The ModelSEED (https://modelseed.org) biochemistry database serves as the foundation of biochemical data underlying the in-silico reconstruction of metabolism in ModelSEED and KBase (https://kbase.org). The database is continually updated with improvements regarding several properties that distinguish it from other published biochemical resources, such as compartmentalization and transport, charging of molecular structures, balancing of protons within reactions, and the integration of thermodynamics. The paper describes the collation of public resources, such as KEGG and MetaCyc; the identification of redundancies and the computation of thermodynamic properties using multiple approaches such as eQuilibrator. The ModelSEED biochemistry database is open-source, and extensible by the user community, with all data found in GitHub (https://github.com/ModelSEED/ModelSEEDDatabase). The repository contains scripts and documentation that describes how users may maintain their local instance, serving the data via SOLR. Finally, the ModelSEED biochemistry database is fully searchable, at both ModelSEED (https://modelseed.org/biochem/reactions) and KBase (https://narrative.kbase.us/#biochem-search/).
metabolic and signaling pathways