Dear Anisha, If you do not specify knownTargets, GPRankTargets() will just use the TF mRNA profile to fit the model of unobserved TF protein profile and target mRNA profile. If you specify for example knownTargets=c('a', 'b', 'c'), GPRankTargets() will first fit a model to targets a, b, c to infer the TF protein profile. It will then augment this model with each candidate target one at a time, while keeping the parameters related to a, b, c fixed to essentially fix the protein profile. The use of TF mRNA profile in this mode is optional - it can help if it is informative about the active TF protein concentration, but it should not be used if the TF activity is mainly determined by some other factor, e.g. some post-translational modification. Antti On 2014-02-24 14:30 , Solanki, Anisha wrote: > Dear Antti, > > Hope you are well. I had a question about the command GPRankTargets(). > It has an input knownTargets. How different would the command run if you > specify some known targets compared to when you don't specify them. > If you don't specify the targets how does the command screen for target > genes. Does it only use the protein concentration predict from the mRNA > Expression? > > Thanks > Anisha -- Antti Honkela antti.honkela at hiit.fi - http://www.hiit.fi/u/ahonkela/

Dear Antti, Hope you are well. I had a question about the command GPRankTargets(). It has an input knownTargets. How different would the command run if you specify some known targets compared to when you don't specify them. If you don't specify the targets how does the command screen for target genes. Does it only use the protein concentration predict from the mRNA Expression? Thanks Anisha [[alternative HTML version deleted]]