sbml-qual Mailing List for Systems Biology Markup Language (SBML)

Thanks very much, Claudine.

On Wed, Jan 31, 2018 at 10:12 AM, Claudine Chaouiya <
cha...@ig...> wrote:

> Hi Sarah,
>
> Thank you Justin for rising this issue...
> Indeed, for logical models this rule should be respected because their
> should not be several transitions defining the regulation of a component…
> we kept this as a recommendation because for Petri net models, it would be
> possible to have distinct transitions sharing the same output.
> I would thus keep it as a recommendation…
> Best regards
> Claudine
>
>
> > On 29 Jan 2018, at 09:17, Sarah Keating <ske...@ca...> wrote:
> >
> > Hi Qual people
> >
> > We received a query (see below) from a libsbml user (Justin McManus
> cc'ed) about an anomaly between the qual specification and the libsbml
> implementation:
> >
> > We have rule qual-20311:
> >
> > A QualitativeSpecies that is referenced by an Output with the
> qual:transitionEffect at-tribute set to “assignmentLevel” cannot be
> referenced by any other Output with the same transitionEffect throughout
> the set of transitions for the containing model.
> >
> > In the spec it is marked as 'recommendation' but libSBML flags it as an
> 'error' - i.e. must not be violated.
> >
> > A recommendation would suggest that it is a good idea to stick to this
> rule but it should not cause problems if you violate it.
> >
> > Before changing libSBML code I wanted to check with people more familiar
> with these models as to whether the rule is fine as a recommendation :-)
> >
> > i.e. which is right -  the specification or libSBML implementation ?? :-)
> >
> > Thanks
> >
> > Sarah
> >
> > -------- Forwarded Message --------
> > Subject: Re: [libsbml-team] Incorrect entry in QualSBMLErrorTable.h
> version 5.16.0
> > Date: Fri, 26 Jan 2018 09:01:05 +0000
> > From: Sarah Keating <ske...@ca...>
> > To: Justin McManus <ju...@ka...>, lib...@go...
> >
> > On 25/01/2018 15:07, Justin McManus wrote:
> >> To the SBML developer team,
> >> First, thank you so much for the profoundly important software you
> maintain.
> >> I noticed one issue that I'd like to report. In what I believe is the
> latest specification of the qual package for SBML, Rule qual-20311 is
> listed as a /recommendation/, not a /requirement/. However, the source code
> (libsbml v 5.16.0) treats this rule as a requirement, and will result in an
> error (rather than a warning) when the recommendation is violated.
> >> In src/sbml/packages/qual/validator, please see QualSBMLErrorTable.h.
> Specifically, please see the entry for QualQSAssignedOnlyOnce, which
> corresponds to this rule (Rule qual-20311) and error code 3020311. This
> entry treats the occurrence of this flag as a LIBSBML_SEV_ERROR rather than
> LIBSBML_SEV_WARNING.
> >> I first noticed the issue while using the R package rsbml, which relies
> on the libSBML, and which throws the error encoded in the libsbml when it
> reads an SBML containing multiple transition assignments for the same
> qualitative species.
> >> Please note that I did /not/ run the library through a debugger to
> confirm that the issue in QualSBMLErrorTable.h is the cause of the
> misspecified behavior I describe above. That's merely my guess based on a
> cursory look at the source code.
> >> Warm regards, and thank you for considering my comments here,
> >> Justin
> >> --
> >> Justin McManus, Ph.D.
> >> Senior Scientist
> >> Lead Computational Biologist
> >> Kallyope, Inc.
> >> 430 East 29th Street, Suite 1050
> >> New York, NY 10016
> >> (646) 596-3471
> >> --
> >> You received this message because you are subscribed to the Google
> Groups "libsbml-team" group.
> >> To unsubscribe from this group and stop receiving emails from it, send
> an email to lib...@go... <mailto:
> lib...@go...>.
> >> For more options, visit https://groups.google.com/d/optout.
> >
> >
> > ------------------------------------------------------------
> ------------------
> > Check out the vibrant tech community on one of the world's most
> > engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> > _______________________________________________
> > sbml-qual mailing list
> > sbm...@li...
> > https://lists.sourceforge.net/lists/listinfo/sbml-qual
>
>


-- 
Justin McManus, Ph.D.
Senior Scientist
Lead Computational Biologist
Kallyope, Inc.
430 East 29th Street, Suite 1050
New York, NY 10016
(646) 596-3471

Hi Sarah,

Thank you Justin for rising this issue...
Indeed, for logical models this rule should be respected because their should not be several transitions defining the regulation of a component… we kept this as a recommendation because for Petri net models, it would be possible to have distinct transitions sharing the same output. 
I would thus keep it as a recommendation…
Best regards
Claudine

 
> On 29 Jan 2018, at 09:17, Sarah Keating <ske...@ca...> wrote:
> 
> Hi Qual people
> 
> We received a query (see below) from a libsbml user (Justin McManus cc'ed) about an anomaly between the qual specification and the libsbml implementation:
> 
> We have rule qual-20311:
> 
> A QualitativeSpecies that is referenced by an Output with the qual:transitionEffect at-tribute set to “assignmentLevel” cannot be referenced by any other Output with the same transitionEffect throughout the set of transitions for the containing model.
> 
> In the spec it is marked as 'recommendation' but libSBML flags it as an 'error' - i.e. must not be violated.
> 
> A recommendation would suggest that it is a good idea to stick to this rule but it should not cause problems if you violate it.
> 
> Before changing libSBML code I wanted to check with people more familiar with these models as to whether the rule is fine as a recommendation :-)
> 
> i.e. which is right -  the specification or libSBML implementation ?? :-)
> 
> Thanks
> 
> Sarah
> 
> -------- Forwarded Message --------
> Subject: Re: [libsbml-team] Incorrect entry in QualSBMLErrorTable.h version 5.16.0
> Date: Fri, 26 Jan 2018 09:01:05 +0000
> From: Sarah Keating <ske...@ca...>
> To: Justin McManus <ju...@ka...>, lib...@go...
> 
> On 25/01/2018 15:07, Justin McManus wrote:
>> To the SBML developer team,
>> First, thank you so much for the profoundly important software you maintain.
>> I noticed one issue that I'd like to report. In what I believe is the latest specification of the qual package for SBML, Rule qual-20311 is listed as a /recommendation/, not a /requirement/. However, the source code (libsbml v 5.16.0) treats this rule as a requirement, and will result in an error (rather than a warning) when the recommendation is violated.
>> In src/sbml/packages/qual/validator, please see QualSBMLErrorTable.h. Specifically, please see the entry for QualQSAssignedOnlyOnce, which corresponds to this rule (Rule qual-20311) and error code 3020311. This entry treats the occurrence of this flag as a LIBSBML_SEV_ERROR rather than LIBSBML_SEV_WARNING.
>> I first noticed the issue while using the R package rsbml, which relies on the libSBML, and which throws the error encoded in the libsbml when it reads an SBML containing multiple transition assignments for the same qualitative species.
>> Please note that I did /not/ run the library through a debugger to confirm that the issue in QualSBMLErrorTable.h is the cause of the misspecified behavior I describe above. That's merely my guess based on a cursory look at the source code.
>> Warm regards, and thank you for considering my comments here,
>> Justin
>> -- 
>> Justin McManus, Ph.D.
>> Senior Scientist
>> Lead Computational Biologist
>> Kallyope, Inc.
>> 430 East 29th Street, Suite 1050
>> New York, NY 10016
>> (646) 596-3471
>> -- 
>> You received this message because you are subscribed to the Google Groups "libsbml-team" group.
>> To unsubscribe from this group and stop receiving emails from it, send an email to lib...@go... <mailto:lib...@go...>.
>> For more options, visit https://groups.google.com/d/optout.
> 
> 
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> sbml-qual mailing list
> sbm...@li...
> https://lists.sourceforge.net/lists/listinfo/sbml-qual

Hi Qual people

We received a query (see below) from a libsbml user (Justin McManus 
cc'ed) about an anomaly between the qual specification and the libsbml 
implementation:

We have rule qual-20311:

A QualitativeSpecies that is referenced by an Output with the 
qual:transitionEffect at-tribute set to “assignmentLevel” cannot be 
referenced by any other Output with the same transitionEffect throughout 
the set of transitions for the containing model.

In the spec it is marked as 'recommendation' but libSBML flags it as an 
'error' - i.e. must not be violated.

A recommendation would suggest that it is a good idea to stick to this 
rule but it should not cause problems if you violate it.

Before changing libSBML code I wanted to check with people more familiar 
with these models as to whether the rule is fine as a recommendation :-)

i.e. which is right -  the specification or libSBML implementation ?? :-)

Thanks

Sarah

-------- Forwarded Message --------
Subject: Re: [libsbml-team] Incorrect entry in QualSBMLErrorTable.h 
version 5.16.0
Date: Fri, 26 Jan 2018 09:01:05 +0000
From: Sarah Keating <ske...@ca...>
To: Justin McManus <ju...@ka...>, lib...@go...

On 25/01/2018 15:07, Justin McManus wrote:
> To the SBML developer team,
> 
> First, thank you so much for the profoundly important software you 
> maintain.
> 
> I noticed one issue that I'd like to report. In what I believe is the 
> latest specification of the qual package for SBML, Rule qual-20311 is 
> listed as a /recommendation/, not a /requirement/. However, the source 
> code (libsbml v 5.16.0) treats this rule as a requirement, and will 
> result in an error (rather than a warning) when the recommendation is 
> violated.
> 
> In src/sbml/packages/qual/validator, please see QualSBMLErrorTable.h. 
> Specifically, please see the entry for QualQSAssignedOnlyOnce, which 
> corresponds to this rule (Rule qual-20311) and error code 3020311. This 
> entry treats the occurrence of this flag as a LIBSBML_SEV_ERROR rather 
> than LIBSBML_SEV_WARNING.
> 
> I first noticed the issue while using the R package rsbml, which relies 
> on the libSBML, and which throws the error encoded in the libsbml when 
> it reads an SBML containing multiple transition assignments for the same 
> qualitative species.
> 
> Please note that I did /not/ run the library through a debugger to 
> confirm that the issue in QualSBMLErrorTable.h is the cause of the 
> misspecified behavior I describe above. That's merely my guess based on 
> a cursory look at the source code.
> 
> Warm regards, and thank you for considering my comments here,
> 
> Justin
> 
> -- 
> Justin McManus, Ph.D.
> Senior Scientist
> Lead Computational Biologist
> Kallyope, Inc.
> 430 East 29th Street, Suite 1050
> New York, NY 10016
> (646) 596-3471
> 
> -- 
> You received this message because you are subscribed to the Google 
> Groups "libsbml-team" group.
> To unsubscribe from this group and stop receiving emails from it, send 
> an email to lib...@go... 
> <mailto:lib...@go...>.
> For more options, visit https://groups.google.com/d/optout.