pathogenie

This is a desktop and command line program for annotating draft bacterial and viral genomes. It may also be used for quickly detecting arbitrary sequences such as antibiotic resistance genes (AMR) proteins in nucleotide sequences. It uses Blast to find hits to known gene sequences from sequence databases. The inputs are fasta files. Annotation is performed in a similar manner to Prokka and first requires an assembled genome if you have sequenced reads. The program is written in Python. Currently it is available as a graphical desktop application. A command line tool will also be added. You can also use it from inside Python scripts.

Usage

The graphical application can be launched from the terminal using:

pathogenie-gui

From the GUI you may load fasta files into a table and then run genome annotation or gene finding with custom databases. This program utilises the sequence databases for gene finding compiled by abricate:

card
resfinder
arg-annot
resfinder
ncbi
ecoh

The GUI layout is shown below:

Using in python

Run an annotation on a fasta file like a set of contigs:

import pathogenie
featdf,recs = pathogenie.app.run_annotation(filename, threads=10, kingdom='bacteria')
#save to genbank
pathogenie.tools.recs_to_genbank(recs, gbfile)

Installation

All operating systems with Python (>=3.6 required) installed:

pip install -e git+https://github.com/dmnfarrell/pathogenie.git#egg=pathogenie

Dependencies

You require ncbi-blast+ tools and clustalw for basic gene finding. The following programs are used for genome annotation:

prodigal
hmmer3
aragorn

Linux

The pip command should work fine but you can also install the snap using:

sudo snap install pathogenie

Or download the AppImage. Download from the latest release page and run as follows:

chmod +x pathogenie-0.4.0-x86_64.AppImage
./pathogenie-0.4.0-x86_64.AppImage

The external binaries can all be installed on Debian/Ubuntu based systems using:

sudo apt install ncbi-blast+ clustal prodigal aragorn hmmer

These are NOT needed if you are using the snap or AppImage.

Windows

The external executables will be downloaded for you when you first launch the program.

Blast also requires the Visual Studio 2015 C++ redistributable runtime package: https://www.microsoft.com/en-us/download/details.aspx?id=48145

OSX

Not yet tested but may work if you can install the dependencies. You can probably install them with bioconda.

Use from python

Run an annotation and save the results:

import pathogenie
for file in fastafiles:  
    name = os.path.basename(file)
    out = name+'.gb'    
    res,recs = pathogenie.run_annotation(file, threads=10)
    #save genbank
    pathogenie.recs_to_genbank(recs, out)
    #save gff
    pathogenie.recs_to_gff(recs, name+'.gff')

dmnfarrell/pathogenie