DeepMPSF

The source code, training and test datasets of paper 'DeepMPSF: A Deep Learning Network for Predicting General Protein Phosphorylation Sites Based on Multiple Protein Sequence Features'

Install dependency:

• conda create -n phros python == 3.8
• conda activate phros
• pip install --no-cache-dir torch==1.7.0+cu110 -f https://download.pytorch.org/whl/torch_stable.html
• pip install pandas
• pip install scikit-learn

Prerequisites:

• python: 3.8
• CUDA: 10.1
• pytorch: -1.7.0

All relevant inputs：

Due to the limitation of Github, some inputs larger than 25MB are not uploaded. Please contact me directly at 20215227108@stu.suda.edu.cn.

Reproduce experimental results:

The prediction of phosphorylation sites by DeepMPSF includes the prediction of S/T and Y. Therefore, there are corresponding datasets and features for S/T and Y, respectively. Here is the example of S/T. And for the prediction or training of Y, you can replace '_ST' with '_Y' in My_Model.py and the corresponding training/prediction script.

Dataset

D_human:Table_All_train_general_ST.csv/Table_All_train_general_Y.csv
T_human:Table_All_test_general_ST.csv/Table_All_test_general_Y.csv
T_homo:Table_blind_human_ST_2k.csv/Table_blind_human_Y_2k.csv T_mus:Table_blind_musculus_ST_2k.csv/Table_blind_musculus_Y_2k.csv
T_rattus:Table_blind_Rattus_ST_2k.csv/Table_blind_Rattus_Y_2k.csv

Test on T_human

Step 1: cd code
Step 2: python predic_my_model.py

Test on T_homo

Step 1: cd code
Step 2: python predic_my_model_blind_homo.py

Test on T_mus

Step 1: cd code
Step 2: python predic_my_model_blind_musculus.py

Test on T_rattus

Step 1: cd code
Step 2: python predic_my_model_blind_Rattus.py

Train your own model DeepMPSF

Step 1: cd code
Step 2: python train_my_model.py