System Introduction
This system is based on deep learning models to predict whether proteins are acidophilic. It uses the ESMC model for sequence encoding combined with PS-MoE(Parameter-Sharing Mixture-of-Experts) architecture for classification prediction.
Usage
- Single Sequence Prediction: Enter a single protein sequence for prediction
- Batch Prediction: Supports multiple sequence prediction in FASTA format, via text input or file upload
Input Requirements
- Supports only standard 20 amino acids:
ACDEFGHIKLMNPQRSTVWY
- Sequence length: 100-2000 amino acids
- FASTA format: Header line starting with
>, followed by sequence
Result Explanation
- Prediction Result: Acidophilic or Non-acidophilic
- Confidence: Model's confidence in the prediction (0-1)
- Probability Distribution: Prediction probabilities for both categories
Example Sequences
Acidophilic Protein Example:
MKALIVLGLVLLSVTVQGKVFERCELARTLKRLGMDGYRGISLANWMCLAKWESGYNTRATE
Non-acidophilic Protein Example:
MVLSEGEWQLVLHVWAKVEADVAGHGQDILIRLFKSHPETLEKFDRFKHLKTEAEMKASED