Re: rxn direction in SBML

Jeremy Zucker

unread,

May 23, 2005, 9:04:12 PM5/23/05

to Debuggin...@googlegroups.com, Markus Krummenacker

Hi Markus,

I cc'ed debugging-the-bug@googlegroups, because I thought they might
be interested in my response.

On May 23, 2005, at 2:28 PM, Markus Krummenacker wrote:

> Hi Jeremy,
>
> Just a quick question, as you are the SBML expert:
>
> How is the reaction direction represented in SBML ?
>
> Is there only one "right" way, or are there several options we need to
> be aware of ?
>
> In the current SBML format that is emitted from the Ptools, I see that
> every rxn has a "reversible" flag, that can be true or false.
>

This is the official, sanctioned way to represent reaction direction in
SBML Level 2, version 1 for the purpose of
qualtiative analyses such as elementary flux modes, or Extreme
pathways. If you want a more detailed, quantitative account of
reversibility then you must use the kineticLaw for each
reaction.

> If this is false, then the direction of the reaction is from the
> listed reactants to products.
>
> If it is true, then the rxn is still only listed once, as opposed to
> as 2 rxns that go in opposite directions.
>
> Is this all correct ?
>

Yes.

> Is there any distinction between several levels of "irreversibility" ?
> Our schema has values that say "this is irreversible under
> physiological conditions" (even if in vitro, reversibility could maybe
> be demonstrated).
>

This is a very good point. It would be nice if reversibility was not
boolean so that deeper questions can be asked. In the next release of
SBML, we would like to specify the concept of constraints, which
will enable one to specify the reaction direction based on additional
criteria. This is what is needed for Flux balance analysis, too.

> Also, is there a way to state the certainty about this information ?
> I.e. is the default just reversible ? Is there any way to state that
> the default has been chosen ?
>

In SBML L2v1, the default value for reversible is "true". To add
additional information, every reaction has an optional notes and
annotation tag.
The convention for the annotation tag is to add RDF or XML. the
convention for the notes tag is to add XHTML. Thus, if you wanted to
add hang a few more pieces of information about the reaction,
such as reversibility type, and certainty, one way to do it is to
annotate the reaction with BioPAX metadata using the metaid attribute.

For example:
<reaction id="PGLUCISOM_RXN" metaid="PGLUCISOM-RXN" reversible="true">
<annotation
xmlns:biopax="http://www.biopax.org/release/biopax-level1.owl#"
xmlns:ecocyc="http://biocyc.org/ECOLI/#">
<rdf:RDF rdf:about="#PGLUCISOM-RXN">
<biopax:EVIDENCE
rdf:resource="http://biocyc.org/ECOLI/reference?type=EVIDENCE-
GLYPH&object=(PGLUCISOM-ENZRXN+:EV-EXP:3277835749:pkarp)"/>
</rdf:RDF>
</annotation>
<listOfReactants>
<speciesRef species="GLUCOSE_6_P" stoichiometry="1"/>
</listOfReactants>
<listOfProducts>
<speciesRef species="FRUCTOSE_6_P" stoichiometry="1"/>
</listOfProducts>
<listOfModifiers>
<modifierSpeciesRef species="PGLUCISOM">
</listOfModifiers>
</reaction>

> Thanks for any feedback you may have.
>
> --
> --
> Regards
> Markus Krummenacker
>

I hope this helps,

By the way, Aaron and I were wondering if you had any suggestions about
how to deal with generic metabolites, such as NAD(P)H and dNTP, and
such. In Ecocyc 9.0, we have observed that all reactions that used to
involve NAD+ or NADP+ now all involve NAD(P)+. However, NAD(P)+ does
not contain atomic mass, so none of those reactions are balanced.

or what about the case of reaction 1.1.1.-, which is a particular
instantiation of the generalized reaction 1.1.1.35?

What do you recommend as the best way to handle these situations?
In the first case, it seems pretty straightforward, just make two
separate reactions for NAD and NADP.
In the second case, It is less clear how to identify and correct these
problems in general, since the link between these reactions is only
implied in the comments, and there is no direct link between them.
The best thing to do it seems like to me is simply to enumerate the
list of problems, and hopefully it will be of a manageable size to do
manually.

Sincerely,

Jeremy Zucker
Bioinformatics Specialist
Dana-Farber Cancer Institute
url: http://research.dfci.harvard.edu
email: zuc...@research.dfci.harvard.edu
work: 617-632-6852
cell: 617-833-3196

Message has been deleted

Thomas, GH

unread,

May 24, 2005, 4:14:39 AM5/24/05

to Debuggin...@googlegroups.com

Dear group,
I am sorry this mail was posted accidently to the group.
Please ignore.
Cheers,
Gavin Thomas.

---------------------------------
Dr. Gavin H. Thomas,
Lecturer in Molecular Microbiology,
Department of Biology, Area 10,
University of York.
PO Box 373,
York. YO10 5YW

Tel: 01904 328678 Fax: 01904 328825 Email: gh...@york.ac.uk

York page: http://www.york.ac.uk/depts/biol/staff/ght.htm
EchoBASE: http://www.ecoli-york.org/
The E. coli index: http://ecoli.bham.ac.uk/
---------------------------------

Markus Krummenacker

unread,

May 24, 2005, 10:44:05 PM5/24/05

to Debuggin...@googlegroups.com, Markus Krummenacker

Jeremy Zucker writes:
> Hi Markus,
>
> I cc'ed debugging-the-bug@googlegroups, because I thought they might
> be interested in my response.

Hi Jeremy,

Sure, sounds good to me.

Thank you for your clarification regarding the SBML format, and your
suggestion of using annotation tags for adding evidence information.

Are you planning to implement something like that in the SBML exporter
code ?

> By the way, Aaron and I were wondering if you had any suggestions about
> how to deal with generic metabolites, such as NAD(P)H and dNTP, and
> such. In Ecocyc 9.0, we have observed that all reactions that used to
> involve NAD+ or NADP+ now all involve NAD(P)+. However, NAD(P)+ does
> not contain atomic mass, so none of those reactions are balanced.

I have been experimenting with some code to automatically convert
reactions that refer to compound classes to a corresponding set of
reactions that use compound instances instead. It sort of works for a
number of cases, especially the fairly easy case of the NAD(P)+ class,
but there are definitely still problems with it.

So yes, I think replacing many of those reactions with instance
versions is a valid technique for modelling. I am not decided yet
regarding whether we should permanently store those instance
reactions in EcoCyc, or whether they simply should be generated
on-the-fly. Any opinions on this ?

If we would want to permanently store such reactions, it would be
really helpful to develop a better reaction-class concept, to
hierarchically group transformations according to narrowness of
substrate specificity. This would involve some thought and effort
first, which has not currently been scheduled for.

Note that this still will leave open other thorny questions, such as
fatty acid polymerization reactions, where it is not clear to me
whether it would be beneficial to just expand out to all species that
could possibly be generated.

> or what about the case of reaction 1.1.1.-, which is a particular
> instantiation of the generalized reaction 1.1.1.35?

I can't find (in ecocyc-9.1+) which reaction you are referring to with
1.1.1.- (you did not indicate the frame ID).

But I think I understand what you mean. In my experimental
class-to-instance substitution code, I do check whether a duplicate
reaction might already exist, which a curator may have manually
entered as a specific example of a more generic reaction. In that
case, the duplicate generated reaction is suppressed.

> What do you recommend as the best way to handle these situations?
> In the first case, it seems pretty straightforward, just make two
> separate reactions for NAD and NADP.
> In the second case, It is less clear how to identify and correct these
> problems in general, since the link between these reactions is only
> implied in the comments, and there is no direct link between them.

Yes, we would need a better reaction-class hierarchy.

> The best thing to do it seems like to me is simply to enumerate the
> list of problems, and hopefully it will be of a manageable size to do
> manually.
>
> Sincerely,
>
> Jeremy Zucker
> Bioinformatics Specialist
> Dana-Farber Cancer Institute
> url: http://research.dfci.harvard.edu
> email: zuc...@research.dfci.harvard.edu
> work: 617-632-6852
> cell: 617-833-3196

Markus Krummenacker

unread,

May 24, 2005, 10:51:09 PM5/24/05

to Debuggin...@googlegroups.com, Markus Krummenacker

Jeremy Zucker writes:
> In Ecocyc 9.0, we have observed that all reactions that used to
> involve NAD+ or NADP+ now all involve NAD(P)+.

Oh, I forgot to address this specifically in my previous reply.

Are you sure about this statement ? All reactions ?

I am looking at ecocyc-9.1+ (BTW, ecocyc-9.1 has just been released,
through the www). As far as I can tell, NAD is linked to a lot more
reactions that the |NAD(P)| class, which is what I would expect.

If you think a bunch of reactions were changed from NAD to |NAD(P)| ,
could you please provide a more detailed list ? If there is some kind
of data bug, we should probably investigate.

Pedro R. Romero

unread,

May 25, 2005, 11:22:07 AM5/25/05

to Debuggin...@googlegroups.com

Markus Krummenacker wrote:

...(snip)...

So yes, I think replacing many of those reactions with instance
versions is a valid technique for modelling.  I am not decided yet
regarding whether we should permanently store those instance
reactions in EcoCyc, or whether they simply should be generated
on-the-fly.  Any opinions on this ?

If we would want to permanently store such reactions, it would be
really helpful to develop a better reaction-class concept, to
hierarchically group transformations according to narrowness of
substrate specificity.  This would involve some thought and effort
first, which has not currently been scheduled for.

As I said before, I believe we need subclasses of reactions that can have many instances, and save those instances in the KB. Generating them on the fly, as you point out, can be very easy in some cases, but it can become more than non-trivial in others.

Note that this still will leave open other thorny questions, such as
fatty acid polymerization reactions, where it is not clear to me
whether it would be beneficial to just expand out to all species that
could possibly be generated.

These could be handled by expanding, as you said, the species at both sides of the reaction. For example, for the reaction

A-X(n) <=> A-X(n+1)

where n goes from, say, 1 to 12 in E. coli, we could have

A-X or A-X(2) or .... A-X(11) <=> A-X(2) or A-X(3) or ... A-X(12)

So, you can produce any of the species on the right hand side if you have any of the species on the left hand side (if I have A-X(2), I can produce A-X(10), and any of the other combinations)

Just my two cents...

Pedro

Marcin Imielinski

unread,

May 25, 2005, 12:20:55 PM5/25/05

to Debuggin...@googlegroups.com

Hi,

Not to complicate things further, but one could also describe species specificity of enzymes / reactions according to functional groups or moieties of molecules, i.e. R-OH à R=0. Since recognition of a substrate by an enzyme usually depends on some key structural element in the molecule and is invariant to some other structural elements, this could be specified kind of like a regular expression on the Smiles structure. This could also be more “readably” implemented by annotating each species in terms of its moieties / subunits, and then specifying reactions as converting a certain moiety in a given context to another moiety.

This kind of encoding would become especially efficient as annotated networks start to include larger species, since the number of possible species will undergo combinatorial explosion due to mixing of subunits. In this case it is best to specify reactions only according to the “business end” of the target species.

Actually some work has been done in this direction by Michael Blinov, who is a post doc at Los Alamos, whose talk I recently saw.

Also on the topic of moieties / structural subunits – I think it would be very useful to annotate species in terms of component functional groups (and have a “hierarchy” of functional groups including other functional groups). I realize this would demand a lot more annotation to the DB, however such a representation of species would be very useful in 1) determining “real” connectivity of species in the network (a la Arita’s paper in PNAS 2004) and 2) correlating maximal conserved moieties on species in the network with extreme semipositive conservation laws.

-Marcin

Markus Krummenacker

unread,

May 25, 2005, 2:24:47 PM5/25/05

to Debuggin...@googlegroups.com

Marcin Imielinski writes:
> Also on the topic of moieties / structural subunits - I think it would

> be very useful to annotate species in terms of component functional
> groups (and have a "hierarchy" of functional groups including other
> functional groups). I realize this would demand a lot more annotation
> to the DB,

Hi Marcin,

Thanks for your suggestions. I have also been thinking along these
lines for a while now (but I haven't done anything yet). As you say,
it would involve a substantial reworking of the way molecules and
reactions are dealt with, and it is not clear yet how soon we will be
able to get to that.

dasher484

unread,

Jun 22, 2005, 11:46:50 PM6/22/05

to Debuggin...@googlegroups.com

The topic sounds interesting, so I would like to join if possible!

Greetings dasher484!

>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>

Markus Krummenacker ha escrito:

zuc...@research.dfci.harvard.edu

unread,

Jun 23, 2005, 12:22:35 AM6/23/05

to Debuggin...@googlegroups.com

Hello Victor,

You are more than welcome to join the discussion. But if you don't
mind, could you tell me a little about yourself? Many of the issues we
are dealing with on this discussion list have to do with issues of
semantics and syntax in modeling metabolic pathways of bacterial
organisms. Is this something you would find interesting?

Sincerely,

Jeremy

aar...@media.mit.edu

unread,

Jul 7, 2005, 12:24:23 PM7/7/05

to Debuggin...@googlegroups.com, Markus Krummenacker

aar...@media.mit.edu

unread,

Jul 7, 2005, 12:50:19 PM7/7/05

to Debuggin...@googlegroups.com, aar...@media.mit.edu, Debuggin...@googlegroups.com, Markus Krummenacker

Hi All,
I am working on a project with Xiaoxia Lin at the Church lab at the Harvard
Medical School in which I am using MetaCyc and organism specific data bases in
order to determine which reactions might be responsible for the metabolism of
P. marinus. In order to do this I need to set up steady state flux balance
constraints. In such a system reactions with |NAD(P)| and |NAD(P)H| will not
balance with reactions that specifically have NAD, NADP, NADH and
NADPH. I have
written software to create specific reactions to replace the general
metabolites
with the specific ones. (It would create two new reactions in the preceding
case.)
Here's my question: What about reactions with |Acceptor| and |Donor-H2| .
These are special cases of |Redox-Electron-Carriers|.
Is it always safe to assume NAD, NADP, FAD, FMN, and NADH, NADPH, FADH2, FMNH2
can be substituted for these?
Is there more specific info available to resolve this?
What about the other Redox-Electron-Carriers?

I have similar questions for making specific reactions with |Ubiquinols| and
|Ubiquinones| and |Menaquinols| and |Menaquinones|
I will pose them in more detail later.

Thanks,
Aaron
--- Class |Redox-Electron-Carriers| ---
Supers: |Coenzymes|
Subs: |Acceptor|, |Donor-H2|, |Menaquinols|, |Menaquinones|,
|NAD(P)|, |NAD(P)H|, |Ubiquinols|, |Ubiquinones|
Instances: FAD, FADH2, FMN, FMNH2

SCHEMA?: T

--- Class |Acceptor| ---
Supers: |Redox-Electron-Carriers|
Subs:
Instances:

APPEARS-IN-LEFT-SIDE-OF: R362-RXN,
3-HYDROXYCYCLOHEXANONE-DEHYDROGENASE-RXN,
SARCOSINE-DEHYDROGENASE-RXN, RXN-2841,
RXN0-2301, RXN0-2101, 1.4.99.5-RXN,
1.20.99.1-RXN, 1.3.99.20-RXN, 1.97.1.8-RXN,
RXN0-1, RXN-1301,
|ALCOHOL-DEHYDROGENASE-(ACCEPTOR)-RXN|,
|ALKAN-1-OL-DEHYDROGENASE-(ACCEPTOR)-RXN|,
RXN-945, CYT-ACC-RXN, R282-RXN, RXN-944,
RXN-905, RXN-865, LACTOSE-DEHYDRO-RXN,
SUCC-FUM-OXRED-A-RXN, RXN-641, R22-RXN,
R221-RXN, R161-RXN, 1.3.99.16-RXN,
1.3.99.17-RXN, 1.3.99.18-RXN, 1.3.99.19-RXN,
1.5.99.10-RXN, NITRATREDUCT-RXN, MOXXANAU-RXN,
SULFITE-REDUCTASE-RXN, MANDELATE-OXY-RXN,
4-HYDROXYMANDELATE-OXY-RXN, 1.17.99.1-RXN,
CARBON-MONOXIDE-DEHYDROGENASE-RXN,
FORMYLMETHANOFURAN-DEHYDROGENASE-RXN,
CARBOXYLATE-REDUCTASE-RXN, 1.2.99.3-RXN,
3-OXOSTEROID-1-DEHYDROGENASE-RXN,
3-OXO-5-BETA-STEROID-4-DEHYDROGENASE-RXN,
2-METHYLACYL-COA-DEHYDROGENASE-RXN,
2-FUROYL-COA-DEHYDROGENASE-RXN, 1.3.99.5-RXN,
GLUTARYL-COA-DEHYDROGENASE-RXN,
DIHYDROOROTATE-DEHYDROGENASE-RXN,
CYCLOHEXANONE-DEHYDROGENASE-RXN,
BETA-CYCLOPIAZONATE-DEHYDROGENASE-RXN,
TAURINE-DEHYDROGENASE-RXN,
ARALKYLAMINE-DEHYDROGENASE-RXN,
AMINE-DEHYDROGENASE-RXN,
SPERMIDINE-DEHYDROGENASE-RXN,
TRIMETHYLAMINE-DEHYDROGENASE-RXN,
NICOTINE-DEHYDROGENASE-RXN,
L-PIPECOLATE-DEHYDROGENASE-RXN,
METHYLGUTAMATE-DEHYDROGENASE-RXN,
DIMETHYLGLYCINE-DEHYDROGENASE-RXN,
NADPH-DEHYDROGENASE-RXN,
|NADPH-DEHYDROGENASE-(QUINONE)-RXN|,
|NADH-DEHYDROGENASE-(QUINONE)-RXN|,
NITRIC-OXIDE-REDUCTASE-RXN,
HYDROXYLAMINE-REDUCTASE-RXN,
NITRITE-REDUCTASE-RXN,
NITROUS-OXIDE-REDUCTASE-RXN, 1.7.99.5-RXN,
HYDROGENSULFITE-REDUCTASE-RXN,
ADENYLYLSULFATE-REDUCTASE-RXN,
PYRIDOXINE-5-DEHYDROGENASE-RXN,
D-2-HYDROXY-ACID-DEHYDROGENASE-RXN,
D-SORBITOL-DEHYDROGENASE-RXN,
DEHYDROGLUCONATE-DEHYDROGENASE-RXN,
FRUCTOSE-5-DEHYDROGENASE-RXN,
GLUCONATE-2-DEHYDROGENASE-RXN,
|GLUCOSE-DEHYDROGENASE-(ACCEPTOR)-RXN|,
1.1.99.27-RXN,
2-HYDROXYGLUTARATE-DEHYDROGENASE-RXN,
|CELLOBIOSE-DEHYDROGENASE-(ACCEPTOR)-RXN|,
URACIL-DEHYDROGENASE-RXN, 1.1.99.23-RXN,
SORBOSE-DEHYDROGENASE-RXN,
|MALATE-DEHYDROGENASE-(ACCEPTOR)-RXN|,
GLYCEROL-3-PHOSPHATE-DEHYDROGENASE-RXN,
|GLYCEROL-DEHYDROGENASE-(ACCEPTOR)-RXN|,
GLUCOSIDE-3-DEHYDROGENASE-RXN, CHD-RXN,
NADH-DEHYDROGENASE-RXN, DAADEHYDROG-RXN,
GLYCOLATEDEHYDRO-RXN

APPEARS-IN-RIGHT-SIDE-OF: VCREDCHLOR-RXN, DCEREDCHLOR-RXN,
TCEREDCHLOR-RXN, 1.14.99.21-RXN,
1.14.19.2-RXN,
STEROID-9-ALPHA-MONOOXYGENASE-RXN,
PROSTAGLANDIN-ENDOPEROXIDE-SYNTHASE-RXN,
1.14.19.1-RXN, 1.14.99.37-RXN,
2-HYDROXYPYRIDINE-5-MONOOXYGENASE-RXN,
1.97.1.1-RXN, 1.14.19.3-RXN, 1.14.99.35-RXN,
1.14.99.33-RXN, R132-RXN, R265-RXN,
TOLUENE-3-MONOOXYGENASE-RXN, 1.14.99.30-RXN,
MALERED-RXN, CHLOROMALERED-RXN,
1.3.99.15-RXN, IMIDHYD-RXN, PEROXID-RXN,
TOLUENE-4-MONOOXYGENASE-RXN,
TOLUENE-2-MONOOXYGENASE-RXN,
3-HYDROXYBENZOATE-2-MONOOXYGENASE-RXN,
1.14.99.6-RXN, 1.14.99.20-RXN,
1.14.99.15-RXN, 1.14.99.12-RXN,
STEROID-21-MONOOXYGENASE-RXN,
STEROID-17-ALPHA-MONOOXYGENASE-RXN,
STEAROYL-COA-DESATURASE-RXN,
SQUALENE-MONOOXYGENASE-RXN,
PROGESTERONE-MONOOXYGENASE-RXN,
PROGESTERONE-11-ALPHA-MONOOXYGENASE-RXN,
PLASMANYLETHANOLAMINE-DESATURASE-RXN,
N-ACETYLNEURAMINATE-MONOOXYGENASE-RXN,
METHYLSTEROL-MONOOXYGENASE-RXN,
LINOLEOYL-COA-DESATURASE-RXN,
LINALOOL-8-MONOOXYGENASE-RXN,
|KYNURENINE-7,8-HYDROXYLASE-RXN|,
JUGLONE-3-MONOOXYGENASE-RXN,
|HEME-OXYGENASE-(DECYCLIZING)-RXN|,
ESTRADIOL-6-BETA-MONOOXYGENASE-RXN,
ECDYSONE-20-MONOOXYGENASE-RXN,
DEOXYHYPUSINE-MONOOXYGENASE-RXN,
THYROXINE-DEIODINASE-RXN,
SELENOCYSTEINE-LYASE-RXN, CROBETREDUCT-RXN,
CROBETRED-RXN, THTOREDUCT-RXN,
DIMESULFREDUCT-RXN COMMON-NAME: "acceptor"

CREATION-DATE: 3051978518

CREATOR: |kr|

MOLECULAR-WEIGHT: 0.0d0

NAMES: "acceptor", "oxidized electron acceptor", "oxidized donor", "A"

SCHEMA?: T

SYNONYMS: "oxidized electron acceptor", "oxidized donor", "A"

-------------
--- Class |Donor-H2| ---
Supers: |Redox-Electron-Carriers|
Subs:
Instances:

APPEARS-IN-LEFT-SIDE-OF: VCREDCHLOR-RXN, DCEREDCHLOR-RXN,
TCEREDCHLOR-RXN, 1.14.99.21-RXN,
1.14.19.2-RXN,
STEROID-9-ALPHA-MONOOXYGENASE-RXN,
PROSTAGLANDIN-ENDOPEROXIDE-SYNTHASE-RXN,
1.14.19.1-RXN, 1.14.99.37-RXN,
2-HYDROXYPYRIDINE-5-MONOOXYGENASE-RXN,
1.97.1.1-RXN, 1.14.19.3-RXN, 1.14.99.35-RXN,
1.14.99.33-RXN, R132-RXN, R265-RXN,
TOLUENE-3-MONOOXYGENASE-RXN, 1.14.99.30-RXN,
MALERED-RXN, CHLOROMALERED-RXN, 1.3.99.15-RXN,
IMIDHYD-RXN, TOLUENE-4-MONOOXYGENASE-RXN,
TOLUENE-2-MONOOXYGENASE-RXN, PEROXID-RXN,
3-HYDROXYBENZOATE-2-MONOOXYGENASE-RXN,
1.14.99.6-RXN, 1.14.99.20-RXN, 1.14.99.15-RXN,
1.14.99.12-RXN, STEROID-21-MONOOXYGENASE-RXN,
STEROID-17-ALPHA-MONOOXYGENASE-RXN,
STEAROYL-COA-DESATURASE-RXN,
SQUALENE-MONOOXYGENASE-RXN,
PROGESTERONE-MONOOXYGENASE-RXN,
PROGESTERONE-11-ALPHA-MONOOXYGENASE-RXN,
PLASMANYLETHANOLAMINE-DESATURASE-RXN,
N-ACETYLNEURAMINATE-MONOOXYGENASE-RXN,
METHYLSTEROL-MONOOXYGENASE-RXN,
LINOLEOYL-COA-DESATURASE-RXN,
LINALOOL-8-MONOOXYGENASE-RXN,
|KYNURENINE-7,8-HYDROXYLASE-RXN|,
JUGLONE-3-MONOOXYGENASE-RXN,
|HEME-OXYGENASE-(DECYCLIZING)-RXN|,
ESTRADIOL-6-BETA-MONOOXYGENASE-RXN,
ECDYSONE-20-MONOOXYGENASE-RXN,
DEOXYHYPUSINE-MONOOXYGENASE-RXN,
THYROXINE-DEIODINASE-RXN,
SELENOCYSTEINE-LYASE-RXN, CROBETREDUCT-RXN,
CROBETRED-RXN, THTOREDUCT-RXN,
DIMESULFREDUCT-RXN

APPEARS-IN-RIGHT-SIDE-OF: GLUTARYL-COA-DEHYDROGENASE-RXN, R362-RXN,
3-HYDROXYCYCLOHEXANONE-DEHYDROGENASE-RXN,
SARCOSINE-DEHYDROGENASE-RXN, RXN-2841,
RXN0-2301, RXN0-2101, RXN0-1, 1.4.99.5-RXN,
1.20.99.1-RXN, 1.3.99.20-RXN, 1.5.99.12-RXN,
1.97.1.8-RXN, RXN-1301,
|ALCOHOL-DEHYDROGENASE-(ACCEPTOR)-RXN|,
|ALKAN-1-OL-DEHYDROGENASE-(ACCEPTOR)-RXN|,
RXN-945, CYT-ACC-RXN, R282-RXN, RXN-944,
RXN-905, RXN-865, LACTOSE-DEHYDRO-RXN,
SUCC-FUM-OXRED-A-RXN, RXN-641, R22-RXN,
R221-RXN, R161-RXN, 1.3.99.16-RXN,
1.3.99.17-RXN, 1.3.99.18-RXN, 1.3.99.19-RXN,
1.5.99.10-RXN, NITRATREDUCT-RXN,
MOXXANAU-RXN, MANDELATE-OXY-RXN,
4-HYDROXYMANDELATE-OXY-RXN, 1.17.99.1-RXN,
CARBON-MONOXIDE-DEHYDROGENASE-RXN,
FORMYLMETHANOFURAN-DEHYDROGENASE-RXN,
CARBOXYLATE-REDUCTASE-RXN, 1.2.99.3-RXN,
3-OXOSTEROID-1-DEHYDROGENASE-RXN,
3-OXO-5-BETA-STEROID-4-DEHYDROGENASE-RXN,
2-METHYLACYL-COA-DEHYDROGENASE-RXN,
2-FUROYL-COA-DEHYDROGENASE-RXN, 1.3.99.5-RXN,
DIHYDROOROTATE-DEHYDROGENASE-RXN,
CYCLOHEXANONE-DEHYDROGENASE-RXN,
BETA-CYCLOPIAZONATE-DEHYDROGENASE-RXN,
TAURINE-DEHYDROGENASE-RXN,
ARALKYLAMINE-DEHYDROGENASE-RXN,
AMINE-DEHYDROGENASE-RXN,
SPERMIDINE-DEHYDROGENASE-RXN,
TRIMETHYLAMINE-DEHYDROGENASE-RXN,
NICOTINE-DEHYDROGENASE-RXN,
L-PIPECOLATE-DEHYDROGENASE-RXN,
METHYLGUTAMATE-DEHYDROGENASE-RXN,
DIMETHYLGLYCINE-DEHYDROGENASE-RXN,
NADPH-DEHYDROGENASE-RXN,
|NADPH-DEHYDROGENASE-(QUINONE)-RXN|,
|NADH-DEHYDROGENASE-(QUINONE)-RXN|,
NITRIC-OXIDE-REDUCTASE-RXN,
HYDROXYLAMINE-REDUCTASE-RXN,
NITRITE-REDUCTASE-RXN,
NITROUS-OXIDE-REDUCTASE-RXN, 1.7.99.5-RXN,
SULFITE-REDUCTASE-RXN,
HYDROGENSULFITE-REDUCTASE-RXN,
ADENYLYLSULFATE-REDUCTASE-RXN,
PYRIDOXINE-5-DEHYDROGENASE-RXN,
D-2-HYDROXY-ACID-DEHYDROGENASE-RXN,
D-SORBITOL-DEHYDROGENASE-RXN,
DEHYDROGLUCONATE-DEHYDROGENASE-RXN,
FRUCTOSE-5-DEHYDROGENASE-RXN,
GLUCONATE-2-DEHYDROGENASE-RXN,
|GLUCOSE-DEHYDROGENASE-(ACCEPTOR)-RXN|,
1.1.99.27-RXN,
2-HYDROXYGLUTARATE-DEHYDROGENASE-RXN,
|CELLOBIOSE-DEHYDROGENASE-(ACCEPTOR)-RXN|,
URACIL-DEHYDROGENASE-RXN, 1.1.99.23-RXN,
SORBOSE-DEHYDROGENASE-RXN,
|MALATE-DEHYDROGENASE-(ACCEPTOR)-RXN|,
GLYCEROL-3-PHOSPHATE-DEHYDROGENASE-RXN,
|GLYCEROL-DEHYDROGENASE-(ACCEPTOR)-RXN|,
GLUCOSIDE-3-DEHYDROGENASE-RXN, CHD-RXN,
NADH-DEHYDROGENASE-RXN, DAADEHYDROG-RXN,
GLYCOLATEDEHYDRO-RXN

CHEMICAL-FORMULA: (H 2)

COMMON-NAME: "reduced acceptor"

CREATION-DATE: 3051978518

CREATOR: |kr|

MOLECULAR-WEIGHT: 2.016d0

NAMES: "reduced acceptor", "reduced electron acceptor", "donor", "A(H2)"

SCHEMA?: T

SYNONYMS: "reduced electron acceptor", "donor", "A(H2)"

Markus Krummenacker

unread,

Jul 7, 2005, 4:58:31 PM7/7/05

to aar...@media.mit.edu, Debuggin...@googlegroups.com, Markus Krummenacker, ca...@ai.sri.com

aar...@media.mit.edu writes:
> Hi All,
> I am working on a project with Xiaoxia Lin at the Church lab at the Harvard
> Medical School in which I am using MetaCyc and organism specific data bases in
> order to determine which reactions might be responsible for the metabolism of
> P. marinus. In order to do this I need to set up steady state flux balance
> constraints. In such a system reactions with |NAD(P)| and |NAD(P)H| will not
> balance with reactions that specifically have NAD, NADP, NADH and
> NADPH. I have
> written software to create specific reactions to replace the general
> metabolites
> with the specific ones. (It would create two new reactions in the preceding
> case.)

Hi Aaron,

This makes a lot of sense, and it is easy in this case.

> Here's my question: What about reactions with |Acceptor| and |Donor-H2| .
> These are special cases of |Redox-Electron-Carriers|.
> Is it always safe to assume NAD, NADP, FAD, FMN, and NADH, NADPH, FADH2, FMNH2
> can be substituted for these?
> Is there more specific info available to resolve this?

From my overall understanding of reaction equations coming from the
ENZYME database, I unfortunately get the impression that often the
specific nature of the redox cofactors is unknown. So I think there
usually will not be an easy solution that can be applied automatically
on a large scale.

Presumably, if humans do careful reviews of the current literature,
some of these enzymes may have been characterized better by now, such
that some publication could be found that settles the case. But this
involves a lot of work. Our MetaCyc and EcoCyc curators perform these
types of work, but they already have long todo lists.

So I think until all uncertainties are cleared up (if this will ever
happen :-) , I think it is important to factor in some amount of
uncertainty in simulations. The |Acceptor| and |Donor-H2| pair is
intended as a collection place for those reactions where the detailed
redox factor is not known yet in our databases. It should be possible
to just use this pair like you would use some other pair of redox
factors. For mass balancing purposes, you may assume that |Donor-H2|
supplies an H2 equivalent.

However, I think expanding out such reactions to all versions using
all known redox factors on the chance that one of them has to be the
right one is probably dangerous, and less correct than to leave the
uncertainty specificed explicitly. (So just to clarify, |Acceptor|
and |Donor-H2| are generally used when the exact factor is unknown,
not when the enzyme is known to have such broad specificity to
accept every known redox factor.)

I have copied in Ron Caspi, just in case he would want to correct my
assessment.

--
--
Regards
Markus Krummenacker

gh...@york.ac.uk

unread,

Jul 7, 2005, 5:58:28 PM7/7/05

to Debuggin...@googlegroups.com, aar...@media.mit.edu, Markus Krummenacker

Dear Aaron, As a biochemist, I would say that you certainly cannot assume
that the alternatives you suggest can be added to these reactions 'safely'.

Again, certain electron transport chains use specific quinones and you
cannot assume that any of these will operate with a particular
dehydrogenase/reductase. However, the quninone associations have only
really been worked out for a few bugs like E. coli, so it is difficult to
be able to assign them for an uncharacterised organism.

Cheers,
Gavin Thomas.

>
> Hi All, I am working on a project with Xiaoxia Lin at the Church lab at
> the Harvard Medical School in which I am using MetaCyc and organism
> specific data bases in order to determine which reactions might be
> responsible for the metabolism of P. marinus. In order to do this I need
> to set up steady state flux balance constraints. In such a system
> reactions with |NAD(P)| and |NAD(P)H| will not balance with reactions
> that specifically have NAD, NADP, NADH and NADPH. I have written software
> to create specific reactions to replace the general metabolites with the
> specific ones. (It would create two new reactions in the preceding case.)
> Here's my question: What about reactions with |Acceptor| and |Donor-H2| .
> These are special cases of |Redox-Electron-Carriers|. Is it always safe
> to assume NAD, NADP, FAD, FMN, and NADH, NADPH, FADH2, FMNH2 can be
> substituted for these? Is there more specific info available to resolve
> this? What about the other Redox-Electron-Carriers?
>
>
>

Reply all

Reply to author

Forward