mixtures of chemical entities
Bjoern Peters
During OBI development, the need to model mixtures of chemical entities
came up. Examples that we need are reagents such as PBS buffer (e.g.
http://www.histosearch.com/histonet/May01A/Re.PBSbufferrecipesA.html ).
In general, we would like to describe mixtures in terms of constituents
their concentrations.
Is Chebi planning to include modeling these? If not, should we
coordinate how to best do it?
Thanks,
Bjoern
--
Bjoern Peters
Assistant Member
La Jolla Institute for Allergy and Immunology
9420 Athena Circle
La Jolla, CA 92037, USA
Tel: 858/752-6914
Fax: 858/752-6987
http://www.liai.org/pages/faculty-peters
Bjoern Peters
During OBI development, the need to model mixtures of chemical entities
came up. Examples that we need are reagents such as PBS buffer (e.g.
http://www.histosearch.com/histonet/May01A/Re.PBSbufferrecipesA.html ).
In general, we would like to describe mixtures in terms of constituents
their concentrations.
Is ChEBI planning to include modeling these? If not, should we
Bjoern Peters
I think we are exactly on the same page: we just want to create the
framework to describe mixtures, not ennumerate them. Our thoughts are
roughly like this. Some of this may still be controversial, so don't
take this to be blessed from everyone on our side:
* mixture is a material entity that has_part at least two components.
* component of mixture are collections of entities, such as 'portion of
NaCl', 'portion of H20'. This means we also need to have the relation
between molecules (with Chebi ids) and collections thereof. This could
be done by a 'molecule is granular part_of collection of molecules
relation'. There are some hidden minefields with this, but let's skip
that for now.
* components of mixtures can have roles, such as 'buffer'
* components of mixtures have a concentrations in the mixture. That
could be an interesting relation to figure out, as it can be by volume,
or in terms of molecules.
Let us know what you think.
- Bjoern
Kirill Degtyarenko wrote:
> Dear Bjoern
>
> Bjoern Peters wrote:
>> Dear ChEBI Team ,
>>
>> During OBI development, the need to model mixtures of chemical
>> entities came up. Examples that we need are reagents such as PBS
>> buffer (e.g.
>> http://www.histosearch.com/histonet/May01A/Re.PBSbufferrecipesA.html
>> ). In general, we would like to describe mixtures in terms of
>> constituents their concentrations.
>>
>> Is ChEBI planning to include modeling these? If not, should we
>> coordinate how to best do it?
>>
>
> We plan to introduce an upper-level term for mixtures (of chemical
> entities) since we have to include some special cases of
> stoichiometric mixtures (e.g. racemic mixtures). Saying that, we'd
> prefer to concentrate on pure substances (which can include salts and
> addition compounds) but not mixtures. There could be an infinite
> number of mixtures.
>
> We certainly have to coordinate our efforts. I imagine one could
> express a mixture as a set of components (each with a CHEBI ID) with
> their corresponding (starting?) concentrations.
>
> With best regards,
> Kirill
>
Kirill Degtyarenko
> I think we are exactly on the same page: we just want to create the
> framework to describe mixtures, not ennumerate them. Our thoughts are
> roughly like this. Some of this may still be controversial, so don't
> take this to be blessed from everyone on our side:
>
> * mixture is a material entity that has_part at least two components.
> * component of mixture are collections of entities, such as 'portion
> of NaCl', 'portion of H20'. This means we also need to have the
> relation between molecules (with Chebi ids) and collections thereof.
> This could be done by a 'molecule is granular part_of collection of
> molecules relation'. There are some hidden minefields with this, but
> let's skip that for now.
In other words, there has to be a relation between microscopic entity
(e.g. one molecule of ethanol = instance of CHEBI:16236) and macroscopic
entity (e.g. 100 ml of ethanol). Mixture is macroscopic entity, i.e. 100
ml of ethanol + 100 ml of water is a mixture whereas one molecule of
water + one molecule of ethanol is *not*.
> * components of mixtures can have roles, such as 'buffer'
yes. The entire mixture can have role as well, such as "buffer".
> * components of mixtures have a concentrations in the mixture. That
> could be an interesting relation to figure out, as it can be by
> volume, or in terms of molecules.
or it could be mixed, e.g. 10 mg (mass) of A per 100 ml (volume) of
solution. Mass of course can be rewritten in mols if the molecular mass
is known.
It may be hair-splitting but in my view there also should be a
relationship between starting concentrations and equilibrium
concentrations (e.g. you start with so many grams of KH2PO4 and finish
with a solution with several types of phosphate anions)
Kirill