> Sounds an extremely ambitious project. First question is what is the
> software product that you're going to convert?
> Second question is how will you go about using the static information in
> genomic databases to work out an activities path?
The genomic databases establish a vocabulary of words and phrases
that are used to describe genetic cellular activities. For example,
the names of proteins and genes can be extracted. The activities of
each gene may take decades for researchers to tease out, but some
gene functions are known today. Its a start. The estimated 50,000
genes that will eventually be found constitutes an end point.
> Assuming that you can do
> this, the number of possible paths from your initial to your final
> conditions will be huge,
Yes, the combinations are enormous. I have a plan to limit the
combinatorics to some degree, but the algorithm will definitely
be very computation intensive. Since we start with a small
proportion of the genes known and described, it should be
workable with today's computer speeds. As we get deeper
into the gene count, we'll need better methods, faster computers,
and more knowledge about how to limit the searches. But all
that is in the future; it will be here when we need it.
> how will you know which ones occur in the cell and
> which ones are fake?
Validation of theories is far past the simulation. The simulation helps
people think about genetic expression, but doesn't validate it. The
validation of genetic expression theories is work for the lab, and
possibly good reason for a more automated lab that 'understands'
> Do you honestly think it is possible to specify a
> cancerous cell in detail (if it were possible surely someone would have
> developed a cure by now?)? I didn't think they knew that much about the
> detailed mechanism of cancer?
The following is an example of genetic research related to genetic models
CRISP search keyword: logic
NIH Grant Number: 5R37GM17709-28 made in FY97, started 9/1/75, ends 2/28/01
"GENETIC ANALYSIS OF THE EUKARYOTIC CELL CYCLE"
Funded by NATIONAL INSTITUTE OF GENERAL MEDICAL SCIENCES
HARTWELL, LELAND H.
FRED HUTCHINSON CANCER RESEARCH CENTER
BOX 19024, 1100 FAIRVIEW AVE N
SEATTLE, WA 98109
CRISP search keyword: logic AND gene
NIH Grant Number: 3R01CA21104-21S1 in FY99, started 6/1/77, ends 2/28/01
"CELL-CELL COMMUNICATION CARCINOGENESIS"
Funded by NATIONAL CANCER INSTITUTE
TROSKO, JAMES E.
MICHIGAN STATE UNIVERSITY, PEDIATRICS AND HUMAN DEVEL
EAST LANSING, MI 48824
There are many others, not necessarily cancer related. Genetic function is
just being teased out now that we know the common genome sequences.
The rationale behind a simulation of gene function is to provide a
that people can use to combine their knowledge with other people's, and
to promulgate the results in a form (i.e., a simulation model) that others
can use to discover more and more complex mechanisms.
My own theory is that there are unknown genetic programs that work
over several generations, possibly repeating after N generations. One
way to find them is to build genetic expression models that are exact.
But that doesn't mean we have to already know the 50,000 genes and
their functions to derive benefits from the simulation and modeling.
I know the ultimate result may be decades away, but there is some
immediate benefit, and that benefit can grow incrementally.
Thanks for your interest,
> Herbert Sauro
>> "Rich Cooper" <richcooper1 at mindspring.com> wrote in message
> news:8skpug$1at$1 at slb6.atl.mindspring.net...> > I'm planning to convert my software product into a cellular activity
> > simulation product. The researcher would specify the initial condition
> > the cell, and specify the final condition, and then the software would
> > genetic function databases to stitch together the sequence of activities
> > objects that would drive the cell from the initial condition to the
> > condition. The researcher would then browse through the stitched
> > activities that a cell might use to perform the sequence of activities.
> > For example, you might specify a healthy cell in the details that
> > you. Then you might specify a cancerous cell in details using the same
> > vocabulary of genes, proteins, activities, and physical objects.
> > My background is PhD in CS&EE. My medical engineering background is
> > primarily in physiology, at the organ level, with little chemical
> > background. I've read all the popular genetics books, as well as
> > "Recombinant DNA" classic. I'm aware of the MitoMap and the Gene
> > Consortium databases. These might provide a starting vocabulary for
> > simulation specifications.
> > I'm looking for biochemist partners to help define what the product
> > do, and how it should be provided. If you have any interest in guiding
> > directions that a cell simulation product should take, please let me
> > You can answer the post here, or you can email me directly at:
> > mailto:richcooper1 at mindspring.com> >
> > Sincerely,
> > Rich Cooper