DeepCapTail is deep learning package that predicts capsid and tail proteins of phage genomes. Basically, the input for this package is a protein sequence, and the output is the propabality that the given sequence is capsid or tail protein. This package is entirely written in Python.
This project should run on all Unix platform, although it was only tested on Ubuntu 16.04.4 LTS
DeepCapTail requires the following packages:
- Python - version 3.6 or later
- Biopython - version 1.7 or later
- keras - version 2.1.6 or later
- Clone this repository
- % cd DeepCapTail/
- python predict_sequence.py --p_fasta protein.fasta --capsid_tail capsid --p_output prediction.csv
where 'protein.fasta': fasta file of the sequences that we want to predict. 'capsid ' to predict capsid probability and 'tail' to predict tail probability. 'prediction.csv': csv file that will include the prediction of probabilty of the sequence being capsid or tail.
% python predict_sequence.py --help
The output file is a csv file that has two columns. The first columns is the id of the protein sequence and the second column is the probabilty of the sequence being capsid or tail.