This page contains a MATLAB implementation of the transcription factor target ranking methodology based on Gaussian process differential equation models.
An R implementation of the same methods is available from the authors upon request.
This package was created by Antti Honkela.
Current release is 0.11.
-
netlab mainly used for optimization utilities (like scg).
-
prior prior distributions.
-
optimi optimization constriant mappings.
-
kern covariance functions.
-
ndlutil various utility functions.
-
mltools machine learning models.
-
gpsim Gaussian process single input motif software.
Minor fix to demPlotMef2Models.m.
This version includes scripts used to run the experiments in the PNAS paper.
The PNAS experiments depend on a Drosophila data set (35 MB), which includes a pre-processed version of the developmental expression time series data of Tomancak et al. as well as various validation data sets. We thank Dr. Tomancak for the kind permission to redistribute the data.
In order to run the demos, you need the above Drosophila data set. The command drosLoadData (also used internally by many of the scripts) will attempt to load it from the current working directory and its subdirectory data.
Models can be fitted for individual genes and the results visualised using the command
>> demPlotModels
>>This will create independent visualisations of model for the three repeated experiments, the last of which are:
The package can also be used to rerun the rankings reported in the paper. This is a time-consiming process that may take several days in serial execution. If desired, these may nevertheless be run using the command
>> demRunRankings
>>Page updated on Thu Apr 15 12:13:57 2010