2009-02-27 15:49:46 UTC
Hi Luke,
I'm John Braisted, an project admin for the SourceForge MeV project and I work at the JCVI/PFGRC on tool development and I generate EASE file systems that correspond to our microarrays and annotation files.
The A. fumigatus array version 2 does have an EASE file system and does contain GO Biological Process, GO Cellular Component, and GO Molecular Function annotations. These annotations link directly to the loci (oligo targets) specified in the array's annotation file. Since you don't need to convert from one accession type to another, there are no converter files needed (these are just used for accession conversions like). The 'Class' folder contains the key files.
Here's a description of the ease file archives from the PFGRC:
http://pfgrc.jcvi.org/index.php/microarray/ease_archive.html
This also describes what to do with the files. The manual describes the folder structure in detail.
Here's a link to the page that will provide the .ann file and the EASE file archive for your array.
http://pfgrc.jcvi.org/index.php/microarray/array_description/aspergillus_fumigatus/version2.html
**Please email me at
mev@jcvi.org and I can provide further assistance regarding running ease.
Here's an overview:
*Download and unzip the ease archive.
*Replace the sample 'ease' folder in the 'data' directory in the MeV folder with the 'ease' file from the fumigatus ease zip. This eliminates the need to point EASE toward your ease directory.
*Create a cluster through a stat or clustering method.
*Right click in a cluster viewer and select 'Store Cluster'. Any stored cluster can be analyzed by EASE.
*Click on EASE.
*Select to specify the population of genes as all loaded in MeV.
*Select the cluster from the table in the lower part of the first page.
*On the annotation page, select 'Locus' as your gene identifier.
*Don't use a converter file (mapping one accession type to another, it's not needed)
*Select all three GO files in the lower section.
*The Stats pane contains options such as p-value corrections. We usually use Fisher Exact as the reported statistic.
*Then just kick off the process.
You wrote earlier that you can't retrieve EASE files from PFGRC through the EASE interface. Only DFCI's Harvard repository is represented in the interface. You can get PFGRC ease annotation archives from the PFGRC website.
Please contact me if you need further assistance (
mev@jcvi.org)
John Braisted