This script will query the KnetMiner Knowledge Graph with a user provided gene list (one gene ID per line) and a user provided keyword file (one keyword per line). It returns the knetscore, genome location, and KnetMiner genepage URL for each user gene that can be associated with any of the keywords. The script was tested with upto 3000 wheat genes and 10 keywords. It works for various species including Wheat, Rice, Arabidopsis, F.graminearum, Z.tritici and more.

The script requires Python3+ with following dependencies (last test with py3.8.2):

• pandas

• numpy

• argparse

• requests

• resource (included in Linux/Unix but needed for Windows conda)

• Python virtual environments, e.g. pyvenv for python3. If the user does not have root permission on Easybuild a virtual environment is required for installation of the dependencies through pip.

• The script uses the KnetMiner REST API and therefore requires an internet connection.

Clone or download this repository using the green "Clone or Download" button. To clone this repository via git (command-line), simple execute the following:

git clone https://github.com/Rothamsted/genelist-api.git

You'll find all the relevant files in the genelist-api folder, cloned to whatever directory you cloned it in.

Login to a compute node:

srun --pty bash -i

Check available versions of python:

module avail Python
module load <Python3 version>

A virtualenv is required if you lack permissions to pip install

virtualenv <name of Python virtual environment>
source </path to env>/bin/activate/

Install missing python packages:

pip install pandas
pip install numpy
pip install argparse
pip install requests

Change to the cloned/downloaded script folder. To see the script help page run:

python3 genepage_insight.py -h

Supported arguments are:

• -g OR --gene Text file which contains your list of gene ids or names (one per line)
• -k OR --keywords Text file which contains the search terms or keywords of interest to you (one er line).
• -s OR --species Currently spporting rice, wheat, or arabidopsis (ara)
• -o OR -output Output directory. If npt provided, a file will be created using your gene file name & appending '_output' to it, where your results & dependent files will be found.

Example command:

python3 genepage-insight.py -g example_list.txt -k keywords_heat.txt -s wheat -o /home/$USER/test_output.txt

The output will be a tabular text file containing 5 columns: Gene ID, Knetscore, Chromosome, Gene start and Network URL.

The Knetscore indicates the relevance of a gene to the provided keywords as described in Hassani-Pak (2017), PhD thesis.

The URL links to an interactive knowledge network of that gene with links to publications, ontologies, pathways etc that contain the keywords.

Keywan-Hassani Pak

Colin Li

Joseph Hearnshaw