[R-sig-dyn-mod] Turbidostat

[Thomas Petzoldt]

On 12/8/2010 9:00 PM, Uta Berger wrote:
> Thomas,
>
> thanks again.
>
>>
>> 2) Why does the simulation start at time step 1 and not at zero like
>> the data?
> It was my mistake.
>>
>> 3) Is there any evidence why this inhibition occurs at the specified
>> times only? Is there any chance to describe this a little bit more
>> process oriented?
>
> Toxic substances were given during the initial phase (first four time
> steps) and before the last time step. During the "middle time steps",
> algae could grow without inhibition.
> Unfortunately, I do not have a chance for a mechanistic explanation
> since we have only the data for the algae. I did not present all of
> them. Three experiments were done made with three different levels of
> toxic substances. (I know the levels by number, e.g., Uran = 0; 0.2 and
> 0.5mu mol, but do not know the absorption by the algae).
>
> The three data sets thus differ mainly in their initial phase and in the
> last phase. The "mid phase" is determined by the last concentration of
> the initial phase (time step four in the example).

I see. So you may consider to handle korr_s as an ordinary parameter of 
the model that can be fitted with the usual procedure. To model the 
presence of the inhibitor one may use either (1) a signal set to 0 resp. 
1 or (2) a state variable for the toxic substance that is added at 
specified times (by help of a signal or an event) and that is removed by 
dilution of the culture -- and optionally something like a 1st order 
decay or an uptake by the cells.

Thomas P.