Dear all,
I would like to know if someone can provide me an estimate for the total number of transcriptional regulatory interactions (transcription factor -> target gene) in E. coli. Currently, there are ~4500 such interactions in RegulonDB (according to the file downloaded on October 13th).
Based on these figures, for example, how many transcriptional regulatory interactions remain to be discovered?
This information would be of great value for me because the main goal of my research project is to predict certain phenotypes using network topological measures. As constructing networks is a dynamic task as new interactions are continually added to databases, so we expect the topological measures are always changing as a consequence.
So, it would be very useful to know the estimated total amount of interactions to test the robustness of my prediction models on different networks, including the estimated "complete network".
I am aware that estimates for the total number of protein-protein interactions are rather common. But I could not find any estimates for transcriptional regulatory interactions.
Best regards,
Marcio
Have you seen this: http://biocyc.org/. It is an integrative tool that may help. To find some numbers: http://bionumbers.hms.harvard.edu/. I hope it helps.
Not easy to answer. But probably you don't need to know this. The topological properties will change also in vivo, because every time you have a biological network in your hands it is first a partial reconstruction, and then it is a static snapshot of a dynamical system. In Klein et al., 2013 (Briefings in Functional Genomics) we called it a potential network and then if you would be able to see the network in any given condition you would see that the realization of the network is smaller than the potential network (depending on the gene expression of the TFs, that clearly change in time and depending on the environmental condition that you are considering). Therefore, as we pointed out in that review, also the topological characteristics will change in time/condition. So, instead of looking for the estimated number of regulations, the best way would be to integrate gene expression data and refine the network on its basis to calculate topological properties. Networks are not static in the cell and therefore its topological characteristics are also changing.
- Badges
- Science method