Developing an Event Prediction and Correction Framework for Microbial Management in Drinking Water Systems.

饮用水中充满了微生物生命。事实上，饮用水每升可以含有数百万到数亿个微生物细胞，它们的进化历史和能力都截然不同。例如，微生物可以(1)通过致病影响人类健康，(2)腐蚀基础设施，(3)还会恶化水的美学质量。为了限制这些有害的情况，饮用水公司投入大量的劳动力、能源和资金，通过使用消毒来限制微生物的存在。尽管消毒方法在减少水传播疾病的发病率方面是有效的，但它们并不是100%成功的，微生物群落仍然存在。因此，饮用水公司还通过实施严格的采样计划来从事微生物管理，以期及早发现微生物污染事件。这些早期检测程序本质上是反应性的，只有在问题发生后才能检测到，并且仅限于通知试图减轻对消费者构成的迫在眉睫的风险的策略。此外，它们通常也只关注致病微生物，而忽视所有其他微生物影响(例如，引起腐蚀的细菌使供水管道恶化)。微生物管理中的这些低效率可以通过从饮用水行业现有的早期事件检测和缓解方法过渡到事件预测和纠正(EPC)框架来补救。EPC框架将使饮用水公司能够预测一系列有害微生物(疾病/腐蚀/气味/味道致病者)在操作上相关的时间尺度上带来的风险，并允许启动衡量和积极主动的纠正行动战略，以在这种风险显现之前消除它。制定强有力的EPC框架的关键是(1)确定饮用水系统中可作为预测指标的关键位置，(2)量化这些位置的时间动态及其与整个饮用水系统的关系，以及(3)开发根据微生物群落数据提供信息的统计模型，以预测污染事件。在这项研究中，我们将进行广泛的努力，以确定苏格兰多个饮用水系统中的细菌、原生动物和真菌群落。随后将在一个有代表性的饮用水系统进行长期采样活动，以量化饮用水微生物群的空间和时间动态。通过利用正在进行的核酸革命，我们将能够以前所未有的详细程度描述饮用水微生物群落。这种详细的定量洞察将被用来参数化和塑造一个统计模型，该模型描述复杂微生物群落的组装并预测它们在饮用水系统中的命运。该项目在一系列时间尺度上有几个预期的好处。在短期内，这个项目将大大提高我们对饮用水微生物群落的了解，而饮用水微生物群落传统上一直处于研究不足的状态。在中期，它将实现饮用水系统的预测性管理，这将有助于防止微生物污染问题，而不是一旦发生就解决它们，这可能是危险、昂贵和费力的。此外，预测模型还可以使我们隔离和处理饮用水处理厂内的微生物风险源，从而防止其进入分配系统。从中长期来看，我们预计，在饮用水部门建立可预测的微生物管理能力将使我们能够有益地操纵饮用水微生物群，以改变我们处理和输送水的方式。

Drinking water is teeming with microbial life. In fact, drinking water can contain anywhere from millions to hundreds of millions of microbial cells per litre, all with extremely different evolutionary histories and abilities. For example, microbes can (1) affect human health by causing diseases, (2) corrode infrastructure, and (3) also deteriorate the aesthetic quality of water. In an effort to limit these detrimental scenarios, drinking water companies invest significant amounts of labour, energy, and money towards limiting microbial presence through the use of disinfection. Though disinfection approaches have been effective in reducing the incidence of waterborne diseases, they are not 100% successful and microbial communities persist. As a result, drinking water companies also engage in microbial management by implementing rigorous sampling programs with the goal of early detection of microbial contamination events. These early detection programs are reactive in nature and can only detect a problem once it has occurred and are limited to informing strategies that try to mitigate the imminent risk posed to consumers. Further, they also typically focus only on pathogenic microorganisms and ignore all other microbial impacts (e.g. corrosion causing bacteria that deteriorate water supply pipes). These inefficiencies in microbial management can be remedied by transitioning from the existing Early Event Detection and Mitigation approach to an Event Prediction and Correction (EPC) framework in the drinking water industry.An EPC framework would enable the drinking water companies to predict the risk presented by an array of detrimental microbes (disease/corrosion/odour/taste causers) over operationally relevant time-scales and allow for the initiation of measured and proactive corrective action strategies to eliminate this risk before it is manifested. The key towards developing a robust EPC framework would be to (1) identify key locations in the drinking water system that can serve as predictive indicators, (2) quantify the temporal dynamics of these locations and how it correlates with the whole drinking water system, and (3) develop statistical models informed by microbial community data to predict contamination events. In this study, we will engage in an extensive effort to characterize the bacterial, protozoal, and fungal communities at multiple drinking water systems in Scotland. This will be followed by a long-term sampling campaign at one representative drinking water system to quantify the spatial and temporal dynamics of the drinking water microbiome. By tapping into the on-going nucleic acid revolution, we will be able to describe the drinking water microbial communities at an unprecedented level of detail. This detailed quantitative insight will be used to parameterize and shape a statistical model that describes assembly of complex microbial communities and predicts their fate in the drinking water system.This project has several anticipated benefits over a range of time-scales. In the short-term, this project will substantially improve our understanding of the drinking water microbial communities, which has been traditionally under-studied. In the medium-term, it will enable the predictive management of the drinking water systems, which will help prevent microbial contamination problems, rather than tackle them once they have occurred which can be risky, expensive, and laborious. Further, predictive models may also allow us to isolate and treat sources of microbial risk within the drinking water treatment plant, thus preventing its entry into the distribution system. Over the medium-long term, we anticipate that building a predictive microbial management capacity in the drinking water sector will enable us to beneficially manipulate the drinking water microbiome to transform the way we treat and deliver water.