Re: GeneMANIA on gene list of ~2000 genes

[Quaid Morris]

Hi Robert,

Thanks for your email!

We normally suggest that people use the Cytoscape app for larger gene lists: https://apps.cytoscape.org/

https://academic.oup.com/bioinformatics/article/26/22/2927/228740

The Cytoscape app comes with some command line tools, so you can run the analyses offline (and in a batch mode) which might make your comparisons easier. I am not sure if they still being maintained, Gary (cc'ed) will know. The interface is described here: http://pages.genemania.org/command-line-tools/

This paper describing using the command line interface: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4032932/

You have to choose a weighting method. This is the method that assigns weights to different networks. To recreate the automatic method of the website, I think that you just leave this command option blank. We also find that the "bp" method works well when predicting biological function or phenotypes.

It should run much faster than overnight; our experience is that queries take a few minutes at the most. It would help if you sent us the list. One thing that slows website queries down is that they have to download the network viewing app which can be quite large and quite slow for large, 1000 node networks.

Best,

Q

Quaid Morris, PhD
Professor, The Donnelly Centre, University of Toronto
Departments of Molecular Genetics and Computer Science

Associate Investigator, Ontario Institute for Cancer Research

CIFAR AI Chair, Vector Institute Faculty

http://morrislab.ca
cell:  (416) 220 5796

On Mon, Dec 16, 2019 at 4:27 PM Robert Ietswaart <robert_i...@hms.harvard.edu> wrote:

Dear Prof. Morris,

Thank you for your attention. My name is Robert ietswaart, I am a postdoc in the lab of Prof. Stirling Churchman at Harvard Medical School and I am contacting you with a question on GeneMANIA:

How can I run GeneMANIA on a input gene list with 2000 - 3000 human or mouse genes?

I have tried your webserver but it did not allow me to input more than 1024 genes. It ran fine with the first 1000 of my input genes with an overnight runtime.

We are trying to systematically compare the gene function prediction results of GeneMANIA to GeneWalk, for the revisions of our manuscript. So it would be great if I can get GeneMANIA to work in order to include these comparison results of the different methods in our revised manuscript.

Thank you for your advice on how to best proceed, I look forward to hearing from you.

Kind regards,

Robert ietswaart

Robert Ietswaart

Postdoctoral Research Fellow

Churchman Laboratory

Department of Genetics

Harvard Medical School

77 Avenue Louis Pasteur

New Research Building, Room 356

Boston, MA 02115