PHsymm Server
About PHsymm
PHsymm is a single-sequence–based deep learning method for predicting the symmetry of homologous protein complexes. PHsymm is the first to apply persistent path homology to extract topological features from protein sequences. The model performs end-to-end symmetry prediction directly from sequence data, without relying on structural information or multiple sequence alignment.
Contact
If you have any questions during operation, please contact us:
- Email: zhaijq93@ruc.edu.cn
- Lab: Lab of Bioinformatics and Machine Learning
Usage Instructions
- Step 1: Select the FASTA file containing the protein sequence(s) to be predicted, and click "Choose FASTA file" to upload it.
- Step 2: Click the "Submit Files" button to upload the FASTA file.
- Step 3: Click "Run Analysis" to perform symmetry prediction using the PHsymm model.
- Step 4: If the interface shows no change for an extended period, click the "Refresh" button to reload the page.
- Step 5: Download the compressed results file.
Source Code
Download the complete PHsymm implementation
Code Components
Click the “Download Source Code” button below to download the compressed folder. After decompressing the folder, it contains all the code for the PHsymm method as well as the protein data used. Detailed instructions for usage can be found in the included README file.
Submit Protein Sequence
Upload the protein sequence to be predicted
Upload Status
Run PHsymm Analysis
Execute PHsymm prediction on uploaded sequences
PHsymm Results
Please upload files and click 'Run Analysis' to start prediction.