Project Management & Overview
| Project Name | Submission Time | Sequence Count | Status | Actions |
|---|
| Project Name | Submission Time | Sequence Count | Status | Actions |
|---|---|---|---|---|
| Marine Microbial Sample | 2025-08-28 14:30 | 5 | Complete | |
| Soil Sample Test | 2025-08-27 09:15 | 12 | Complete |
3D Structure Viewer
Interactive protein structure will be displayed here
Due to uncontrollable network fluctuations during data packet parsing on the MSA server, ColabFold is temporarily unable to return structural information.
| Sequence ID | Length | Sequence Fragment |
|---|
| Sequence ID | Plastic Type | Degradation Probability | Confidence |
|---|
Plaszyme is an online platform dedicated to predicting plastic-degrading enzymes. Utilizing advanced machine learning models (including algorithms trained on sequence patterns) and cutting-edge structure-based deep learning models, the platform analyzes protein sequences to predict their ability to degrade various types of plastics. This platform aims to provide convenient tools for biotechnology researchers, environmental scientists, and industry professionals to accelerate plastic degradation research and applications.
Our prediction capability is robust: the models are trained on extensive experimentally validated plastic-degrading enzyme data (PlaszymeDB), accurately identifying both sequence patterns and structural features critical for plastic degradation.
The platform now primarily offers prediction functions through PlaszymeAlpha and PlaszymeX. Users have the flexibility to select their preferred model based on the task mode or utilize the metagenomic prediction mode to process environmental samples. All underlying models are rigorously trained on PlaszymeDB data.
Before preparing your data, first select the appropriate analysis path based on your research goals:
Standard Prediction Mode (PlaszymeAlpha / PlaszymeX): For analyzing individual or small batches of protein sequences. You must choose one of the core models (PlaszymeAlpha or PlaszymeX) based on your preference for speed or in-depth analysis (e.g., structure-based prediction).
Metagenomic Prediction Mode: For processing large environmental or uncurated protein sequence datasets. This mode utilizes specialized algorithms optimized for high-throughput, environmental screening.
Prepare the protein sequences you want to analyze. Ensure sequences are properly formatted with unique identifiers.
Format Supported: Supports single sequence text input or batch sequences in FASTA format files (.fasta, .fa, .txt).
Navigate to the "Predict" page.
Enter an optional project name.
Paste your sequences in the text box or drag and drop your FASTA files into the upload area.
Confirm your chosen Prediction Mode and Model selection (from Step 1) on the submission interface.
Click the "Submit Prediction" button to start the analysis.
The system will start processing your sequences. Processing time depends on the chosen model (Standard vs. Metagenomic Mode) and the number/length of the sequences. After processing is complete, a "View Report" button will appear.
On the report page, you can view the detailed prediction results.
Data Display: Results show each sequence's degradation probability against different plastic types, displayed in intuitive progress bars with confidence annotations.
Export Options: You can export the full results in CSV or PDF format for further analysis or reporting.