iACVP

This package is used for prediction of anti-coronavirus peptides (ACVPs).

Environment

anaconda 4.11.0
python 3.8.8
pytorch 1.9.0
scikit-learn 1.0.2
gensim 4.0.1

Execution

1 Setting iACVP directory

The directory structure given in the github should be conserved.

2 Construction of training and test datasets

./
$sh dataconst.sh
(train_division.py, test_fasta.py)

3 Word2vec model construction

./program/w2v
$sh w2vconst.sh
(word2vec_acvp.py)

NOTE
Use epochs=20 for prediction of AVPs
Use epochs=4 for prediction of ACVPs

4 Training and testing of W2V-based and BE-based classifiers

4-1 Simulation

./program/network
$sh main.sh
(deep_train_test.py, machine_train_test.py)

4-2 Analysis (calculation of measures)

./program
$python analysis_net.py 　

5. Training and testing of RF and SVM classifiers with conventional encodings

5-1 Simulation

./program/ml
$sh submain.sh
(conv_train_test.py)

5-2 Analysis (calculation of measures)

./program
$python analysis_ml.py