Search for a command to run...
Background: Genome-scale metabolic models (GEMs) are foundational tools for systems biology, enabling quantitative interrogation of human metabolism across physiological and pathological states. However, many legacy reconstructions exhibit heterogeneous identifier usage, incomplete pathway integration, and limited thermodynamic refinement, constraining reproducibility, interoperability, and translational applicability. Methods: We present EHMN 2026, an update of the Edinburgh Human Metabolic Network. The reconstruction was refined through systematic identifier reconciliation using MetaNetX and ChEBI mappings, duplicate reaction consolidation, thermodynamic directionality assessment, and structured pathway annotation via Reactome. The final model was encoded in Systems Biology Markup Language (SBML) Level 3 Version 2 with the Flux Balance Constraints (FBC2) package, ensuring explicit gene–protein–reaction (GPR) representation and compatibility with modern constraint-based modelling toolchains. Results: EHMN 2026 comprises 11 compartments, 14,321 metabolites (species), and 22,642 reactions, supported by 3996 gene products. Of all reactions, 9638 (42.6%) contain GPR associations, linking metabolic transformations to 2887 unique Ensembl gene identifiers (ENSG). Pathway integration yielded 2194 unique Reactome identifiers, providing structured pathway-level organisation of metabolic functions. Thermodynamic refinement reduced infeasible energy-generating cycles and improved reaction directionality coherence while preserving global network connectivity. The reconstruction is fully SBML-compliant and portable across major modelling platforms. Compared with Recon3D and Human1, EHMN 2026 uniquely combines native Reactome reaction-level annotation, systematic MetaNetX identifier harmonisation, documented thermodynamic cycle elimination (37 cycles, 0 remaining), and an 11-compartment architecture supporting organelle-specific modelling—features designed for QSP and multi-layer integration applications. Conclusions: EHMN 2026 delivers a rigorously harmonised, thermodynamically refined, and pathway-annotated human metabolic reconstruction with enhanced annotation depth and standards-based interoperability. By combining genome-scale coverage with structured gene and pathway integration, the model establishes a robust computational backbone for reproducible metabolic analysis and provides a scalable foundation for future multi-layer systems pharmacology and integrative modelling frameworks.