MicroRNA
Expression Profiles Classify Human Cancers
Nature 435, 834-838, June 9th, 2005
Frequently
Asked Questions
Thank
you for your interest in our work!
This
page contains answers to frequently asked questions on the above
referenced paper.
This page is constantly being
updated.
NOTE: Supplementary files on
Nature website are from an earlier
version.
Please refer to
this website for the most updated information.
Questions
1.
Where can I find the expression data?
2.
What are the formats of the expression data?
3.
I want to run the same analyses as described in the paper. How can I do
it?
Answers
1.
Where can I find the expression data?
Expression data are
available at two locations.
MicroRNA expression data
have been depositted into GEO (http://www.ncbi.nlm.nih.gov/geo/),
with a serial accession number of GSE2564.
MicroRNA and mRNA
expression data are also available at Broad Institute cancer program
website (http://www.broad.mit.edu/cancer/pub/miGCM).
2.
What are the formats of the expression data?
Expression data at the
Broad cancer website is in the gct format. This is a tab-delimitted
text file format. For a detailed explanation of this file format,
please refer to the GenePattern tutorial (at http://www.broad.mit.edu/cancer/software/genepattern/tutorial/gp_tutorial_fileformats.html)
3.
I want to run the same analyses as described in the paper. How can I do
it?
We generated GenePattern
modules and pipelines that reproduce the results presented in the
paper. In order to run these pipelines please perform the following
steps:
- Download GenePattern 1.4.1 (or later) from http://www.broad.mit.edu/cancer/software/genepattern/
(GenePattern is a free analysis software produced by the Broad
Institute's cancer program).
- Install all the modules and pipelines in the last step of the
installation. These include the pipelines and modules attached to this
paper.
- There are 7 pipelines for this paper, all start with
Lu.Getz.Miska.Nature.June.2005:
- Lu.Getz.Miska.Nature.June.2005.clustering.miGCM218.pipline
reproduces Figure 2a of the paper.
- Lu.Getz.Miska.Nature.June.2005.clustering.ALL.pipline
reproduces Figure 2b of the paper.
- Lu.Getz.Miska.Nature.June.2005.clustering.ep.miRNA.pipeline
reproduces the upper dendrogram of Figure 2c.
- Lu.Getz.Miska.Nature.June.2005.clustering.ep.mRNA.pipeline
reproduces the lower dendrogram of Figure 2c.
- Lu.Getz.Miska.Nature.June.2005.mouse.lung.pipeline
reproduces the results of Figures 3a and 3b.
- Lu.Getz.Miska.Nature.June.2005.PDT.miRNA.pipeline
reproduces the results of predicting the poorly-differentiated tumors
using miRNAs.
- Lu.Getz.Miska.Nature.June.2005.PDT.mRNA.pipeline
reproduces the results of predicting the poorly-differentiated tumors
using mRNAs.
- Note that apart from
Lu.Getz.Miska.Nature.June.2005.mouse.lung.pipeline all other pipelines
run only on Windows and include an installation of MATLAB Component
Runtime libraries.
The installaion of the MATLAB Component Runtime libraries is performed
only once and requires administrator rights.
If you have problems installing or running the software, please contact
gp-help@broad.mit.edu.
Update Logs
05/10/2005, created, by Jun Lu
06/08/2005, modified, by Jun Lu
06/17/2005, added GenePattern modules in Q3, by Gad
Getz