Accurate and high-resolution classification of synthetic and natural community amplicon data

在过去的十年中，已经开发了许多用于分析扩增子数据的计算方法，并成功地应用于与植物相关的微生物群落的研究。尽管速度和准确性有所提高，但目前的计算方法仍存在许多局限性，其中最严重的局限性与使用人工构造指定的操作分类单位（otu）有关。这些序列群，估计起源于相同的微生物物种，是基于任意的序列相似性阈值，不对应有意义的功能生物实体。在这里，我们提出实现一种精确的算法，专门设计用于处理从合成群落实验中获得的扩增子数据，利用我们对输入微生物菌株的精确了解。我们还将使用递归分类方法和动态数据库方案将该算法扩展到处理自然群落扩增子数据。最后，我们将使用模拟合成社区进行实验，以验证这些算法并优化其实现。

Over the past decade, numerous computational methods for the analysis of amplicon data have been developed and successfully apply to the study of microbial communities associated with plants. Despite improvements in speed and accuracy, current computational methods suffer from a number of limitations, the most severe of which are associated with the use of the artificial constructs designated Operational Taxonomic Units (OTUs). These groups of sequences, estimated to originate from the same microbial species, are based on arbitrary sequence similarity thresholds which do not correspond to meaningful functional biological entities. Here we propose to implement an accurate and precise algorithm specifically designed to process amplicon data obtained from experiments with synthetic communities by taking advantage of our precise knowledge of the input microbial strains. We will also extend this algorithm for processing natural community amplicon data by using a recursive classification approach and a dynamic database scheme. Finally, we will perform experiments with mock synthetic communities to validate these algorithms and optimize their implementation.