DeepMethy: Prediction of Protein Methylation Sites with Deep Learning
Developer: XinCheng from Data Science and Big Data Technology, College of Software, Jilin University
keras==2.0.0
numpy>=1.8.0
backend==tensorflow
The related data is stored in '/dataset/test_file.csv'.
The input file is an csv file, which includes , postion, sequences and labels
To use the model for prediction on your test data, prepare a CSV file with two columns: position and sequence.
You can then run predict.py to generate predictions. The output will be a text file with results in the format:
"21" "0.9999963" "21" "0.95067513" "21" "1.0669616e-24" "21" "3.7860446e-30" "21" "0.72186846" "21" "1.16561736e-07" "21" "1.8712221e-07" "21" "1.2668259e-24"
You can modify parameters in the predict.py main function to customize the prediction process according to your needs.
If you want to train your own network,your input file is an csv fie, while contains 3 columns: label, postion, sequence label is 1 or 0 represents methylation and non-methylation site.
If you are interested in adding more function to the code, welcome to show your talent!
The methods folder contains train_BP(length-concat).py, predict_BP(length-concat).py.
The detailed model structure is shown in the train_BP(length-concat).py
Please feel free to contact us if you need any help: [email protected]