EMERALD METAGENOMICS ANNOTATIONS PIPELINE

EMERALD METAGENOMICS ANNOTATIONS PIPELINE extracts and integrates metadata from research publications into metagenomics databases to enable comparative and contextual analyses of microbial community sequence data and to help account for confounding factors across studies.

Key Features:

• Metadata Extraction: A machine learning framework automatically extracts essential metadata from the narrative sections of research publications, including sample characteristics and molecular methods.
• Integration with Databases: Extracted metadata is integrated and annotated into existing metagenomics resources such as ENA (European Nucleotide Archive) and MGnify.
• Extensive Coverage: The framework processed over 114,099 publications in Europe PMC and extracted and annotated metadata from 19,900 publications describing metagenomic research in ENA and MGnify.

Scientific Applications:

• Comparative analyses: Enables cross-study comparisons of metagenomic sequence data by augmenting records with publication-derived contextual metadata.
• Longitudinal and cross-sectional studies: Supports longitudinal and cross-sectional study designs by providing consistent metadata to track microbial community changes over time or between cohorts.
• Ecological and evolutionary inference: Facilitates analyses of ecological dynamics, evolutionary adaptations, and functional roles of microbiota within different environments by supplying contextual sample and method metadata.

Methodology:

Application of a machine learning framework to automatically extract metadata from narrative sections of Europe PMC publications and the subsequent integration and annotation of those metadata into ENA and MGnify; processing coverage reported as >114,099 Europe PMC publications and metadata extracted/annotated from 19,900 metagenomic publications.