19 February 2020 in NAR Genomics and Bioinformatics
Large-scale metagenomic assemblies have uncovered thousands of new species greatly expanding the known diversity of microbiomes in specific habitats. To investigate the roles of these uncultured species in human health or the environment, researchers need to incorporate their genome assemblies into a reference database for taxonomic classification. However, this procedure is hindered by the lack of a well-curated taxonomic tree for newly discovered species, which is required by current metagenomics tools. Here we report DeepMicrobes, a deep learning-based computational framework for taxonomic classification that allows researchers to bypass this limitation. We show the advantage of DeepMicrobes over state-of-the-art tools in species and genus identification and comparable accuracy in abundance estimation. We trained DeepMicrobes on genomes reconstructed from gut microbiomes and discovered potential novel signatures in inflammatory bowel diseases. DeepMicrobes facilitates effective investigations into the uncharacterized roles of metagenomic species.
20 October 2020 in Allergy
14 October 2024 in iMetaOmics
Gene expression is a molecular phenotype shaped by the interplay between genotype and environment. The microbiome represents a critical environmental exposure for the host. However, the genotype‐microbiome interactions (GMIs) shaping the host transcriptome remain largely unexplored. Here, we integrated conjunctival multiomics data from 120 pairs of twins to investigate GMIs. We identified 272,972 expression quantitative trait loci associated with 5946 genes and 241,073 genotype‐controlled correlations between gene expression and microbial abundance. We developed a modeling approach, MicroGenix, that screens for GMIs from host genome, transcriptome, and microbiome data and identifies GMIs associated with the disease through gene‐based association tests. We applied MicroGenix to the conjunctival data set and found that incorporating GMIs into gene expression prediction models significantly increased prediction accuracy. The genes with increased accuracy were overrepresented by those encoding cell adhesion molecules. We further used MicroGenix to predict the transcriptome from conjunctival metagenomes and identified GMIs associated with ocular surface disease. Our work provides a resource for studying host‐microbiome interactions at the conjunctiva and a computational approach to investigate GMIs using multiomics data.
20 June 2026 in Nature Communications
Selective enrichment of target bacteria from complex communities, such as the human microbiome, has remained a challenge. Here, we report precision single-cell culturomics based on label-free morphology, Raman spectrometry, and Laser-Induced Forward Transfer (LIFT) technology. This approach operates at the level of single microbial cells, many generations before these cells form visible colonies. We develop a machine learning-based framework that achieves species-level identification of single cells in complex microbiome and achieve selective culturing for or against specific bacteria in fecal or vaginal samples, and quantify some of the cellular components based on Raman spectra. Genomic analysis of single-cell cultures reveals that short-term antibiotic use promotes both pre-existing resistance and de novo mutations of gut commensals, alongside convergent evolution across species. Our precision culturomics method provides a powerful tool for morphological, metabolic, and genomic analysis of microbial phenotypic variations at the single-cell level in microbiome studies.