XJTLU-AI-China: PlaszymeDB

PlaszymeDB is part of the iGEM 2025 XJTLU-AI-China project Plaszyme. This database focuses on curated protein entries related to plastic degradation, integrating sequence, annotation, structure, and degradation phenotype information into a unified resource for research and machine learning. The database serves as both a research reference and a resource for machine learning, supporting the training of models such as PlaszymeAlpha and PlaszymeX. It is also accessible via our WebApp: http://plaszyme.org/plaszymedb, which provides interactive browsing, search, and visualization tools. Key Features 474 protein entries related to plastic degradation 721 experimentally verified enzyme–plastic relationships 34 plastic types (including PET, PE, PS, PU, PLA, Nylon, and members of the PHA family) 82 experimentally resolved protein structures and complete AlphaFold high-confidence predictions Rich metadata per entry, including: • Protein sequence, enzyme name, EC number, taxonomy, host organism • Cross-references: GenBank, UniProt, PDB, RefSeq • Experimental or predicted structures (with manual quality review) • Verified degradation phenotypes across 34 plastic types Manually reviewed dataset ensuring quality and reproducibility