Forums

Hi, Mike,
I meant have both. Always have the full rate law, but where possible put
in a well-defined rate-law identifier, along the lines of the proposal
Howard Salis and others have been discussing. This would simplify model
parsing in many, if not most cases.

-- Upi
--
Upinder S. Bhalla National Centre for
bhalla@ncbs.res.in Biological Sciences, TIFR
+91-80-2363-6420X3230 UAS-GKVK Campus
Fax: +91-80-23636662 Bangalore 560065, INDIA
Web: http://www.ncbs.res.in/~faculty/upi.html

> bhalla> Having said this, I completely agree that rate
> bhalla> laws are needed for generality and for permitting
> bhalla> SBML to encode pretty much all models. But if the
> bhalla> idea is to consciously adopt a dual-tier
> bhalla> representation (rate laws + convenience
> bhalla> information like the current concept of modifiers)
> bhalla> then I would push for a strictly defined and
> bhalla> minimal second tier composed of a small enumerated
> bhalla> set of rate laws. I believe that about 90% of
> bhalla> models could be fully defined with a half-dozen
> bhalla> basic rate laws.
>
> Hi Upi,
>
> When you say a "second tier", what that imply in terms of
> constructs? Are you thinking that we would have an SBML
> scheme whereby models could provide
>
> 1. either a full rate law expression (as it is now)
> 2. *or* skip providing the whole expression and instead
> reference an identifier standing for a well-known law?
>
> I'm just trying to make sure I understand what you have in
> mind.
>
> MH
>
>
>