You can subscribe to this list here.
2005 |
Jan
|
Feb
|
Mar
|
Apr
(2) |
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
---|---|---|---|---|---|---|---|---|---|---|---|---|
2006 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
(4) |
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
From: Padraig G. <p.g...@uc...> - 2006-07-04 17:09:21
|
While it should be obvious for most quantities what the SI units should be, we could specify these somewhere for completeness. We could use something like: ** <xs:attribute name="rise_time" type="xs:double" use="required"> <xs:annotation> <xs:documentation>The characteristic rise time of the conductance waveform <meta:units>s</meta:units> </xs:documentation> </xs:annotation> </xs:attribute> in the schema document. This approach wouldn't be too intrusive in the ChannelML files themselves, but looking at the HTML view of the schema will clear up any confusion on units (and will provide the extra documentation). The question is then if there is a units framework/schema we can reuse, which leads to the question of whether to follow the SBML or BrainML approach, a question which can be addressed at CNS. Regards, Padraig Josef Svitak wrote: > XML should generally be self-documenting, though I agree that a few small > comments can go a long way in improving human readability. > > My main concern is that we be able to compare apples to apples. While I'd be > hard-pressed to find an alternate unit for an electrical charge, isn't it > conceivable that conductances might be expressed in some Ye Olde English > (non-metric) unit? So, what I'm really wondering is if that unit actually > appears somewhere in the (or an associated) XML doc. > > joe > > --- Padraig Gleeson <p.g...@uc...> wrote: > > >> Think it's a chan_dist issue... >> >> The channels element specified that units are Physiological Units. This >> means that within a channels spec all units are consistent, and you >> don't have to add extra unit information each time you add a value. It >> should be clear from the definition of the elements/attributes what type >> of physical quantity it is. >> >> The comment is only an aid for readability. This is to move away from >> XML files intended for machines only, for NEURON/GENESIS users, who'll >> want to poke around in the source files. >> >> Padraig >> >> >> >> >> Josef Svitak wrote: >> >>> Ahoy! >>> >>> I noticed: >>> >>> <cml:ion name="non_specific" charge="1" default_erev="-54.3"/> <!-- phys >>> units: mV --> >>> >>> in the CellWithChannels sample file. Does the unit information in the >>> >> comment >> >>> have a (more visible) home in the XML hierarchy? There is other unit info a >>> couple lines farther down in this file as well. >>> >>> Also, this message doesn't quite fit neatly under any of the mailing list >>> topics. Do we need a catch-all list? >>> >>> joe >>> >>> js...@ya... >>> Software Engineer >>> Linux/OSX C/C++/Java >>> >>> Using Tomcat but need to do more? Need to support web services, security? >>> Get stuff done quickly with pre-integrated technology to make your job >>> >> easier >> >>> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo >>> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 >>> _______________________________________________ >>> Neuroml-chan_dist mailing list >>> Neu...@li... >>> https://lists.sourceforge.net/lists/listinfo/neuroml-chan_dist >>> >>> >> Using Tomcat but need to do more? Need to support web services, security? >> Get stuff done quickly with pre-integrated technology to make your job easier >> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo >> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 >> _______________________________________________ >> Neuroml-chan_dist mailing list >> Neu...@li... >> https://lists.sourceforge.net/lists/listinfo/neuroml-chan_dist >> >> > > > js...@ya... > Software Engineer > Linux/OSX C/C++/Java > > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Neuroml-chan_dist mailing list > Neu...@li... > https://lists.sourceforge.net/lists/listinfo/neuroml-chan_dist > |
From: Hugo C. <hug...@gm...> - 2006-07-03 22:14:41
|
I took a quick glance at the network level xsd. Some questions and suggestions : 1. In NeuroML, projections can only be specified between populations (that's what the docs say). Neurospaces allows projections between networks too (I will demo it at the CNS workshop). Is there a specific reason not to allow projections between networks ? 2. The element 'Instances' is about cell instances only. Is the intent to allow 'PopulationInstances' and 'NetworkInstances' in the future ? 3. I suggest to be very clear about the difference between a 'network connection' and a 'cell-to-cell connection'. I guessed that the element 'Connection' is about a 'cell-to-cell' connection, yet it is documented as a 'network connection'. 3. Synapses : 3.1. Naming : synapses are specified using (1) at the population level 'from' and 'to' populations, and (2) at the cell level 'source' and 'target' elements. I think that 'from' and 'to' are to generic to be used and that in the literature 'source' and 'target' are commonly associated with populations or networks. I suggest using 'source' and 'target' for the network level and 'pre' and 'post' for the pre- and post-synaptics at the cell level respectively. That is also the way it is done in Neurospaces. 3.2. Currently lacking : the synaptic location element must allow to specify a post-synaptic channel. 3.3. Is it right that SynapseProperties can occur only once in a projection ? Is it right that in the future this element will include a distinction between pre- and post- synaptic delay, and delay due to the synaptic cleft ? Overall it looks very consistent, let's discuss this further at the CNS meeting. Hugo On 6/29/06, Padraig Gleeson <p.g...@uc...> wrote: > > Hi, > > A new version of the NeuroML specs has been put live on the website. > It's termed v1.3alpha as there may be some changes in this after CNS, > and the new network part (Level 3) is quite basic. The changes from v1.2 > are: > > MorphML/Metadata: > No significant changes here, just some new simple types. > > ChannelML: > No large differences, the HTML representation of the channel information > has been updated. Ion role has been changed to PermeatedSubstance, > ModulatingSubstance and SignallingSubstance > > Biophysics: > The information on the biophysical aspects of the model cells has been > restructured to a more logical layout (e.g. capacitance is no longer > described as a mechanism) and more information on what's needed for a > full cell model in NEURON/GENESIS is included. > > NetworkML: > A new network specification is included. It's an initial suggestion for > how to specify the cell populations and network connections, and I'm > sure there will be a lot of input needed from anatomists and those who > deal with point process neural networks. A key point is that there > should be two ways to describe cell populations/network connections, > based on instances (where all cell positions are specified, as are > connections, with details on which segments contain the synapses) and > templates (where a population of a specific number is specified along > with a region in which the cells are found, and patterns describe how > the populations should be connected). > > Feedback is encouraged and I'm sure there will be some interesting > discussions on them at CNS. > > Regards, > Padraig > > > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job > easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Neuroml-chan_dist mailing list > Neu...@li... > https://lists.sourceforge.net/lists/listinfo/neuroml-chan_dist > -- Hugo Cornelis Ph.D. Research Imaging Center University of Texas Health Science Center at San Antonio 7703 Floyd Curl Drive San Antonio, TX 78284-6240 Phone: 210 567 8112 Fax: 210 567 8152 |
From: Josef S. <js...@ya...> - 2006-06-30 15:28:16
|
XML should generally be self-documenting, though I agree that a few small comments can go a long way in improving human readability. My main concern is that we be able to compare apples to apples. While I'd be hard-pressed to find an alternate unit for an electrical charge, isn't it conceivable that conductances might be expressed in some Ye Olde English (non-metric) unit? So, what I'm really wondering is if that unit actually appears somewhere in the (or an associated) XML doc. joe --- Padraig Gleeson <p.g...@uc...> wrote: > Think it's a chan_dist issue... > > The channels element specified that units are Physiological Units. This > means that within a channels spec all units are consistent, and you > don't have to add extra unit information each time you add a value. It > should be clear from the definition of the elements/attributes what type > of physical quantity it is. > > The comment is only an aid for readability. This is to move away from > XML files intended for machines only, for NEURON/GENESIS users, who'll > want to poke around in the source files. > > Padraig > > > > > Josef Svitak wrote: > > Ahoy! > > > > I noticed: > > > > <cml:ion name="non_specific" charge="1" default_erev="-54.3"/> <!-- phys > > units: mV --> > > > > in the CellWithChannels sample file. Does the unit information in the > comment > > have a (more visible) home in the XML hierarchy? There is other unit info a > > couple lines farther down in this file as well. > > > > Also, this message doesn't quite fit neatly under any of the mailing list > > topics. Do we need a catch-all list? > > > > joe > > > > js...@ya... > > Software Engineer > > Linux/OSX C/C++/Java > > > > Using Tomcat but need to do more? Need to support web services, security? > > Get stuff done quickly with pre-integrated technology to make your job > easier > > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > > _______________________________________________ > > Neuroml-chan_dist mailing list > > Neu...@li... > > https://lists.sourceforge.net/lists/listinfo/neuroml-chan_dist > > > > > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Neuroml-chan_dist mailing list > Neu...@li... > https://lists.sourceforge.net/lists/listinfo/neuroml-chan_dist > js...@ya... Software Engineer Linux/OSX C/C++/Java |
From: Padraig G. <p.g...@uc...> - 2006-06-30 11:42:36
|
Think it's a chan_dist issue... The channels element specified that units are Physiological Units. This means that within a channels spec all units are consistent, and you don't have to add extra unit information each time you add a value. It should be clear from the definition of the elements/attributes what type of physical quantity it is. The comment is only an aid for readability. This is to move away from XML files intended for machines only, for NEURON/GENESIS users, who'll want to poke around in the source files. Padraig Josef Svitak wrote: > Ahoy! > > I noticed: > > <cml:ion name="non_specific" charge="1" default_erev="-54.3"/> <!-- phys > units: mV --> > > in the CellWithChannels sample file. Does the unit information in the comment > have a (more visible) home in the XML hierarchy? There is other unit info a > couple lines farther down in this file as well. > > Also, this message doesn't quite fit neatly under any of the mailing list > topics. Do we need a catch-all list? > > joe > > js...@ya... > Software Engineer > Linux/OSX C/C++/Java > > Using Tomcat but need to do more? Need to support web services, security? > Get stuff done quickly with pre-integrated technology to make your job easier > Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo > http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 > _______________________________________________ > Neuroml-chan_dist mailing list > Neu...@li... > https://lists.sourceforge.net/lists/listinfo/neuroml-chan_dist > |
From: Josef S. <js...@ya...> - 2006-06-29 19:11:48
|
Ahoy! I noticed: <cml:ion name="non_specific" charge="1" default_erev="-54.3"/> <!-- phys units: mV --> in the CellWithChannels sample file. Does the unit information in the comment have a (more visible) home in the XML hierarchy? There is other unit info a couple lines farther down in this file as well. Also, this message doesn't quite fit neatly under any of the mailing list topics. Do we need a catch-all list? joe js...@ya... Software Engineer Linux/OSX C/C++/Java |
From: Padraig G. <p.g...@uc...> - 2006-06-29 18:16:35
|
Hi, A new version of the NeuroML specs has been put live on the website. It's termed v1.3alpha as there may be some changes in this after CNS, and the new network part (Level 3) is quite basic. The changes from v1.2 are: MorphML/Metadata: No significant changes here, just some new simple types. ChannelML: No large differences, the HTML representation of the channel information has been updated. Ion role has been changed to PermeatedSubstance, ModulatingSubstance and SignallingSubstance Biophysics: The information on the biophysical aspects of the model cells has been restructured to a more logical layout (e.g. capacitance is no longer described as a mechanism) and more information on what's needed for a full cell model in NEURON/GENESIS is included. NetworkML: A new network specification is included. It's an initial suggestion for how to specify the cell populations and network connections, and I'm sure there will be a lot of input needed from anatomists and those who deal with point process neural networks. A key point is that there should be two ways to describe cell populations/network connections, based on instances (where all cell positions are specified, as are connections, with details on which segments contain the synapses) and templates (where a population of a specific number is specified along with a region in which the cells are found, and patterns describe how the populations should be connected). Feedback is encouraged and I'm sure there will be some interesting discussions on them at CNS. Regards, Padraig |
From: <neu...@li...> - 2005-04-29 14:20:56
|
Just to let everyone know at what point the NEURON developers are starting from, here is the tentative xml that ModelView is emitting for the neurondemo stylized cell. Since the file is 549 lines long, instead of enclosing it, I have put it at http://www.neuron.yale.edu/ftp/hines/tmp/stylized.xml You may have to look at the page source since I see that that my browser complains that it is not well formed. The bulk of the file is the morphml points and setOfSegments that correspond to NEURON sections. However, when you get to our tentative <biophysics> text you can see the definition of subsets of sections (ModelView currently has no way to give reasonable names to subsets) which are then used to compactly specify parameters. I should mention that the CellBuilder now can specify inhomogeneous parameters by means of subset domain iterators. That is, parameters in one or more sections can vary with location according to a function of some independent variable (the "subset domain iterator"). Subset domain iterators provided are: arc distance from soma, physical distance (radial distance) from a point, and projection onto a straight line (distance along an axis). I believe these correspond to NeuroML mask instances. -Michael Hines |
From: <neu...@li...> - 2005-04-28 21:23:04
|
Hello all, As you all know, since the WAM-BAMM meeting, some of us have been = revisiting MorphML to extend it in ways that will better accomodate everyone's = needs.=20 Tom Morse is also taking the lead on revising "ChannelML" to address = some of the issues that we discussed at the meeting...separating channel info = from ion=20 concentration info, converging on an agreement of how to represent info = about the (in)activation functions for gating variables, etc. Very shortly we will need a way to connect these two things (morphology = and channels/ions) in a way that specifies the distribution of channels = throughout the cell membrane. I propose that we get started by looking at the way that = this is done in neuron and genesis. I hope that Graham Cummins will chime in (if I can twist his arm to get him signed up here) for the neurosys = group and his parameter searching studies. Are there other people who we can = contact to see if we can get their input? We need to keep in mind that if we put = in the time=20 right now to be inclusive and as general as possible, then we will end = up with fewer revisions down the road. Best, Sharon |