W3C home > Mailing lists > Public > public-semweb-lifesci@w3.org > April 2009

Re: blog: semantic dissonance in uniprot

From: Egon Willighagen <egon.willighagen@gmail.com>
Date: Thu, 2 Apr 2009 16:56:31 +0200
Message-ID: <6aeb064b0904020756w3e153e0ex6ba67f06724004c@mail.gmail.com>
To: Matthias Samwald <samwald@gmx.at>
Cc: Phillip Lord <phillip.lord@newcastle.ac.uk>, Oliver Ruebenacker <curoli@gmail.com>, Pat Hayes <phayes@ihmc.us>, public-semweb-lifesci <public-semweb-lifesci@w3.org>
On Tue, Mar 31, 2009 at 2:09 PM, Matthias Samwald <samwald@gmx.at> wrote:
>>> Reaction equations describe stochastic processes, that's why you can
>>> have non-integer molecule numbers
>> I think you can't have non-integer molecule numbers because it makes
>> no chemical sense. Half a molecule is a whole molecule of a different
>> kind.
> You can have reaction equations that look like
> N2O5 ---> 2 NO2 + 1/2 O2
> Which means that the number of O2 molecules that would be produced if the
> equilibrium would be shifted to the absolute right side is 1/2 of the number
> of molecules of N2O5 that would exist if the equilibrium would be shifted to
> the absolute left. This only makes sense if we interpret reaction equations
> as descriptions of pools of molecules and their stochastic processes, rather
> than single molecules. Representing reaction equations as processes where
> the participants are single molecules is wrong. In that case, one cannot
> blame OWL if one is running into inconsistencies.

Actually, I'd say OWL is to blame here... that is, the OWL class was
not properly defined.

Reaction Equations are a difficult concept, and depending on the
context mean different things. This discussions is just caused by
different meaning people give to Reaction Equation. Very typically,
you would even mix both types in reaction schemes (e.g. schemes where
both the overall as well as mechanistic reactions are given), even
though they refer to different concepts.


Post-doc @ Uppsala University
Received on Thursday, 2 April 2009 14:57:12 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 14:52:39 UTC