Thanks a lot! What kind of data should I used to benchmark it, chip-chip data? I still have a question: even though I could know exactly the distribution, if it will make little contribution for me to understand gene expression data, due to PTM and cellular environment?
Dear all,
I am interested in the distribution of protein numbers regulation by transcription factor (TF) based on TF PWM matrix. Specifically, the strength of regulation effects on downstream genes computed based on Position Weighted Matrix.
I have read the paper
http://www.sciencedirect.com/science/article/pii/0968000488900850
http://www.ploscompbiol.org/article/info%3Adoi%2F10.1371%2Fjournal.pcbi.1000039
Further, I propose to use this information to infer the probability of regulation relationship between TF and gene. As I known, besides PWM matching other factors play crucial roles as well, such as post translation modification, environment conditions. So, based solely on this data, to what extend could we believe the performance of this method.
Thanks a lot!
Zhe
1 answer
Your question cannot be answered with any simple statement. One important aspect is that even without unmeasurable aspects such as post-translational modification in a genome-wide context, the performance differs strongly by transcription factor/PWM.
If you are interested in a single or a small set of TFs, benchmark it yourself by using external data.
Log in to answer this question.
I find your question hard to read. Is what you mean to ask "how likely is it that transcription factor binding motif really binds the transcription factor for which it is the target?"