Keith,
Thanks for your comments
robison at mito.harvard.edu (Keith Robison) wrote:
>Your flowchart for lacZ expression already presents the danger of
>being interpreted in a linear fashion. The "decisions" made by
>lacI (repressor) and CRP are made in parallel. You probably have
>not made the mistake, but your diagram would suggest that CRP acts
>first and lacI second.
Yes, this is probably a necessary correction. But discussing these
things is just part of the process of refining the model. It wasn't
obvious to me that these "decisions" are made in parallel - at least not
parallel until the activator initially signaled that there was no
glucose present.
(BTW, I know the yes/no were initially mistakenly placed on the wrong
branches of output on the "lactose present" diamond. I asked the
moderator to post a comment about this, because, for some reason, I
can't see my original posting - only Robison's response. Maybe it's my
news server.)
>>I think you'll find such diagrams will suffer from such problems
>increasingly as you try to model a significantly complex system.
>Flowcharts are inherently linear beasts, ill-suited for parallel
>processes, especially biological ones with many non-linearly combined
>inputs.
>
Yes, these problems will definitely increase. The representation is
meant to be suggestive of the large scale structure. Eventually any
flow chart will look like a tangled mess, but that still doesn't
discourage me from making more charts. At least not until I or someone
else comes up with a better representation of these processes. Though
this type of representation is crude and inherently flawed, I haven't
seen significantly better ones of gene expression processes in textbooks
or journal articles.
Maybe someone can give me a good reference?
I know there are diagrams representing interactions, but they usually
relate to the physical relationships of the genes, RNA and ribosomes
rather than abstract diagrams of processes. I'm trying to get away from
the topography - i.e. the physical layout of things - and instead make
sense of the relationships between processes.
Diagrams I've seen of interactions haven't really indicated the overall
flow of events, that's also part of what I'm trying to get at.
What might be some of the intitial processes on a chart for a whole
organism be? Final processes? What's the relationship between
different developmental subroutines? Those are some of the questions
that I think would benefit from a large scale analysis. I want to step
back and try to see the whole elephant - even if my vision is blurred.
You mentioned problems with
>non-linearly combined inputs.
Yeah, that may be tricky, but think about the fuzzy logic chip that
controls the processes taking place in expensive Japanese cars.
>>>Keith Robison
>Harvard University
>Department of Cellular and Developmental Biology
>Department of Genetics / HHMI
>>robison at mito.harvard.edu>
Gary Welz
