I guess this question is extremely difficult to answer.

Because transcription factors act as activators and repressors depending on time, stiuation and target genes. So there might not exist purely repressing and purely activating factors.

Even the lac repressor (one of the first and famous example for a repressor) is temporarily an activator

http://www.sciencedirect.com/science/article/pii/0092867487900936

Thanks for your reply

I edited the question now

great, now the question makes more sense.

Yes, Bionumbers came to my mind - because this is such kind of questions one can only guess ...

But it seems that we could get an answer only by doing some text mining approaches (with a list of these around 1000 TFs).

Still, I fear that the current knowledge is not yet at the state to answer this question accurately.

The difficulty is also in the point that TF binding is difficult to predict today, especially in mammalian cells (in Drosophila TFs tend to have longer binding sequences, but in mammals we have often 6 or even 4 bp motifs - leading to many false positive binding site predictions - statistically, every few thousand bps in the genomes there is a potential binding site for each of such motifs ...)

Some groups even have seen evidence that the "same" (homolog) genes in different species have very different binding sites for their corresponding target genes.

And it seemed that the DNA sequence dictates the binding of TFs (human chromosome 21 sequence bound in mouse environment the TFs in a similar way like the human TFs bind the chromosome 21 sequence e.g.)

And in addition, the binding sites seem to evolve rapidly (which makes proper estimation also difficult).

http://www.nature.com/nrg/journal/v15/n4/full/nrg3481.html

and combinations of TFs seem to improve expression predictions in certain cell types (perhaps because they reduce the false positives some how?):

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4359374/

In which way you want to simulate TF-gene activations and repressions?

Dear

In definition: In molecular biology and genetics, a transcription factor (sometimes called a sequence-specific DNA-binding factor) is a protein that binds to specific DNA sequences, thereby controlling the rate of transcription of genetic information from DNA to messenger RNA. Transcription factors perform this function alone or with other proteins in a complex, by promoting (as an activator), or blocking (as a repressor) the recruitment of RNA polymerase (the enzyme that performs the transcription of genetic information from DNA to RNA) to specific genes.

Please find more details according to your question in the following publication entitled: Recent advances in the structural molecular biology of Ets transcription factors: interactions, interfaces and inhibition. Christopher D.O. Cooper, Joseph A. Newman, Opher Gileadi Biochemical Society Transactions Feb 01, 2014, 42 (1) 130-138.

Introduction:  The Ets TF (transcription factor) family is found throughout the metazoa, comprising 28 members in humans [1,2], all containing the evolutionarily conserved DNA-binding Ets domain which binds the invariant DNA sequence 5′-GGA(A/T)-3′ [2]. Ets TFs play important roles in normal cellular development and differentiation [3], but when deregulated are significant mediators of tumorigenesis in various cancers [4,5]. Ets proteins are subclassified by the presence of further domains associated with PPIs (protein–protein interactions) or transcriptional regulation [2,3], including the TCF (containing the B-box [6]) and PEA3 (containing the unstructured PEA3 transactivation domain [7]) subfamilies. PNT (pointed) domains are also frequently found N-terminal to the Ets domain, involved in PPI and homodimerization

You can download the publication at the direct link below: http://www.biochemsoctrans.org/content/42/1/130

Hoping it will be hepful

Marius

I read somwhere (I have to find this reference) that a mammalian cell has 5000-6000 expressed genes.

If the human genome has let's say 30000 genes, then around 1/5th of the genes are active in each cell type. And  around 4/5th of the genes then are repressed. Maybe this gives a starting point for some guesstimation work?

Thanks for your participation

I will read the articles provided and continue discussion

It should be noted that I want to know the percentage of different kinds of regulation in experimentally identified and known TF-gene interactions. I do not prefer to predict new TF or other things.  

You are welcome. :)

Yes, I understood your concern. But I talk about prediction because I guess you will be in the situation that you have to estimate the numbers yourself, because I don't think that there will be a reference which responds accurately your question. The current state of the research is that we don't understand really well how genes are regulated as you already know. How should there exist then a reliable number about promoting or inhibiting TF-gene interactions? And even if, the estimation will be very inaccurate ...

My PhD was about a transcription factor Sox9, its downstream targets and its regulation, so I read about quite some transcription factors at that time. I am aware that there are many other people which read more about even more transcription factors, but the field itself won't have been changed that much meanwhile. In the thesis, we identified a far upstream regulatory region for Sox9. While writing the thesis, I realized that there are not that many examples for far upstream regulatory elements as I expected. So overall, I would say, we are currently understanding very little about gene regulation especially in human cells.

Sometimes, TFs binding in the neighbour gene's intronic regions can influence the regulation of a gene. And how far a far upstream regulatory region can be located is also unclear. And it is also possible that sequences in trans (inter-chromosomal) can have influence on the expression of a gene. (just for imagination in picture:

https://www.google.de/search?q=chromosome+3D+localization+nucleus&client=ubuntu&hs=JQz&channel=fs&source=lnms&tbm=isch&sa=X&ved=0ahUKEwiKltDBvNfPAhWGbRQKHVh4AGIQ_AUICCgB&biw=1488&bih=901

)

Personally, I also would avoid predictions of TF binding sites, because false positives are unavoidable. And even if the TF if this binding the predicted binding site, it is again unclear, if it is involved in the regulation of the gene by this binding site ...

So a vast amount of uncertainities ...