Work

D2.2 Report on integrative database construction

Year: 2025

Type: article

Abstract: One of the key challenges in developing a computationally framework to support the design of synergistic biosystems for pollutant degradation is the fragmentation and heterogeneity of the underlying biological data. Functional information on chemical compounds, reactions, pathways, enzymes, and organisms involved in pollutant biodegradation is distributed across multiple databases, varies in annotation quality, and often lacks standardized formats that allow seamless integration. This lack of harmonized resources hinders systematic identification of candidate microorganisms with biodegradation potential, reconstruction of degradation pathways, and the implementation of reproducible computational workflows for the assistance in synergistic biosystem design, capable of efficiently degrading and sequestering pollutant mixtures. To address this gap, the BIOSYSMO project has developed BIOSYSMOdb, a structured, cross-referenced database that consolidates and organizes diverse biological, metabolic and biochemical data relevant to the degradation and sequestration of environmental pollutants. Developed as part of Task 2.1 of the project BIOSYSMOdb has been conceived as a centralized resource to support the identification of organisms with functional traits relevant to pollutant transformation or uptake. It integrates curated information on metabolic pathways, biochemical reactions, enzymatic functions, and associated taxa, enabling the design and evaluation of biosystems for bioremediation applications. Structured as a relational SQL database, BIOSYSMOdb allows integrative querying and consistent cross-referencing across biological levels. It contains data extracted from reference public repositories (e.g., UniProt1, NCBI2, RefSeq3, MetaCyc4, EAWAG-BBD5), complemented with curated literature information and prepared to include experimental results generated within BIOSYSMO. The resource provides a harmonized, extensible relational design that supports integrative querying, traceability, and scalability, enabling seamless incorporation of that new experimental or curated information throughout the project and after. Beyond static data storage, BIOSYSMOdb serves as the foundation for key bioinformatic workflows, including the generation of Hidden Markov Models (HMMs) to characterize biodegradation gene families, the annotation of microbial communities (via metabarcoding and 16S rRNA amplicon sequencing), and the simulation of metabolic capabilities through genome-scale metabolic (GEMs) models. It is a central component of the computational infrastructure developed under WP2 and is expected to support both internal project needs and broader research efforts. The BIOSYSMOdb is being developed under a methodological framework aligned with the FAIR principles, ensuring data are Findable, Accessible, Interoperable, and Reusable within and beyond the consortium through availability on Zenodo6. Particular emphasis is placed on enhancing findability through the assignment of persistent identifiers and structured metadata, accessibility via open-access deposition, Interoperability through the use of controlled vocabularies and standard ontologies, and reusability by ensuring transparent data provenance, version control, and clear licensing This deliverable provides a comprehensive technical overview of BIOSYSMOdb, detailing its design, implementation, content, and practical utilities. Section 1 introduces the database, while Section 2 describes the data processing methodology, including protocols for collection, curation, harmonisation, and integration. Section 3 outlines the content, structure, and accessibility of BIOSYSMOdb, complemented in Section 4 by additional datasets on bacterial, fungal, and plant metal resistance proteins. Section 5 illustrates the application of BIOSYSMOdb and the complementary datasets through selected case studies, followed by conclusions and references. Key Insights•BIOSYSMOdb integrates critical biodegradation knowledge by integrating chemical, metabolic, enzymatic, and organism-level data from multiple heterogeneous sources into a unified, queryable SQL database aligned with FAIR principles.•Complementary curated protein datasets on metal resistance in bacteria, fungi, and plants expand the functional utility of BIOSYSMOdb beyond organic pollutant degradation, enabling trait screening in diverse biological contexts.•The database supports both data-driven discovery and applied annotation, serving as the backbone for workflows such as HMM-based model generation, genome-scale metabolic reconstructions, and trait detection in taxonomic or metagenomic datasets.•Practical case studies confirm the robustness and versatility of the system, showcasing its ability to validate protein clusters, infer functional traits from real environmental samples, and guide experimental prioritization.•BIOSYSMOdb is a dynamic and extensible platform, designed to accommodate new data and analytical tools over the course of the project and beyond, supporting long-term research in environmental microbiology and bioremediation.

Source: Zenodo (CERN European Organization for Nuclear Research)

DOI: 10.5281/zenodo.21002744