The University of Sheffield
Neil Lawrence ML@SITraN
GPy and Bioconductor on Gene Expression Data

Running GPy with Biological Data

These instructions relate to machines in the EBI teaching room with Biolinux installed. For other linux machines instructions should be similar.

On the BioLinux machines in the EBI lab you should be able to install GPy by the following instructions. In a terminal at the prompt write.

git clone https://github.com/SheffieldML/GPy.git
cd GPy
git checkout devel

This checks out the latest version of the code. Normally you would use the master branch, but we are currently in the middle of developing a major new release, some of the features are used in the example, for that reason we've switched here to the devel branch, which is very much an Alpha release.

Now we start the IPython notebook and change to the GPy directory we've created, so that the code can be found.

ipython notebook

Change directory to the 'GPy' directory so that the GPy module will naturally be in the path. Then open a new notebook and download the notebook for processing the Tomancak fruitfly data by pasting in the following commands into the notebook.

import urllib
urllib.urlretrieve('http://staffwww.dcs.sheffield.ac.uk/people/N.Lawrence/ebi14/TomancakDataWithGPy.ipynb', 'Tomancak Data with GPy.ipynb')

Now go back to the main notebook page and click on the notebook you've just downloaded.

For the single cell example, to run the parts that make use of locally linear embedding and isomap, you will also need to install scikit-learn. From a terminal window write:

sudo pip install scikit-learn

You can then download the relevant notebook with the following commands:

import urllib
urllib.urlretrieve('http://staffwww.dcs.sheffield.ac.uk/people/N.Lawrence/ebi14/SingleCellDataWithGPy.ipynb', 'Single Cell Data with GPy.ipynb')

Finally Mu presented our work on modelling the Drosophila gap gene network with a latent force model based on a temporo-spatial partial differential equation. The notebook to recreate these results is available here.

import urllib
urllib.urlretrieve('http://staffwww.dcs.sheffield.ac.uk/people/N.Lawrence/ebi14/SpatioTemporalGeneProteinDrosophila.ipynb', 'Spatio-temporal Gene and Protein Data from Drosophila Melanogaster.ipynb')

This document last modified Wednesday, 14-May-2014 11:14:41 UTC