Team:UNAM-Genomics Mexico/Modeling/MM



Plasmid Stability

So we decided to model the stability of our working plasmid. The data we had were the surviving colony counts after a certain time period. Therefore, we were interested in the probability of loss for the plasmid, at the moment of celular division. We assumed the following in our model:

  1. Time-discrete
    1. Because the event can only occur at the passage from one generation, to the next. This model attributes the blunt of the loss probability to plasmid segregation issues at cell division.
  2. Equal-probability per iteration
    1. Since the event of the plasmid being kept from one generation to the next is sufficiently similar as to be considered as the same.
  3. Plasmid can only be inherited, never generated.
    1. The plasmid can not be spontaneously generated from nothing, nor can the phenotype suddenly present a gain-of-function mutation to resist our selection markers.


As always, our data is available: here.


Applying a Markov model to the stability, with <X> the event of plasmid keeping for passage of <t> to <t+1>, we found a value of P(X=True)=0.999774 and consequently P(X=False)=0.000226.