Graphage

Graphage, a phage analysis tool that incorporates phage and ICE discrimination, phage integration site prediction, phage lifestyle prediction, and phage host prediction. Graphage utilizes a Gapped Pattern Graph Convolutional Network (GP-GCN) framework for phage representation learning. The GP-GCN framework is available for various sequence analysis at Github.

Prerequisite

The scripts are written in Python3. Following Python packages should be installed:

• cython
• docopt
• numpy
• Biopython
• pytorch 1.9.0
• pytorch_geometric 1.7.0

For phage lifestyle prediction task, please also install Glimmer and Hmmer.

• Glimmer: conda install -c bioconda glimmer
• Hmmer: conda install -c bioconda hmmer

Install

git clone https://github.com/deepomicslab/Graphage.git
cd Graphage/scripts
python setup.py build_ext --inplace

Please download the model dictionary from Zenodo and put it under the scripts dictionary. We also provide the sequences on the phage and bacterial integration site. Other data used in the manuscript can be obtained from public datasets.

Data preparation

For phage integration site prediction, phage lifestyle prediction, and phage host prediction, only fasta file is needed (you can use fasta file with multiple sequences). For phage and ICE discrimination, please provide fasta file and gff file. Please check our example data for details.

Usage

Phage and ICE discrimination

cd Graphage/scripts/
python Graphage_ice.py phage_genomes.fasta phage_gff_dic --out output.txt --batch 100 --thread 10

• phage_genomes.fasta contains phage genome sequences in fasta format. Graphage supports both single genome and multiple genomes in one file.
• phage_gff_dic contains phage annotations in gff format. Each phage should have one gff file under the dictionary, with the same name as the identity in the fasta file (after >).
• The input of --out is the filename of the output file (default: Graphage_ice_output.txt).
• The input of --thread is the number of worker processes to use for genome encoding (default:10).
• The input of --batch is the batch size for the prediction process (default:100).

For example:

python Graphage_ice.py ../example_data/example.fasta ../example_data/gff_data

There are two columns in the output, with the first showing the sequence id and the second showing the predicted type.

For more information, please use the command:

python Graphage_ice.py -h

Phage integration site prediction

cd Graphage/scripts/
python Graphage_int.py phage_genomes.fasta seq_type --out output.txt --batch 100 --thre
ad 10

• phage_genomes.fasta contains phage genome sequences in fasta format. Graphage supports both single genome and multiple genomes in one file.
• seq_type should be phage or bacterial, indicating whether your sequences are phage sequences or bactrial sequences.
• The input of --out is the filename of the output file (default: Graphage_int_output.txt).
• The input of --thread is the number of worker processes to use for genome encoding (default:10).
• The input of --batch is the batch size for the prediction process (default:100).

For example:

python Graphage_int.py ../example_data/example.fasta phage

There are three columns in the output, with the first showing the sequence id, the second showing the location on the sequence, and the third showing the prediction (0 for non-integration-site and 1 for integration-site).

For more information, please use the command:

python Graphage_int.py -h

Phage lifestyle prediction

cd Graphage/scripts/
python Graphage_lifestyle.py phage_genomes.fasta --out output.txt --batch 100 --thread 10

• phage_genomes.fasta contains phage genome sequences in fasta format. Graphage supports both single genome and multiple genomes in one file.
• The input of --out is the filename of the output file (default: Graphage_lifestyle_output.txt).
• The input of --thread is the number of worker processes to use for genome encoding (default:10).
• The input of --batch is the batch size for the prediction process (default:100).

For example:

python Graphage_lifestyle.py ../example_data/example.fasta

There are two columns in the output, with the first showing the sequence id and the second showing the predicted lifestyle.

For more information, please use the command:

python Graphage_lifestyle.py -h

Phage host prediction

cd Graphage/scripts/
python Graphage_host.py phage_genomes.fasta --out output.txt --batch 100 --thread 10

• phage_genomes.fasta contains phage genome sequences in fasta format. Graphage supports both single genome and multiple genomes in one file.
• The input of --out is the filename of the output file (default: Graphage_host_output.txt).
• The input of --thread is the number of worker processes to use for genome encoding (default:10).
• The input of --batch is the batch size for the prediction process (default:100).

For example:

python Graphage_host.py ../example_data/example.fasta 

There are two columns in the output, with the first showing the sequence id and the second showing the predicted host.

For more information, please use the command:

python Graphage_host.py -h

Maintainer

WANG Ruohan ruohawang2-c@my.cityu.edu.hk

Reference

Wang, R. H., Ng, Y. K., Zhang, X., Wang, J., & Li, S. C. (2024). Coding genomes with gapped pattern graph convolutional network. Bioinformatics, 40(4), btae188.