Skip to content

Latest commit

 

History

History
28 lines (17 loc) · 1.12 KB

README.md

File metadata and controls

28 lines (17 loc) · 1.12 KB
Raw

Combined Embedding Model for MiRNA-Disease Association Prediction（CEMDA）

Implemented evironment

Python>=3.6

###Required libraries numpy,numba,openpyxl,xlrd，torch,itertools,sys,os,importlib

We recommended that you could install Anaconda to meet these requirements

How to run CEMDA?

####Data All datas or mid results are orgnized in DATA fold, which contains miRNA-disease associations,disease semantic similarity, miRNA functional similarity, encode result of disease and miRNA.

####The starting point for running CEMDA is:

(1)meta_path_instance.py：gereating meta-paths from the dataset of miRNA-disease associations,disease semantic similarity, miRNA functional similarity. all the result is saved in the folds named "5.mid result" and "6.meta path", which need to be created by yourselves.

(2)CEMDA.py: training the model of PESLMDA which will referece GRU.py, MLP.py,SelfAttention.py. And it outputs the parameter of CEMDA

####other relateive files: GRU.py: a GRU model in CEMDA MLP.py: a MLP model in CEMDA SelfAttention.py:SelfAttention model in CEMDA