DeepKla

an attention mechanism-based deep neural network for protein lysine lactylation site prediction in Rice (Oryza sativa)

DeepKla provides a deep-learning method for lactylation site prediction. It is implemented by deep learning library Keras and Tensorflow backend. At present, DeepKla only provides prediction of rice lactylation sites; however, it also provides customized model training that enables users to train other PTM prediction models by using their own training data sets.

Installation Download DeepKla by git clone https://github.com/linDing-group/DeepKla

Installation has been tested in Mac OS X with Python 2.7.

Since the package is written in python 2.7, python 2.7 with the pip tool must be installed first. DeepKla uses the following dependencies: numpy, h5py, keras version=2.0.6

You can install these packages first, by the following commands:

pip install numpy

pip install h5py

pip install -v keras==2.0.6

pip install tensorflow

Predict on your own data : cd to the predict folder which contains load_model.py, run:

python load_model.py <your_test_file>.fa <predict_result>.txt

Example:

python load_model.py ../data/fungiForTest.fa fungiForTest_predict_result.txt fungiForTest.pkl

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

For advanced users who want to perform training by using their own data:

For training:

python ../SEL_CNN_BiGRU_Attention.py (Computational time: 6453.411189s)

Note: You need to change fastafile, modelfile, pklfile, and model_save's name on your own.

Hao Lv, Fu-Ying Dao, Hao Lin. 2022. DeepKla: An attention mechanism-based deep neural network for protein lysine lactylation site prediction. iMeta 1: e11. https://doi.org/10.1002/imt2.11