now take microbiome‐informed therapy decisions based on a current picture of your microbiome. This is particularly relevant in your case. We now know that the pathogen causing your infection is resistant to most antibiotics and viruses that may target and kill it. But please do not worry. Fortunately the spread of antibiotic resistance during the past decades pushed the development of alternative methods and we now have several new options available for treating you, such as using close, non‐pathogenic relatives of the offending bacteria to out‐compete it and chase it out of your gut.

Patient: Does that mean that instead of a drug, the treatment you are suggesting is actually a microbe itself? Is that going to work?

Doctor: Yes, exactly. The chances for success of this approach are very good, but they do depend on several factors. Using the information we have on your immune system, as well as the composition of your microbiome, I browsed a strain database and selected the best candidate for the job. There are sometimes situations where this approach does not work out, for instance if the combination of a patient's immune and microbiome profile does not allow for a predictive outcome. But the science is now quite mature and allows us in many cases to restore a perturbed microbiome. And sometimes, AI uses the data collected from patients over the past couple of decades to inform us about therapeutic strategies, although we do not always know the exact mechanisms. These microbial strain‐mediated strategies are being used in more and more cases. For instance, oncologists have used select strains to stimulate the immune system during cancer immunotherapy.

Patient: Fascinating stuff! Although if you ask me, as long as it works, it does not matter how it works, does it?

It's October 2031 and after the all‐time high summer temperatures a parent and their child are off to enjoy the school break on a picturesque but crowded beach on the English Channel. After walking past the recently renovated botanical garden, which now uses soil with a microbiome engineered to enhance the growth of exotic‐plants, they finally have sight of the sea. Just before reaching the pebbles, the child stops to read an information panel…

Child: Hey, look! There is a map of the beach. Look, it also shows the quality of the water. That's so cool! So, I think here are the rivers and streams that go to the sea. They're green so it means good! Ah no, except this orange one. But maybe it's because of the sheep over there? Anyway, it says the water at the beach is excellent, so it's safe. Let us go swim!

Parent, pulling out a mobile device from their bag: Wait, come have a look at the augmented version of the map. You can see the water quality for every week over the last few years, and for several places along all of the rivers and streams on the map. Actually, this is pretty interesting. I remember being here before you were born and the old map was nothing like this. They did things quite differently. I think they would filter some water and specifically grow indicator gut bugs to see how much human waste was getting into the water. Look, here it says that with high‐throughput and high‐resolution sequencing strategies, they can now have detailed profiles with many more time points, and a spatial sampling. It says that they now know exactly, which strains of which species are dominating. I guess that's good because you can really tell the difference between bugs from the beach, animals or sewage.

Child: Hey, look! You can press that panel to see where the bugs come from. It says that by decoding their DNA, they can track the bacteria back to their source. Cool, most of the E. coli at the beach is not from humans and actually comes from the stream next to the meadow. There's also a bit coming from the water treatment plant upstream. I guess that makes sense.

Parent: Interesting, you'll have to tell your teacher about this after the holidays. Look, it also says that you can use similar microbial profiles to monitor air quality, although this is still just at the research stage. But for the water, they also have a second analysis: this one is about toxic blooms. What does it say?

Child: So, it says that naturally present algae and cyanobacteria can, under certain circumstances, form harmful blooms. When that happens we cannot go swimming where the bloom is. But based on the monitoring, they can predict when and where this will happen. It says that there is a 42.26% chance of a bloom starting on the other side of the beach in 2 weeks. So maybe that part will be closed soon?

Parent: I think so! It says they are testing a virus‐based method to control the blooms, but we better dive in while we still can!

Child, running towards the water: Way ahead of you!!!

For these imaginary conversations to become reality, and the field of microbiomics to advance as a whole, we find several challenges need to be overcome. In the case of the human gut, for instance, it is established that microbial communities undergo strong day‐to‐day compositional variations at the ‘species’ level (David et al. 2014), and that these ‘species’ harbour a relatively stable intra‐specific diversity (Schloissnig et al. 2013). This microdiversity is subject to both ecological and evolutionary processes (Garud et al. 2019; Zhao et al. 2019) and, as shown for other systems, it can affect the ‘species’ level composition of a community (Gómez et al. 2016; Padfield et al. 2020). However, answers to seemingly straightforward questions about this and other microbiomes remain elusive. For example, how many different genotypes of a given species (e.g., E. coli) does a gram of soil, an individual human, or a lake contain? What is the spatial structure, if any, of strain populations at the scales of micrometres (from epithelium to lumen) to meters (from oral cavity to rectum)? What are the abundance and gene expression dynamics within conspecific strain populations over time (e.g., from hourly to diurnal and seasonal variations in coastal waters)?

We argue that tackling such questions is currently limited by the operational resolution at which we look at microbial communities along spatial, temporal and phylogenomic dimensions. Advances along each one of these axes will contribute to a more holistic understanding of microbial communities and the intricate relationships between their members. Although such a holistic view may not be within reach yet, we believe that we are well on track and highlight some trends in technical, analytical and conceptual developments showing that we are in fact already amid a resolution revolution in microbiomics.

An increase in spatial and temporal resolution of microbiome composition largely relies on a more intensive sampling and sequencing effort, the latter of which is already being driven by dropping sequencing costs. Additional technological advances in low‐input DNA sequencing strategies (Rinke et al., 2016) combined with automated and remote sampling (Zhang et al., 2020) is likely to provide us with the ability to scale up the spatial and temporal frequency at which microbial communities are sampled, including at remote locations (Marx, 2020). The challenge of resolving within‐species genotypic diversity, or microdiversity, of microbiomes through metagenomics has taken the centre stage over the last few years (Yan et al., 2020; Van Rossum et al., 2020). We foresee that additional technological advances such as genome reconstruction from long‐read metagenomes (Moss et al., 2020; Beaulaurier et al., 2020; Kolmogorov et al., 2020) and high‐throughput single‐cell sequencing (Woykeet al., 2017) will provide the ground to further resolve population genotypic heterogeneity.

In addition to such technological developments, we anticipate the emergence of new strategies for data analysis and representation that need to be able to accommodate ongoing addition of new samples. For instance, in light of the increasing amount of sequencing data, graph‐based approaches represent an opportunity to rethink the mainstream data structures encoding genomic information (Mustafa et al., 2017; Karasikov et al., 2020). As we continuously uncover microbial microdiversity, leading to a steady growth of intra‐specific gene pools and information on genome‐scale structural variations, graphs emerge as an inevitable alternative for representing microbial pan‐genomes (Gautreau et al., 2020; Brown et al. 2020; Quince et al., 2020).

Conceptually, we anticipate these technological and analytical developments to prime a more mechanistic and predictive understanding of microbiomes. On the one hand, we expect the ever‐increasing amount of data to drive associations between microbiome read‐outs and disease state, treatment efficacy, ecosystem function or environmental factors (Sunagawa et al., 2015; Spanogiannopoulos et al., 2016; Wirbel et al., 2019). Such data‐driven analyses will keep generating more testable hypotheses on mechanistic causes and more powerful predictions. On the other hand, models of environmental and host‐associated microbial communities, including synthetic ones, and experimental strategies required to test mechanistic hypotheses are already emerging (Bai et al., 2015; Brugiroux et al., 2016; Bell, 2019). Notably, we expect a multidimensional increase in resolution to help researchers better grasp the complexity and diversity of natural microbial systems, and to influence environmental and health applications. Environmental monitoring strategies are almost certain to be impacted by advances in microbiomics approaches, which, in the case of bathing water quality assessment are already presented as viable alternatives to colony‐forming unit counts of coliforms (Council Directive 2006/7/CE, 2006; Tan et al., 2015). Furthermore, studies identifying strain‐level variation and diversity as drivers of community composition (Gómez et al., 2016; Padfield et al., 2020) and host physiology (e.g., cancer immunotherapy, Fluckiger et al., 2020) exemplify the importance of resolving microdiversity for future translational applications.

Although our focus here was set on nucleic acid sequencing‐based omics, other emerging technologies, including meta‐proteomics, meta‐metabolomics and high‐throughput imaging (Özel Duygan et al., 2020), are arguably also experiencing a resolution revolution. Overcoming the challenge of integrating data across these platforms will certainly promote a more integrative understanding of the evolutionary, ecological and physiological mechanisms that underpin microbial community structure, function and dynamics. With computational and experimental microbiomics advancing hand in hand, we will keep paving solid grounds to accurately predict the outcome of targeted modulations of the microbiome and to engineer its desired functional properties (Cullen et al. 2020; Widder et al. 2016). To foster the impact of such developments and the resultant translational applications, we advocate concomitant efforts into improving microbial literacy across all levels of society. This way, a more prominent role for microbiomics‐informed decisions in policy making could become more than just wishful thinking.



中文翻译:

空间、时间和微多样性:迈向微生物组学的分辨率革命

预测我们领域的未来可能是一个令人兴奋且有价值的思想实验。为了使这项工作不仅仅是一厢情愿的想法,我们开始讨论环境微生物学的一些领域,我们认为目前的证据表明正在发生特别迅速的变化。为此,通过三个想象的对话,我们通过我们所说的微生物组学分辨率革命来探索对该领域当前状态的一些预期影响。然后,我们最后讨论了我们认为哪些将有助于克服关键限制,以根据技术、分析和概念的发展推进我们对微生物组结构、功能和动力学的理解。

毕业十年后一位前博士生和他们的导师在一次实体会议上再次重聚由于环境影响这种经历必然很少见并且在反复出现的大流行事件中非常怀念。在追上多年的个人新闻和晚餐八卦之后话题不可​​避免地转向微生物组科学……

顾问:那么,回想一下,在我实验室的那些年里,您的研究面临的主要挑战是什么?

前学生:很难说真的,但除了不断努力跟上每周发表的大量微生物组论文之外......我们正在研究的自然系统。可用的有限空间分辨率模糊了我们对局部异质性的看法,这不可避免地被合并或采样不足。此外,有限的时间分辨率只能提供动态变化的快照。系统基因组分辨率也是一种痛苦,因为它通常会破坏微生物种群内部和之间的自然变异,从而限制了我们对菌株水平微多样性的理解。最后,如果您原谅双关语,您是否认为该领域实际上已经完全解决了所有这些问题?

顾问:我记得我们曾经在一次实验室会议上推测,机器人、微流体、单细胞、无间隙复制子测序的组合将解决我们所有的问题。实际上,我认为该领域在大多数这些领域都取得了巨大的飞跃,尽管总有进一步提高分辨率的空间。

前学生:当然。我认为数据生成的进步已经从字面上改变了我们看待序列空间的方式。我们已经从像 FASTA 文件这样的一维表示转向更丰富的、基于图形的基因组变异可视化。我实际上对此感到非常高兴:图表是捕捉物种内多样性和基因组可塑性的更好方法。我还怀疑计算微生物组学的这些变化对于微多样性主流化至关重要。

顾问:当然!很高兴看到数据表示的这种转变伴随着实验室实验的变化。正如我们当时预测的那样,焦点从描述性的、假设生成的数据相关性转向实验假设检验的转变,真正开始改善我们对这些系统的机械理解。尽管如此,即使在那时,处理未知功能和菌株水平差异的基因在相当长的一段时间内仍然是一个挑战。

前学生:我同意。但你认为是什么促使了这种变化?

顾问:嗯……可以说这一切都可以追溯到 Baas Becking 的宗旨“万物无处不在,但是,环境选择’以及我们能够将其用作可行假设的多种方式。通过调整调焦轮,我们首先要定义:什么是“东西”?是细菌生命吗?微生物种类?基因组变异的生态进化内聚单元?个别菌株?基因?适应性突变?通过单倍型解析分析,我觉得我们在一些长期存在的问题上取得了重大进展。例如:同种菌株种群动态是由基因组变异存量的增长和灭绝、新多样性的产生还是迁移驱动的?单个菌株的迁移(或受控添加/移除)如何影响整个社区内部的相互作用和功能特性?我会说这提升了以微多样性为重点的“N + ”的价值1' 和 'N - 1' 类型的实验。就在过去的几年中,此类实验已经显示了工具性微生物多样性是如何理解微生物组的。例如,“1”对“N”的影响,反之亦然,取决于“1”和“N”的菌株特异性基因含量,就像它们的物种水平组成一样。当然,这通过挑战基于单个实验室培养菌株的结果到自然群落的可移植性使一些鼻子失去了联系,因为许多人使用的模型菌株通常不是我们在自然界中找到的模型菌株,而且自然社区由多样化而不是克隆种群组成。

前学生:非常真实。沿着这些思路,我认为计算和实验微生物组学实际上正在携手并进。你不能再没有功能验证的情况下对计算分析进行对等证明,这个领域不会有它。这对科学来说很棒,但是……我觉得这使我获得终身职位的过程复杂化了!

顾问:是的,因为你认为我们很容易?

以前的学生:其实……没关系。但回到分辨率革命,我希望它能让我们更好地了解社区功能,预测这些功能将如何响应环境和生物变化而发生变化,从而操纵它们,并最终为决策者提供信息。不过好像还有很长的路要走!我们确实有一些具有足够好的机械理解的特定示例来做出可靠的预测,但它们仍然是例外。当然,大量的经验数据加上新的 AI 工具现在使每个人都可以预测微生物组特征和表型之间的关联,尤其是在生物医学领域。但我们知道的是它的工作原理仍然被太多的比知道更重要的看到它是如何工作的,因此可悲的是,解开分子机制的努力被忽视了。当然,这确实为基础研究人员留下了大量黑匣子来破解和照亮。从好的方面来说,微生物组学分辨率革命为建立一致的生态和进化微生物单元铺平了道路,这应该允许该领域解决这些如何影响微生物组结构、功能和动力学,并通过这种方式进一步统一生态学和进化。

顾问:不过,没有改变的一件事是你对概念和学术进步的关注。但不要忘记在完成任期申请时描述高分辨率微生物组学的这些进步现在如何影响日常生活!

在医院从显示的内容的患者是一种胃肠道感染仔细倾听他们的医生根据当天早些时候进行的试验结果是在暗示一种新型的治疗...

医生:如果你还记得,今天早上我说过我们需要你肠道微生物组的最新资料。为此,我们获得了一个样本……在诱导排便时……好吧,你知道我的意思。基于对该样本的分析,我们现在知道您肠道微生物组的确切组成,不仅是什么物种和数量,而且更具体地说是每个物种中的哪些变异,重要的是导致疾病的变异。该微生物组谱将添加到您的统一医疗记录中,我们已经拥有您的基因组、免疫谱和之前微生物组的时间点。这将使我们能够针对您的具体情况做出最合适的治疗决定。

病人:好的医生。我在某处读到,如果您的肠道中有某些虫子,药物就会失去作用。这就是您在决定治疗之前需要这些信息的原因吗?

医生:正确的。您可能会遇到药物-虫子或虫子-药物相互作用,其中一种药物会对您的微生物组产生有害影响,或者您的微生物组会阻止药物正常发挥作用。这就是为什么大部分现代医学已经摆脱了“一刀切”的模式。我们现在根据您的微生物组的当前图片做出基于微生物组的治疗决策。这与您的情况特别相关。我们现在知道导致您感染的病原体对大多数可能靶向并杀死它的抗生素和病毒具有抗药性。但请不要担心。幸运的是,过去几十年抗生素耐药性的蔓延推动了替代方法的发展,我们现在有几种新的治疗方法可供您选择,例如使用关闭、

患者:这是否意味着您建议的治疗实际上是一种微生物本身而不是药物?那行得通吗?

医生:对,就是这样。这种方法的成功机会非常好,但它们确实取决于几个因素。使用我们掌握的关于您的免疫系统的信息以及您的微生物组的组成,我浏览了一个菌株数据库并选择了最适合该工作的候选人。有时这种方法行不通,例如,如果患者的免疫和微生物组谱的组合不允许预测结果。但科学现在已经相当成熟,并且在许多情况下允许我们恢复被扰乱的微生物组。有时,人工智能使用过去几十年从患者那里收集的数据来告知我们治疗策略,尽管我们并不总是知道确切的机制。这些微生物菌株介导的策略正在越来越多的案例中使用。例如,

病人:迷人的东西!虽然如果你问我,只要能用,怎么用都无所谓,不是吗?

现在是 2031 年 10 月,在经历了前所未有的夏季高温之后,父母和他们的孩子将在英吉利海峡风景如画但拥挤的海滩上享受学校假期。走过最近翻修过的植物园后,他们终于看到了大海,该植物园现在使用土壤和微生物组,旨在促进外来植物的生长。就在到达鹅卵石之前,孩子停下来阅读信息面板……

孩子:嘿,看!有海滩地图。看,它也显示了水的质量。这太酷了!所以,我认为这里是流向大海的河流和溪流。它们是绿色的,所以这意味着很好!啊不,除了这个橙色的。但也许是因为那边的羊?不管怎样,它说海滩的水很好,所以很安全。我们去游泳吧!

家长从包里拿出移动设备: 等等,来看看地图的增强版。您可以看到过去几年中每周的水质,以及地图上所有河流和溪流沿线的几个地方的水质。事实上,这很有趣。我记得在你出生之前就在这里,旧地图完全不是这样。他们做事完全不同。我认为他们会过滤一些水并专门种植指示性肠道虫子,以查看有多少人类排泄物进入水中。看,这里说的是通过高通量和高分辨率测序策略,他们现在可以拥有更多时间点的详细轮廓和空间采样。它说他们现在确切地知道哪些物种的哪些菌株占主导地位。我想这很好,因为你真的可以分辨出海滩上的虫子之间的区别,

孩子:嘿,看!您可以按该面板查看错误来自何处。它说通过解码他们的 DNA,他们可以追踪细菌回到它们的来源。很酷,海滩上的大多数大肠杆菌不是来自人类,而是来自草地旁边的溪流。还有一点来自上游的水处理厂。我想这是有道理的。

家长:有意思,假期过后你得把这件事告诉你的老师。看,它还说您可以使用类似的微生物特征来监测空气质量,尽管这仍处于研究阶段。但是对于水,他们还有第二个分析:这是关于有毒水华的。它说什么?

孩子:所以,它说自然存在的藻类和蓝藻在某些情况下会形成有害的水华。当这种情况发生时,我们不能去盛开的地方游泳。但基于监控,他们可以预测何时何地会发生这种情况。它说有 42.26% 的机会在 2 周内在海滩的另一边开始开花。那么也许那部分很快就会关闭?

家长:我也这么认为!它说他们正在测试一种基于病毒的方法来控制花朵,但我们最好趁我们还有机会潜入!

孩子奔向水面:遥遥领先!!!

为了让这些想象中的对话成为现实,以及微生物组学领域的整体发展,我们发现需要克服几个挑战。例如,在人类肠道的情况下,微生物群落在“物种”水平上经历了强烈的日常组成变化(David等人2014 年),并且这些“物种”具有相对稳定的种内多样性(Schloissnig等人2013 年)。这种微多样性受生态和进化过程的影响(Garud et al . 2019 ; Zhao et al . 2019) 并且,如其他系统所示,它可以影响社区的“物种”级别组成(Gómez等人2016 年;Padfield等人2020 年)。然而,关于这个和其他微生物组的看似简单的问题的答案仍然难以捉摸。例如,一克土壤、一个人或一个湖包含多少种不同基因型的给定物种(例如,大肠杆菌)?如果有的话,在微米(从上皮到管腔)到米(从口腔到直肠)尺度上的菌株种群的空间结构是什么?随着时间的推移,同种菌株种群内的丰度和基因表达动态是什么(例如,沿海水域从每小时到昼夜和季节性变化)?

我们认为,解决这些问题目前受到我们在空间时间系统发育维度上观察微生物群落的操作分辨率的限制。沿着每一个轴的进步将有助于更全面地了解微生物群落及其成员之间的复杂关系。尽管这种整体观点可能还遥不可及,但我们相信我们已经走上正轨,并突出了技术、分析和概念发展的一些趋势,表明我们实际上已经处于微生物组学的分辨率革命之中。

微生物组组成的空间和时间分辨率的提高很大程度上依赖于更密集的采样和测序工作,后者已经受到测序成本下降的推动。低输入 DNA 测序策略(Rinke等人2016 年)与自动和远程采样(Zhang等人2020)相结合的其他技术进步可能会为我们提供扩大空间和时间频率的能力微生物群落被采样,包括在偏远地区(Marx,2020)。在过去几年中,通过宏基因组学解决微生物组的物种内基因型多样性或微多样性的挑战已成为焦点(Yan等人2020 年;Van Rossum等人2020 年)。我们预见到额外的技术进步,例如长读长宏基因组的基因组重建(Moss等人2020 年;Beaulaurier等人2020 年;Kolmogorov等人2020 年)和高通量单细胞测序(Woyke等人2020 年)。 , 2017) 将为进一步解决种群基因型异质性提供基础。

除了这些技术发展之外,我们预计会出现新的数据分析和表示策略,这些策略需要能够适应不断添加的新样本。例如,鉴于测序数据量不断增加,基于图的方法代表了重新思考编码基因组信息的主流数据结构的机会(Mustafa等人2017 年;Karasikov等人2020 年)。随着我们不断发现微生物微多样性,导致特定内基因库和基因组规模结构变异信息的稳定增长,图表成为代表微生物泛基因组的必然选择(Gautreau等人2020 年); 布朗等人2020 年;昆斯等人2020 年)。

从概念上讲,我们预计这些技术和分析发展将为微生物组的更机械化和预测性的理解做好准备。一方面,我们期望不断增加的数据量,以驱动协会之间微生物读出值和疾病状态,治疗效果,生态系统功能或环境因素(砂川等人,。2015 ; Spanogiannopoulos等人,。2016 ; Wirbel2019)。这种数据驱动的分析将不断产生更多关于机械原因的可检验假设和更强大的预测。另一方面,环境和宿主相关微生物群落的模型,包括合成模型,以及测试机械假设所需的实验策略已经出现(Bai等人2015 年;Brugiroux等人2016 年;Bell,2019 年)。值得注意的是,我们期望分辨率的多维增加有助于研究人员更好地掌握天然微生物系统的复杂性和多样性,并影响环境和健康应用。环境监测策略几乎肯定会受到微生物组学方法进步的影响,在沐浴水质评估的情况下,微生物组学方法已经被提出作为大肠菌群形成单位计数的可行替代方案(理事会指令 2006/7/CE,2006 年; Tan等人2015 年)。此外,研究将菌株水平的变异和多样性确定为群落组成的驱动因素(Gómez等人2016 年;Padfield等人2020 年)) 和宿主生理学(例如,癌症免疫疗法,Fluckiger等人2020 年)举例说明了解决微多样性对未来转化应用的重要性。

尽管我们在这里的重点是基于核酸测序的组学,但其他新兴技术,包括元蛋白质组学、元代谢组学和高通量成像(Özel Duygan2020),可以说也正在经历一场分辨率革命。克服跨这些平台整合数据的挑战肯定会促进对支持微生物群落结构、功能和动力学的进化、生态和生理机制的更全面的理解。随着计算和实验微生物组学的推进,我们将继续为准确预测微生物组靶向调节的结果并设计其所需的功能特性奠定坚实的基础(Cullen等人, 2017 年2020 年;威德等人2016 年)。为了促进此类发展和由此产生的转化应用的影响,我们提倡同时努力提高社会各个层面的微生物素养。这样,微生物组学决策在政策制定中的更突出作用可能不仅仅是一厢情愿。

更新日期:2020-10-15
down
wechat
bug