Biological Oceanography, Leibniz Institute for Baltic Sea Research Warnemünde, Rostock, Mecklenburg-Western Pomerania, Germany.
Maritime Graphics, Fraunhofer Institute for Computer Graphics Research, Rostock, Mecklenburg-Western Pomerania, Germany.
Mar Pollut Bull. 2019 Dec;149:110530. doi: 10.1016/j.marpolbul.2019.110530. Epub 2019 Aug 24.
Machine learning algorithms can be trained on complex data sets to detect, predict, or model specific aspects. Aim of this study was to train an artificial neural network in comparison to a Random Forest model to detect induced changes in microbial communities, in order to support environmental monitoring efforts of contamination events. Models were trained on taxon count tables obtained via next-generation amplicon sequencing of water column samples originating from a lab microcosm incubation experiment conducted over 140 days to determine the effects of glyphosate on succession within brackish-water microbial communities. Glyphosate-treated assemblages were classified correctly; a subsetting approach identified the taxa primarily responsible for this, permitting the reduction of input features. This study demonstrates the potential of artificial neural networks to predict indicator species for glyphosate contamination. The results could empower the development of environmental monitoring strategies with applications limited to neither glyphosate nor amplicon sequence data.
机器学习算法可以在复杂的数据集上进行训练，以检测、预测或模拟特定方面。本研究的目的是训练人工神经网络与随机森林模型进行比较，以检测微生物群落中的诱导变化，从而支持对污染事件的环境监测工作。模型是在通过对来源于实验室微宇宙培养实验的水柱样本进行下一代扩增子测序获得的分类计数表上进行训练的，该实验进行了 140 多天，以确定草甘膦对咸水微生物群落演替的影响。草甘膦处理的组合被正确分类；一种子集方法确定了主要负责这一点的分类群，从而可以减少输入特征。本研究证明了人工神经网络预测草甘膦污染指示物种的潜力。结果可以为环境监测策略的发展提供支持，其应用不仅限于草甘膦或扩增子序列数据。
