Detection and classification of pre-miRNA based on deep learning
Training
- cross-validation for human datasets
python cv_human.py
- cross-validation for cross-species data sets
python cv_whole.py
- testing on human datasets
python main_human.py
Usage
python isMiRNA.py -s <RNAsequence> -m <model_path> -f <feature>
for example: python isMiRNA.py -s AGAAUUCUCUUAUCCAACAUCAACAUCUUGGUCAGAUUUGAACUCUUCAA -m model/human/model_cv_human_3.pkl -f feature_data/featureSelectedModel_cv_human_3.pkl
python isMiRNA.py -i <input file> -o <output file>
-
input file: fasta format
-
output file: txt format
-
model_path: the path of trained model
-
feature：Algorithm model for feature extraction
Dependencies
- Python >= 3.6
- RNAFold
- Pytorch
- Numpy
File
- data.zip
- Datasets containing humans and cross-species
- feature_data.zip
- The sequence feature set used in this experimental method, File names with "whole" represent cross-species, and those with "human" represent human.
- model
- The path of trained model