a novel probabilistic model to account for confounding factors in eQTL studies.
Hidden confounding factors, such as unobserved covariates or unknown subtle environmental perturbations can create spurious false associations or mask real genetic association signals. In contrast to previous methods, PANAMA learns hidden factors jointly with the effect of prominent genetic regulators. As a result, this new model can more accurately distinguish true genetic association signals from confounding variation.
N. Fusi,O. Stegle and N. D. Lawrence, "Joint modelling of confounding factors and prominent genetic regulators provides increased accuracy in genetical genomics studies", PLoS Computational Biology, 2012
There are two ways of installing PANAMA.
PANAMA includes a script that lets you access the basic functionalities without too much pain.
After installing PANAMA, from a terminal, run:
$ panama expression_data.csv snp_data.csv
Where expression_data.csv is a Comma Separated Values file containing the gene expression data in the following format
sample1, sample2, ..., sampleN gene1 value , value , ..., value gene2 value , value , ..., value ... geneN
and snp_data.csv contains the SNP data encoded as [0,1,2] in the following format
sample1, sample2, ..., sampleN SNP1 value , value , ..., value SNP2 value , value , ..., value ... SNPQIf you need to load data in another format, just contact us.
ipcluster startfrom the command line.
ipcluster local -n num_processes. PANAMA automatically detects the Ipython version, so no further input is necessary. Read the Ipython documentation for more informations on how to run on networked machines, on clusters and on Amazon EC2
Some of the features of PANAMA can't be accessed using the simple
panama script provided or need further setup. Here's a list of features that have been implemented and can be accessed through the python API:
If you want to call PANAMA directly from python, just take a look at the
PANAMA function in
panama.core.run. The expected arguments are:
If you don't want to hack your way around python, just contact us. We can make some functionalities available from the main script.
If you want to use PANAMA for non-commercial purposes, the code is available under a modified GPL v2 license. The modifications are: