RAMA - Robust Analysis of Microarray
Rama Initialization Dialog
Rama
Rama is an intensity normalization algorithm based on hierarchical Bayesian modeling.
Rama has been designed to work with raw intensities ( Affymetrix or 2 color ), especially in
dye-swap experimental designs. It estimates intensities and as such requires there to be at
least 2 data points for each color label.
Experiment Labeling
Each slide (experiment or hyb) loaded into MeV is listed here. You must tell Rama how you've
labeled your samples.
If you've loaded 2 color data, you are asked to indicate which color ( Cy3 or Cy5 ) dye was
used on your control samples. It is assumed that the other was used for the treated sample
on that slide. Rama will infer from the labeling whether or not your experiment was dye-swapped
based on your input here. Keep in mind that Rama requires at least 2 of any color data.
Advanced Parameters
# Iterations ( B ) - the number of iterations used by Rama
Burn In Period ( num.iter ) - the length of the burn-in period in the MCMC algorithm. This
should be less than # Iterations ( B )
Rserve
This implementation has been built allow the user to connect to any Rserve daemon. Upon
load, the pull down menu is populated with the connection strings from the config file
( config/tmev.cfg ). The user can add a new connection string here if desired.
The format is: ipaddress:port.
For example, if the server running Rserve is at IP address 192.168.0.1 and
Rserve is running on its default port ( 6311 ), the connection string would be 192.168.0.1:6311.
The user can type that into the text field, click add, make sure it is selected in the pull down
menu and then run RAMA. If the connection was made without any errors, the new connection
string will be added automatically to the config file. Of course, the user always has the option
of manually adding/editing the connection strings in the config file.