SAPPHIRE: a stacking-based ensemble learning framework for accurate prediction of thermophilic proteins.

The packages that this program depends on is scikit-learn==0.24.1 or higher. You can run following command in terminal.
pip install scikit-learn==0.24.1

1. Copy your fasta file into ./input and change the name to seq.fasta
   
2. Extract PSSM features from your fasta file using pssmpro (https://github.com/deeprob/pssmpro). Please follow their tutorial carefully.
   
3. After finish feature extraction, copy the required PSSM features for SAPPHIRE into the ./input folder. These required PSSM features are listed as the following.

• psssm_composition.csv
• rpm_pssm.csv
• s_fpssm.csv

4. Run command
   python sapphire.py
5. The result including sequence Header, label and probability will be saved in ./output/predicted_result.csv