机器学习在微生物组宿主表型预测中的应用

Applications of machine learning in predicting host phenotype based on microbiome

摘要:

随着大数据时代的到来，如何将生物组学海量数据转化为易理解及可视化的知识是当前生物信息学面临的重要挑战之一。为了处理复杂、高维的微生物组数据，目前机器学习算法已被应用于人体微生物组研究，以揭示疾病背后的复杂机制。本文首先简述了微生物组数据处理方法及常用的机器学习算法，如支持向量机（SVM）、随机森林（RF）和人工神经网络（ANN）等，然后对机器学习的工作流程及其要点进行阐述，并探讨了机器学习算法在基于微生物组数据预测宿主表型方面的应用。最后以唾液微生物组数据预测口腔异味为例，实现了机器学习算法的模型构建与评估分析，并提供了可用于微生物组研究实践的R/Python代码（https：//github.com/LiLabZSU/microbioML）。

Abstract:

With the advent of the era of big data, how to transform the omics data into easy-to-understand and visualized knowledge is one of the important challenges in bioinformatics. Recently, machine learning techniques had been utilized to analyze the complicated, high-dimensional microbiome data to address the complex mechanisms of human diseases. Here, we firstly summarized microbiome data procession approaches and the most commonly used machine learning algorithms, such as support vector machine (SVM), random forest (RF), and artificial neural networks (ANN). Then, the workflow of machine learning studies was described, and the application of ML algorithms in predicting host phenotypes based on microbiome data was evaluated. Finally, the model construction and validation of machine learning algorithms were demonstrated by using saliva microbiome data to predict oral malodour as an example, and R/Python code for practical data analysis was provided (https://github.com/LiLabZSU/microbioML).