PEPR Tutorial

From the home page:

http://pepr.cnmcresearch.org/

You will first need to register. Registration is free, and we do not release any information to any third party. Registration is simply to provide user demographics to our funding agencies.

If you are a project originator, then specific rights will be assigned to you, so that you can enter or modify meta data for your project, manage public release to PEPR and GEO, and other management functions. Contact the Eric Hoffman ( ehoffman@cnmcresearch.org ) or the PEPR designer/administrator Josephine Chen ( jchen@cnmcresearch.org ) if you have any questions or concerns regarding registration and project management.

Once you have logged in, you will be given access to the PEPR database, including visualization tools and links to GEO.

Here, we give an example of dynamic queries of a human longitudinal data set (each human volunteer had four muscle biopsies during an aerobic exercise training intervention; William Kraus' STRRIDE study).

- left link “search repository”

- search “muscle” “project”

- choose “WKraus STRRIDE Study”

Go to “Analysis” link in middle right

There are both “Chart” and “GEO Clustering Tool” links; we are putting the GEO on as they are completed by GEO staff; at the time of this tutorial, this project does not yet have GEO cluster implementations (although the data is in GEO)

Choose “Chart”

Choose U133A chips

“select all” arrays

You then have a series of options for visualization, as shown:

This particular study is a longitudinal time series, with four muscle biopsies per individual.

One can graph any of the shown options. If you use the options shown above (querying for GOT1 gene), you will get the average for males and females separately, with all replicates shown. Submitting this will give you a drop down list of all probe sets matching your query (one in this instance). Select this, and you get:

From this, it is clear that males respond at time point 1 much more than females.

If you would like to look at longitudinal data individually for each subject, just choose “Time” and “Subject Number” (instead of time and sex), and you will get:

Experiment with the different meta data visualizations, and with different projects. This project is longitudinal in design, but there are many cross-sectional projects on PEPR as well.

If there is a particular project that a PEPR/chart or GEO visualization not let loaded, let us know and we'll prioritize it.

Eric Hoffman
Josephine Chen