FAST-PETase degrades 51 post-consumer PET products in a week

AI system

System: Structure-based machine-learning mutation predictor (ensemble of three 3D convolutional neural networks) applied to PETase engineering
Model: MutCompute.
Release: Exact release not established.
Application developer: University of Texas at Austin (Alper and Ellington groups)
Role: The model selected candidate mutations for enzyme experiments.
Configuration: Not assessed
Inputs and outputs: Not assessed
Autonomy: Not assessed
Workflow: A structure-based machine-learning model trained to predict stabilizing mutations. A wild-type PET hydrolase as the starting scaffold. Untreated post-consumer PET tested at moderate temperatures.
Human oversight: Not assessed
Relation to AI: The machine-learning model chose which mutations to introduce. The designers measured the enzymatic activity experimentally. A separate group at another institution later reproduced it, using FAST-PETase as a component of its own system.
AI contribution: The machine-learning model chose which mutations to introduce. The designers measured the enzymatic activity experimentally. A separate group at another institution later reproduced it, using FAST-PETase as a component of its own system.

Studies: Not assessed
Study design: Not assessed
Measurement: Not assessed
Setting and sample: Not assessed
Measures: degraded 51 post-consumer PET products in a week
Comparison: wild-type PETase and previously engineered PET hydrolases
Results: Result: superior PET-hydrolytic activity relative to both wild-type and engineered alternatives between 30 and 50 °C and a range of pH levels.
Additional value from AI: Not assessed
Limitations: Not assessed
Denominators: Not assessed
Materials for replication: Not assessed

Participants: Developer: University of Texas at Austin (Alper and Ellington groups). Operator: University of Texas at Austin; subsequently Texas A&M University.
Jurisdictions: Benefit: Austin, Texas, United States. Developer: United States. Operator: United States.
Ownership and control: Not assessed
Dependencies: Not assessed
Cross-border connections: Not assessed
Geopolitical implications: Not assessed

Conditions for replication: Moderate reaction temperatures (roughly 30–50 degrees Celsius). Amorphous or thermoformed PET substrate. Enzyme expressed and purified or displayed at sufficient activity.
Replication attempts: 1 follow-up entries. Benefit recurred: 1.
Later findings: 19 March 2025. An independent Texas A&M group adopted FAST-PETase in a yeast whole-cell biocatalyst and reported complete depolymerization of a PET film in seven days.
Availability: Not assessed
Institutional decision: Not assessed
Grounds: Not assessed
Action and result: Not assessed
Success criterion: Not assessed

Assessed benefits: Organizations. Society.
Depth: Organizations: 1. Society: 1.
Reasons: Organizations: A laboratory-scale enzyme that an independent group could pick up and reuse successfully; the sources report bench work, not production deployment. Society: Complete depolymerization of untreated post-consumer PET was shown only in the laboratory in both the original and the independent study. Neither source reports industrial-scale recycling.
Observed or intended: Not assessed
Rubric: Not assessed
Assessor and date: Not assessed
Alternative assessments: Not assessed
Fill: The deployer's account

Kind: Not assessed
Related entries: Not assessed
Field: Enzyme engineering / plastic degradation.
Purpose: Use a structure-based machine-learning model to predict stabilizing mutations that make a PET hydrolase active and robust at moderate temperatures.
Dates: Date basis unassessed: 2022 (precision unassessed). Updated: 21 October 2026.
Stage: Not assessed
Version: SHA-256: d86cb258dc3a…