Abstract:ABSTRACT With the concomitant advances in both the microbiome and machine learning fields, the gut microbiome has become of great interest for the potential discovery of biomarkers to be used in the classification of the host health status. Shotgun metagenomics data derived from the human microbiome is composed of a high-dimensional set of microbial features. The use of such complex data for the modeling of host-microbiome interactions remains a challenge as retaining de novo content yields a highly granular set of microbial features. In this study, we compared the prediction performances of machine learning approaches according to different types of data representations derived from shotgun metagenomics. These representations include commonly used taxonomic and functional profiles and the more granular gene cluster approach. For the five case-control datasets used in this study (Type 2 diabetes, obesity, liver cirrhosis, colorectal cancer, and inflammatory bowel disease), gene-based approaches, whether used alone or in combination with reference-based data types, allowed improved or similar classification performances as the taxonomic and functional profiles. In addition, we show that using subsets of gene families from specific functional categories of genes highlight the importance of these functions on the host phenotype. This study demonstrates that both reference-free microbiome representations and curated metagenomic annotations can provide relevant representations for machine learning based on metagenomic data. IMPORTANCE Data representation is an essential part of machine learning performance when using metagenomic data. In this work, we show that different microbiome representations provide varied host phenotype classification performance depending on the dataset. In classification tasks, untargeted microbiome gene content can provide similar or improved classification compared to taxonomical profiling. Feature selection based on biological function also improves classification performance for some pathologies. Function-based feature selection combined with interpretable machine learning algorithms can generate new hypotheses that can potentially be assayed mechanistically. This work thus proposes new approaches to represent microbiome data for machine learning that can potentiate the findings associated with metagenomic data.

Metagenomic biomarker discovery and explanation

Mining the unknown: a systems approach to metabolite identification combining genetic and metabolic information.

Assisted Selection of Biomarkers by Linear Discriminant Analysis Effect Size (LEfSe) in Microbiome Data.

High Throughput and Quantitative Measurement of Microbial Metabolome by Gas Chromatography/Mass Spectrometry Using Automated Alkyl Chloroformate Derivatization.

Large-scale microbiome data integration enables robust biomarker identification

MarkerML - Marker Feature Identification in Metagenomic Datasets Using Interpretable Machine Learning

A Machine Learning Framework of Functional Biomarker Discovery for Different Microbial Communities Based on Metagenomic Data.

biosigner: A New Method for the Discovery of Significant Molecular Signatures from Omics Data

A Comparative Evaluation of Tools to Predict Metabolite Profiles From Microbiome Sequencing Data

Methodology for biomarker discovery with reproducibility in microbiome data using machine learning

Metagenome comparison (MC): A new framework for detecting unique/enriched OMUs (operational metagenomic units) derived from whole-genome sequencing reads

Infer Disease-Associated Microbial Biomarkers Based on Metagenomic and Metatranscriptomic Data

MetaMarker: a Pipeline for De Novo Discovery of Novel Metagenomic Biomarkers

A systematic machine learning and data type comparison yields metagenomic predictors of infant age, sex, breastfeeding, antibiotic usage, country of origin, and delivery type

Gene-based microbiome representation enhances host phenotype classification

Beyond Samples: A Metric Revealing More Connections of Gut Microbiota Between Individuals

Microbiome Sample Comparison and Search: from Pair-Wise Calculations to Model-Based Matching

Mimatch: a Microbial Metabolic Background Matching Tool for Mitigating Host Confounding in Metagenomics Research

A metabolomics pipeline for the mechanistic interrogation of the gut microbiome

Challenges in benchmarking metagenomic profilers

Benchmarking Metagenomics Tools for Taxonomic Classification