Submission deadline November 11th 2020

In recent years, the human microbiome has been paid great attention. Several large-scale studies have pointed out the microbiome as a key player in intestinal and non-intestinal diseases. High expectations have been put on the use of microbiome data in clinical use for diagnostics, prognostics and therapeutics, as well as to focus on its causality role in diseases. However, these promising applications are still in their infancy.

Machine Learning (ML) methods offer great potential to continue growing microbiome science. ML algorithms are developed to process high dimensional data and to deal with uncertainty and noise, while the aims of the algorithms are multiple: classification, prediction, etc. However, to maximize the combinatory potential of these emerging fields (microbiome and ML), several challenges must first be overcome. One of the reasons is that microbiome data are inherently convoluted, noisy and highly variable, and non-standard analytical methodologies are therefore required to unlock its clinical and scientific potential. In this manner, although there are available a wide range of statistical modelling and ML methods, sub-optimal implementation often leads to errors, over-fitting and misleading results, due to a lack of good analytical practices and ML expertise in the microbiome community.

Considering these facts, our Research Topic is based on the idea that microbiome science has not benefited enough from the interaction with available machine learning methods. This Research Topic will allow the field to advance in this mission, through its dissemination as well as the participation of experts over the world in the discussion.

The aim of this Research Topic is to collect all article types (reviews, research papers, methods papers, …) related to machine learning in relation to microbiome data. We encourage manuscripts about review, evaluation or application of ML-based models, software packages and web servers for specific prediction problems in microbiome data, as well as the development on novel ones. We welcome the submission of research articles that use machine learning as an underlying modeling strategy and/or primary data analysis tool in any kind of microbiome data. Biomedical data is a must-have, but we are open to any kind of field, from plants to environmental, the goal is to advance in the use of ML techniques in microbiome data and gain insight into current problems and methods in ML on gut microbiome data. Thus, we expect submissions from different expertise fields: Biological Sciences, Computer and Information Sciences, Mathematics and Applied Statistics, Translational Medicine, Environmental Biotechnology as well as Physical Sciences including Network Science.

Thus, specific topics may include, but are not limited to:

• Optimization of data preparation of microbiome data to use ML techniques
• Tools and pipelines to analyze microbiome data with ML methods, for experts and non-experts
• Latest ML algorithms with applications in taxa and gene function prediction
• ML models to extract potential microbiome-biomarkers
• ML approaches and applications for integrating multi-level omics data
• ML models for early disease prediction and prevention
• Studies based on dynamical and prospective models of microbiome
• Studies based on multi-omics data, e.g. the combination of genomic, transcriptomic, epigenomic, or proteomic data

Submission closed