Accelerating the discovery of industrial PET-degrading enzymes: Evolving paradigms from bioprospecting to computational discovery.

AI-assisted and structure-guided methods now enable identification of PET-degrading enzymes with industrial relevance, surpassing traditional bioprospecting approaches.

• Why it matters: Efficient enzymatic depolymerization of PET is hindered by substrate heterogeneity, enzyme stability, and activity loss, limiting practical recycling solutions and environmental impact.
• What they did: The review compares phenotype-driven bioprospecting, sequence mining, structure-guided searches, and AI prioritization, emphasizing throughput, novelty, and industrial fitness, with a focus on hierarchical validation.
• The result: Integrating standardized substrate characterization, kinetic data, and open data sharing can improve enzyme selection, enabling the development of robust, application-ready PET-degrading biocatalysts for circular plastic use.