Team:Paris/Modeling/More RBS issue


Let's see on our example why it is not a problem. Our wish is to characterize the strength of the promoter pFliL, in function of the transcriptions factors FlhDC and FliA. As it is explained in the characterisation protocol, it is done in two steps :

  • First, we study the activity of the promoters pTet and pBad, in function of the small molecules we introduce in the medium. It gives, with the standard RBS B0034, respectively the functions ƒ1 and ƒ2.
  • Then, we put the genes flhDC and fliA behind pTet and pBad, but with their natural RBS. In this order, we will get
[FlhDC] = {coefflhDC} * ƒ1 and [FliA] = {coeffliA} * ƒ2.

So,the pFliL added will have a given activity, measured with the RBS B0032 throught the expression of GFP. We have shown in our modeliing that we expect it to have the following analytical expression :


that gives, with previous considerations and by dividing numerator and denominator by {coefgene}n :


Thus, we can still evaluate the expression by keeping ƒ1 and ƒ2, but we must notice that the dissociation constant which we will estimate is not exactly the dissociation constant of the complexation FT><pFliL. From now on, we understand that the calculated quantities of proteins by our global model are not the one in the "reality". Still, the purpose of our "Characterization Approach" is kept, by considering relative expression of the proteins...

We can moreover notice that, by using this method in other circumstances, we can get predictive quantities and real constants, by keeping always the same RBS (for example B0032) in all constructs !