neuroml-python Mailing List for NeuroML

Brought to you by: adrianq, andrew-davison, borismarin, fredhowell, and 4 others

neuroml-python — All things Python related, particularly the Python API for NeuroML

You can subscribe to this list here.

2012	_Jan	_Feb	_Mar	_Apr (1)	_May (21)	_Jun (11)	_Jul (1)	_Aug	_Sep (1)	_Oct	_Nov	_Dec
2013	_Jan	_Feb (2)	_Mar (1)	_Apr (2)	_May	_Jun	_Jul	_Aug	_Sep (1)	_Oct	_Nov (9)	_Dec
2016	_Jan	_Feb (1)	_Mar	_Apr	_May	_Jun	_Jul	_Aug	_Sep	_Oct	_Nov	_Dec
2017	_Jan	_Feb	_Mar	_Apr	_May	_Jun	_Jul	_Aug	_Sep	_Oct (3)	_Nov	_Dec

Flat | Threaded

1 2 3 > >> (Page 1 of 3)

Re: [Neuroml-python] How 2 convert SBML to neuron using python

From: Padraig G. <p.g...@uc...> - 2017-10-16 14:03:04

Hi Robert,

Yes, that seems to be the best way to do it via Python. It does of 
course require that Java is set up correctly in the system, but that's 
required for getting most of the jNeuroML specific functionality in 
pyNeuroML.

Regards,

Padraig



On 13/10/17 12:42, Robert Lindroos wrote:
> Hi again.
> I solved it using the pynml.py description:
>
> https://github.com/NeuroML/pyNeuroML/blob/master/pyneuroml/pynml.py
>
> For reference it was the run_jneuroml() function that should be used 
> for this task.
>
> Here's an example script:
>
> from pyneuroml import pynml
> pynml.run_jneuroml('-sbml-import', 'SBML_file_name.xml', '1 1' )      
>      # creates a LEMS file called SBML_file_name_LEMS.xml
> pynml.run_jneuroml('', 'SBML_file_name_LEMS.xml', '-neuron')          
>    # creates a NEURON .mod file
>
>
> Best,
> R
>
> 2017-10-12 22:02 GMT+02:00 Robert Lindroos <rob...@gm... 
> <mailto:rob...@gm...>>:
>
>     Hi,
>     I am able to export an SBML file to NEURON readable mod file using
>     jnml in the following way.
>
>     Import;
>     ./jnml -sbml-import SBML.xlm 1 1
>
>     Export;
>     ./jnml SBML.xlm_LEMS.xml -neuron
>
>     But how can I do this with the python version?
>
>     Preferably the names of the substrates in the SBML file would be
>     used in the mod file as well if possible (instead of the ID that
>     seems to be the standard). Ultimately I would need this to work in
>     a jupyter notebook.
>
>
>     Thanks and regards,
>     Robert
>
>
>
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
>
>
> _______________________________________________
> Neuroml-python mailing list
> Neu...@li...
> https://lists.sourceforge.net/lists/listinfo/neuroml-python

-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

Re: [Neuroml-python] How 2 convert SBML to neuron using python

From: Robert L. <rob...@gm...> - 2017-10-13 11:42:47

Hi again.
I solved it using the pynml.py description:

https://github.com/NeuroML/pyNeuroML/blob/master/pyneuroml/pynml.py

For reference it was the run_jneuroml() function that should be used for
this task.

Here's an example script:

from pyneuroml import pynml
pynml.run_jneuroml('-sbml-import', 'SBML_file_name.xml', '1 1' )
 # creates a LEMS file called SBML_file_name_LEMS.xml
pynml.run_jneuroml('', 'SBML_file_name_LEMS.xml', '-neuron')             #
creates a NEURON .mod file


Best,
R

2017-10-12 22:02 GMT+02:00 Robert Lindroos <rob...@gm...>:

> Hi,
> I am able to export an SBML file to NEURON readable mod file using jnml in
> the following way.
>
> Import;
> ./jnml -sbml-import SBML.xlm 1 1
>
> Export;
> ./jnml SBML.xlm_LEMS.xml -neuron
>
> But how can I do this with the python version?
>
> Preferably the names of the substrates in the SBML file would be used in
> the mod file as well if possible (instead of the ID that seems to be the
> standard). Ultimately I would need this to work in a jupyter notebook.
>
>
> Thanks and regards,
> Robert
>
>

[Neuroml-python] How 2 convert SBML to neuron using python

From: Robert L. <rob...@gm...> - 2017-10-12 20:03:04

Hi,
I am able to export an SBML file to NEURON readable mod file using jnml in
the following way.

Import;
./jnml -sbml-import SBML.xlm 1 1

Export;
./jnml SBML.xlm_LEMS.xml -neuron

But how can I do this with the python version?

Preferably the names of the substrates in the SBML file would be used in
the mod file as well if possible (instead of the ID that seems to be the
standard). Ultimately I would need this to work in a jupyter notebook.


Thanks and regards,
Robert

[Neuroml-python] generateDS.py now supports Python 3

From: Dave K. <dku...@da...> - 2016-02-16 22:26:05

For future reference, you may be interested in the following.

There is a new version of generateDS.py -- version 2.19b.

This version has improved support for Python 3.  You can now run
generateDS.py with either Python 2 or Python 3, *and* the same
generated file can be run with either Python 2 or Python 3.
generateDS.py no longer generates different code for Python 2 and
Python 3.  The "--py3" command line option has been removed.

I apologize for not generating code that will run under either
Python 2 or 3 in the previous version.  I thought doing that would
be much more difficult than it turned out to be.

You can find it here:

- Python Package Index -- http://pypi.python.org/pypi/generateDS/
- Source Forge -- http://sourceforge.net/projects/generateds/
- Bitbucket -- For those of you who prefer using Mercurial, there is
  also a Mercurial repository at Bitbucket:
  https://bitbucket.org/dkuhlman/generateds

Below are a few notes and details from the README.

If you have comments, suggestions, or problems, please send them
along.  The email list is here:
https://lists.sourceforge.net/lists/listinfo/generateds-users

Dave

# =========================================================

Version 2.19b (02/16/2016)

- Modified generated code so that it will run under both Python 2
  and Python 3.  There is no longer any need to generate different
  code for Python 2 and Python 3.  If fact, the "--py3" command line
  option has been removed.

# =========================================================

-- 

Dave Kuhlman
http://www.davekuhlman.org

Re: [Neuroml-python] [Neuroml-technology] Defining Excitatory/Inhibitory Neurons or Synapses

From: Padraig G. <p.g...@uc...> - 2013-11-25 16:50:22

Hi,

The size of the PSP isn't a primary parameter in any of the cell models 
supported by NeuroML 2, it will be a function of the input resistance of 
the cell, the conductance of the synapse and time courses of the 
membrane & synapse. The best way to see what PSP a synaptic input 
produces would be to create a simple model with I&F cells, and run it. 
The PSP should be fairly linearly scalable by adjusting the max 
conductance of the synapse.

Alternatively you could build a new cell model in LEMS which adjusts its 
membrane potential by a fixed amount on receiving an event (and has a 
leak & thresholding mechanism like the standard I&F) but this would be 
of limited use in standard NeuroML 2 networks, as these are based around 
a separated synapse/cell model.

Regards,
Padraig



On 21/11/13 16:51, Önder Gürcan wrote:
> Hello Padraig,
>
> Thanks a lot for your explanation. The objective of my simulator is to 
> find the right PSP values for the synapses. I am then planning to 
> export the resulting network to NeuroML. However, according to you 
> explanation I suppose I need to make another parameter adjustment in 
> order to find the PSP values I have already found. Isn't there any 
> other ways finding out them? Or isn't there any calculation about how 
> compute it?
>
> Best,
>
> Önder.
> http://das.ege.edu.tr/~ogurcan/ <http://das.ege.edu.tr/%7Eogurcan/>
>
>
> On Wed, Nov 20, 2013 at 6:34 PM, Padraig Gleeson 
> <uc...@li... <mailto:uc...@li...>> wrote:
>
>     Hi Önder,
>
>     The size of the PSP will depend on the transient conductance (or
>     current) of the synapse as well as the the cell model you choose,
>     particularly what its input resistance is. It might be best to
>     choose a simple integrate and fire cell model, and a conductance
>     based synapse, create a network with 2 cells, and see what the
>     voltage response is when one spikes and causes a PSP in the other,
>     and adjust your parameters accordingly.
>
>     Padraig
>
>
>
>
>
>     On 18/11/13 22:11, Önder Gürcan wrote:
>>     Hello Padraig,
>>
>>     Thanks a lot for your answer and sorry for my ignorance about the
>>     topic.
>>
>>     In my simulator, I also modelled excitatory and inhibitory
>>     neurone like you said. However, as far as I know from [Iansek and
>>     Redman, 1973] (attached),  the post-synaptic potential (PSP) for
>>     a unitary synapse can range from 0.07 mV to 0.60 mV, but mostly
>>     it is between 0.10 and 0.20 mV (see Figure 4 from [Iansek and
>>     Redman, 1973] attached also).
>>
>>     In this case, could you please tell me how to define a 0.1 mV PSP
>>     and -0.1 mV PSP in NeuroML?
>>
>>     Thanks in advance for helps,
>>
>>     Best regards,
>>
>>     Önder.
>>
>>     [Iansek and Redman, 1973] Iansek, R. and Redman, S. (1973). The
>>     amplitude, time course and charge of unitary post-synaptic
>>     potentials evoked in spinal motoneurone dendrites. J. of
>>     Neurophysiol., 234:665 – 688.
>>
>>
>>
>>
>>     On Mon, Nov 18, 2013 at 1:01 PM, Padraig Gleeson
>>     <p.g...@uc... <mailto:p.g...@uc...>> wrote:
>>
>>         Hi Önder,
>>
>>         In NeuroML (1&2) there is a separation between the
>>         specification of the cell properties and the synapses. A cell
>>         will be specified in terms of its structure (segments etc.)
>>         and the locations of ion channel densities (in v1 or 2) or as
>>         an abstract cell (izhikevichCell, adExCell, iafCell, etc.
>>         from v2.0), and the synapse model to use will be set when
>>         connections are made between populations of the cells, e.g.
>>         see here:
>>
>>         https://github.com/NeuroML/NeuroML2/blob/master/examples/NML2_InstanceBasedNetwork.nml
>>
>>         A cell is generally considered "excitatory" or "inhibitory"
>>         if the reversal potentials of the synapses with which it
>>         sends input to other cells are around 0mV and below -70mV
>>         respectively (though there are exceptions to these values,
>>         and an inhibitory synapse can be excitatory for a v
>>         hyperpolarised target cell).
>>
>>         Hope this helps.
>>
>>         Padraig
>>
>>
>>
>>         On 16/11/13 21:46, Önder Gürcan wrote:
>>>         Hello all,
>>>
>>>         May be it is a very simple question for you but I was unable
>>>         to find how to define excitatory and inhibitory neurone or
>>>         synapses in neuroml? Could anyone tell me where to define this?
>>>
>>>         Thanks in advance,
>>>
>>>         Best,
>>>
>>>         Önder.
>>>         http://das.ege.edu.tr/~ogurcan/
>>>         <http://das.ege.edu.tr/%7Eogurcan/>
>>>
>>>
>>>         ------------------------------------------------------------------------------
>>>         DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps
>>>         OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
>>>         Free app hosting. Or install the open source package on any LAMP server.
>>>         Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native!
>>>         http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>>>
>>>
>>>         _______________________________________________
>>>         Neuroml-python mailing list
>>>         Neu...@li...  <mailto:Neu...@li...>
>>>         https://lists.sourceforge.net/lists/listinfo/neuroml-python
>>
>>
>>         -- 
>>
>>         -----------------------------------------------------
>>         Padraig Gleeson
>>         Room 321, Anatomy Building
>>         Department of Neuroscience, Physiology&  Pharmacology
>>         University College London
>>         Gower Street
>>         London WC1E 6BT
>>         United Kingdom
>>
>>         +44 207 679 3214  <tel:%2B44%20207%20679%203214>
>>         p.g...@uc...  <mailto:p.g...@uc...>
>>         -----------------------------------------------------
>>
>>
>>         ------------------------------------------------------------------------------
>>         DreamFactory - Open Source REST & JSON Services for HTML5 &
>>         Native Apps
>>         OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External
>>         API Access
>>         Free app hosting. Or install the open source package on any
>>         LAMP server.
>>         Sign up and see examples for AngularJS, jQuery, Sencha Touch
>>         and Native!
>>         http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>>         _______________________________________________
>>         Neuroml-technology mailing list
>>         Neu...@li...
>>         <mailto:Neu...@li...>
>>         https://lists.sourceforge.net/lists/listinfo/neuroml-technology
>>
>>
>
>
>     ------------------------------------------------------------------------------
>     Shape the Mobile Experience: Free Subscription
>     Software experts and developers: Be at the forefront of tech
>     innovation.
>     Intel(R) Software Adrenaline delivers strategic insight and
>     game-changing
>     conversations that shape the rapidly evolving mobile landscape.
>     Sign up now.
>     http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk
>     _______________________________________________
>     Neuroml-technology mailing list
>     Neu...@li...
>     <mailto:Neu...@li...>
>     https://lists.sourceforge.net/lists/listinfo/neuroml-technology
>
>


-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

Re: [Neuroml-python] [Neuroml-technology] Defining Excitatory/Inhibitory Neurons or Synapses

From: Önder G. <ond...@gm...> - 2013-11-21 16:51:17

Hello Padraig,

Thanks a lot for your explanation. The objective of my simulator is to find
the right PSP values for the synapses. I am then planning to export the
resulting network to NeuroML. However, according to you explanation I
suppose I need to make another parameter adjustment in order to find the
PSP values I have already found. Isn't there any other ways finding out
them? Or isn't there any calculation about how compute it?

Best,

Önder.
http://das.ege.edu.tr/~ogurcan/


On Wed, Nov 20, 2013 at 6:34 PM, Padraig Gleeson <uc...@li...>wrote:

>  Hi Önder,
>
> The size of the PSP will depend on the transient conductance (or current)
> of the synapse as well as the  the cell model you choose, particularly what
> its input resistance is. It might be best to choose a simple integrate and
> fire cell model, and a conductance based synapse, create a network with 2
> cells, and see what the voltage response is when one spikes and causes a
> PSP in the other, and adjust your parameters accordingly.
>
> Padraig
>
>
>
>
>
> On 18/11/13 22:11, Önder Gürcan wrote:
>
> Hello Padraig,
>
>  Thanks a lot for your answer and sorry for my ignorance about the topic.
>
>  In my simulator, I also modelled excitatory and inhibitory neurone like
> you said. However, as far as I know from [Iansek and Redman, 1973]
> (attached),  the post-synaptic potential (PSP) for a unitary synapse can
> range from 0.07 mV to 0.60 mV, but mostly it is between 0.10 and 0.20 mV
> (see Figure 4 from [Iansek and Redman, 1973]  attached also).
>
>  In this case, could you please tell me how to define a 0.1 mV PSP and
> -0.1 mV PSP in NeuroML?
>
>  Thanks in advance for helps,
>
>  Best regards,
>
>  Önder.
>
> [Iansek and Redman, 1973] Iansek, R. and Redman, S. (1973). The amplitude,
> time course and charge of unitary post-synaptic potentials evoked in spinal
> motoneurone dendrites. J. of Neurophysiol., 234:665 – 688.
>
>
>
>
> On Mon, Nov 18, 2013 at 1:01 PM, Padraig Gleeson <p.g...@uc...>wrote:
>
>>  Hi Önder,
>>
>> In NeuroML (1&2) there is a separation between the specification of the
>> cell properties and the synapses. A cell will be specified in terms of its
>> structure (segments etc.) and the locations of ion channel densities (in v1
>> or 2) or as an abstract cell (izhikevichCell, adExCell, iafCell, etc. from
>> v2.0), and the synapse model to use will be set when connections are made
>> between populations of the cells, e.g. see here:
>>
>>
>> https://github.com/NeuroML/NeuroML2/blob/master/examples/NML2_InstanceBasedNetwork.nml
>>
>> A cell is generally considered "excitatory" or "inhibitory" if the
>> reversal potentials of the synapses with which it sends input to other
>> cells are around 0mV and below -70mV respectively (though there are
>> exceptions to these values, and an inhibitory synapse can be excitatory for
>> a v hyperpolarised target cell).
>>
>> Hope this helps.
>>
>> Padraig
>>
>>
>>
>> On 16/11/13 21:46, Önder Gürcan wrote:
>>
>>  Hello all,
>>
>>  May be it is a very simple question for you but I was unable to find
>> how to define excitatory and inhibitory neurone or synapses in neuroml?
>> Could anyone tell me where to define this?
>>
>>  Thanks in advance,
>>
>>  Best,
>>
>>  Önder.
>> http://das.ege.edu.tr/~ogurcan/
>>
>>
>>  ------------------------------------------------------------------------------
>> DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps
>> OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
>> Free app hosting. Or install the open source package on any LAMP server.
>> Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native!http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>>
>>
>>
>> _______________________________________________
>> Neuroml-python mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/neuroml-python
>>
>>
>>
>> --
>>
>> -----------------------------------------------------
>> Padraig Gleeson
>> Room 321, Anatomy Building
>> Department of Neuroscience, Physiology&  Pharmacology
>> University College London
>> Gower Street
>> London WC1E 6BT
>> United Kingdom
>> +44 207 679 321...@uc...
>> -----------------------------------------------------
>>
>>
>>
>> ------------------------------------------------------------------------------
>> DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps
>> OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
>> Free app hosting. Or install the open source package on any LAMP server.
>> Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native!
>>
>> http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>> _______________________________________________
>> Neuroml-technology mailing list
>> Neu...@li...
>> https://lists.sourceforge.net/lists/listinfo/neuroml-technology
>>
>>
>
>
>
> ------------------------------------------------------------------------------
> Shape the Mobile Experience: Free Subscription
> Software experts and developers: Be at the forefront of tech innovation.
> Intel(R) Software Adrenaline delivers strategic insight and game-changing
> conversations that shape the rapidly evolving mobile landscape. Sign up
> now.
> http://pubads.g.doubleclick.net/gampad/clk?id=63431311&iu=/4140/ostg.clktrk
> _______________________________________________
> Neuroml-technology mailing list
> Neu...@li...
> https://lists.sourceforge.net/lists/listinfo/neuroml-technology
>
>

Re: [Neuroml-python] [Neuroml-technology] Defining Excitatory/Inhibitory Neurons or Synapses

From: Padraig G. <uc...@li...> - 2013-11-20 16:35:09

Hi Önder,

The size of the PSP will depend on the transient conductance (or 
current) of the synapse as well as the  the cell model you choose, 
particularly what its input resistance is. It might be best to choose a 
simple integrate and fire cell model, and a conductance based synapse, 
create a network with 2 cells, and see what the voltage response is when 
one spikes and causes a PSP in the other, and adjust your parameters 
accordingly.

Padraig




On 18/11/13 22:11, Önder Gürcan wrote:
> Hello Padraig,
>
> Thanks a lot for your answer and sorry for my ignorance about the topic.
>
> In my simulator, I also modelled excitatory and inhibitory neurone 
> like you said. However, as far as I know from [Iansek and Redman, 
> 1973] (attached),  the post-synaptic potential (PSP) for a unitary 
> synapse can range from 0.07 mV to 0.60 mV, but mostly it is between 
> 0.10 and 0.20 mV (see Figure 4 from [Iansek and Redman, 1973]  
> attached also).
>
> In this case, could you please tell me how to define a 0.1 mV PSP and 
> -0.1 mV PSP in NeuroML?
>
> Thanks in advance for helps,
>
> Best regards,
>
> Önder.
>
> [Iansek and Redman, 1973] Iansek, R. and Redman, S. (1973). The 
> amplitude, time course and charge of unitary post-synaptic potentials 
> evoked in spinal motoneurone dendrites. J. of Neurophysiol., 234:665 – 
> 688.
>
>
>
>
> On Mon, Nov 18, 2013 at 1:01 PM, Padraig Gleeson <p.g...@uc... 
> <mailto:p.g...@uc...>> wrote:
>
>     Hi Önder,
>
>     In NeuroML (1&2) there is a separation between the specification
>     of the cell properties and the synapses. A cell will be specified
>     in terms of its structure (segments etc.) and the locations of ion
>     channel densities (in v1 or 2) or as an abstract cell
>     (izhikevichCell, adExCell, iafCell, etc. from v2.0), and the
>     synapse model to use will be set when connections are made between
>     populations of the cells, e.g. see here:
>
>     https://github.com/NeuroML/NeuroML2/blob/master/examples/NML2_InstanceBasedNetwork.nml
>
>     A cell is generally considered "excitatory" or "inhibitory" if the
>     reversal potentials of the synapses with which it sends input to
>     other cells are around 0mV and below -70mV respectively (though
>     there are exceptions to these values, and an inhibitory synapse
>     can be excitatory for a v hyperpolarised target cell).
>
>     Hope this helps.
>
>     Padraig
>
>
>
>     On 16/11/13 21:46, Önder Gürcan wrote:
>>     Hello all,
>>
>>     May be it is a very simple question for you but I was unable to
>>     find how to define excitatory and inhibitory neurone or synapses
>>     in neuroml? Could anyone tell me where to define this?
>>
>>     Thanks in advance,
>>
>>     Best,
>>
>>     Önder.
>>     http://das.ege.edu.tr/~ogurcan/ <http://das.ege.edu.tr/%7Eogurcan/>
>>
>>
>>     ------------------------------------------------------------------------------
>>     DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps
>>     OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
>>     Free app hosting. Or install the open source package on any LAMP server.
>>     Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native!
>>     http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>>
>>
>>     _______________________________________________
>>     Neuroml-python mailing list
>>     Neu...@li...  <mailto:Neu...@li...>
>>     https://lists.sourceforge.net/lists/listinfo/neuroml-python
>
>
>     -- 
>
>     -----------------------------------------------------
>     Padraig Gleeson
>     Room 321, Anatomy Building
>     Department of Neuroscience, Physiology&  Pharmacology
>     University College London
>     Gower Street
>     London WC1E 6BT
>     United Kingdom
>
>     +44 207 679 3214  <tel:%2B44%20207%20679%203214>
>     p.g...@uc...  <mailto:p.g...@uc...>
>     -----------------------------------------------------
>
>
>     ------------------------------------------------------------------------------
>     DreamFactory - Open Source REST & JSON Services for HTML5 & Native
>     Apps
>     OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
>     Free app hosting. Or install the open source package on any LAMP
>     server.
>     Sign up and see examples for AngularJS, jQuery, Sencha Touch and
>     Native!
>     http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>     _______________________________________________
>     Neuroml-technology mailing list
>     Neu...@li...
>     <mailto:Neu...@li...>
>     https://lists.sourceforge.net/lists/listinfo/neuroml-technology
>
>

Re: [Neuroml-python] Defining Excitatory/Inhibitory Neurons or Synapses

From: Padraig G. <p.g...@uc...> - 2013-11-18 11:02:29

Hi Önder,

In NeuroML (1&2) there is a separation between the specification of the 
cell properties and the synapses. A cell will be specified in terms of 
its structure (segments etc.) and the locations of ion channel densities 
(in v1 or 2) or as an abstract cell (izhikevichCell, adExCell, iafCell, 
etc. from v2.0), and the synapse model to use will be set when 
connections are made between populations of the cells, e.g. see here:

https://github.com/NeuroML/NeuroML2/blob/master/examples/NML2_InstanceBasedNetwork.nml

A cell is generally considered "excitatory" or "inhibitory" if the 
reversal potentials of the synapses with which it sends input to other 
cells are around 0mV and below -70mV respectively (though there are 
exceptions to these values, and an inhibitory synapse can be excitatory 
for a v hyperpolarised target cell).

Hope this helps.

Padraig

On 16/11/13 21:46, Önder Gürcan wrote:
> Hello all,
>
> May be it is a very simple question for you but I was unable to find 
> how to define excitatory and inhibitory neurone or synapses in 
> neuroml? Could anyone tell me where to define this?
>
> Thanks in advance,
>
> Best,
>
> Önder.
> http://das.ege.edu.tr/~ogurcan/ <http://das.ege.edu.tr/%7Eogurcan/>
>
>
> ------------------------------------------------------------------------------
> DreamFactory - Open Source REST & JSON Services for HTML5 & Native Apps
> OAuth, Users, Roles, SQL, NoSQL, BLOB Storage and External API Access
> Free app hosting. Or install the open source package on any LAMP server.
> Sign up and see examples for AngularJS, jQuery, Sencha Touch and Native!
> http://pubads.g.doubleclick.net/gampad/clk?id=63469471&iu=/4140/ostg.clktrk
>
>
> _______________________________________________
> Neuroml-python mailing list
> Neu...@li...
> https://lists.sourceforge.net/lists/listinfo/neuroml-python

-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Defining Excitatory/Inhibitory Neurons or Synapses

From: Önder G. <ond...@gm...> - 2013-11-16 21:46:58

Hello all,

May be it is a very simple question for you but I was unable to find how to
define excitatory and inhibitory neurone or synapses in neuroml? Could
anyone tell me where to define this?

Thanks in advance,

Best,

Önder.
http://das.ege.edu.tr/~ogurcan/

Re: [Neuroml-python] [Neuroml-technology] Using NeuroML with Java

From: Önder G. <ond...@gm...> - 2013-11-13 08:47:32

Hello Padraig,

Thanks a lot for your answer. Matteo has already answered me the same way a
few days ago. However, he made "reply-to" my e-mail, rather than
"reply-to-all" I think.

Best,

Önder.
http://das.ege.edu.tr/~ogurcan/




On Tue, Nov 12, 2013 at 1:45 PM, Padraig Gleeson <p.g...@uc...>wrote:

>  Hi Önder,
>
> The best option is the set of packages related to jNeuroML. The "NeuroML
> Java API" is no longer supported.
>
> If you want to read/write NeuroML v1.8.1, use:
> https://github.com/NeuroML/org.neuroml1.model
> An example of reading & writing valid NeuroML v1.8.1 is here:
> https://github.com/NeuroML/org.neuroml1.model/tree/master/src/test/java/org/neuroml1/model/test
> Thanks for enquiring about this, the tests for org.neuroml1.model were not
> validating the generated NML1, and I've updating them to do this now.
>
> If you'd like to work with NeuroML 2 (recommended) use:
> https://github.com/NeuroML/org.neuroml.model
> Examples of reading/writing this are here:
> https://github.com/NeuroML/org.neuroml.model/blob/master/src/test/java/org/neuroml/model/test/NeuroML2Test.java
>
> Installing the source for jNeuroML (
> https://github.com/NeuroML/jNeuroML/blob/master/README.md) should get you
> most of the way, and you'll have compiled jars in the target directories of
> all the Java repositories, and they'll be ready for use locally with Maven.
>
> You'll have to decide whether to go with the stable NeuroML 2 beta1
> release on the master branches (which should be quite adequate for what you
> need), or try the most recent development version (use 'python
> getNeuroML.py development' in jNeuroML)
>
> Padraig
>
>
>
> On 11/11/13 14:59, Önder Gürcan wrote:
>
> Hello all,
>
>  I am developing a Java-based neural network simulator and I am planning
> to integrate NeuroML facility to my simulator. Basically, I want to be able
> to load and save neural network models in the cellular level (neurons and
> their synapses).
>
>  In this sense I checked the following website:
> http://www.neuroml.org/tool_support.php
>
>  However I was not if I need to use jNeuroML or NeuroML Java API. Which
> one shall I use? Which one is the latest and more stable?
>
>  Best regards,
>
>  Önder Gürcan
> http://das.ege.edu.tr/~ogurcan/
>
>
> ------------------------------------------------------------------------------
> November Webinars for C, C++, Fortran Developers
> Accelerate application performance with scalable programming models. Explore
> techniques for threading, error checking, porting, and tuning. Get the most
> from the latest Intel processors and coprocessors. See abstracts and registerhttp://pubads.g.doubleclick.net/gampad/clk?id=60136231&iu=/4140/ostg.clktrk
>
>
>
> _______________________________________________
> Neuroml-technology mailing lis...@li...https://lists.sourceforge.net/lists/listinfo/neuroml-technology
>
>
>
> --
>
> -----------------------------------------------------
> Padraig Gleeson
> Room 321, Anatomy Building
> Department of Neuroscience, Physiology&  Pharmacology
> University College London
> Gower Street
> London WC1E 6BT
> United Kingdom
> +44 207 679 321...@uc...
> -----------------------------------------------------
>
>

Re: [Neuroml-python] [Neuroml-technology] Using NeuroML with Java

From: Padraig G. <p.g...@uc...> - 2013-11-12 11:45:57

Hi Önder,

The best option is the set of packages related to jNeuroML. The "NeuroML 
Java API" is no longer supported.

If you want to read/write NeuroML v1.8.1, use:
https://github.com/NeuroML/org.neuroml1.model
An example of reading & writing valid NeuroML v1.8.1 is here: 
https://github.com/NeuroML/org.neuroml1.model/tree/master/src/test/java/org/neuroml1/model/test
Thanks for enquiring about this, the tests for org.neuroml1.model were 
not validating the generated NML1, and I've updating them to do this now.

If you'd like to work with NeuroML 2 (recommended) use:
https://github.com/NeuroML/org.neuroml.model
Examples of reading/writing this are here: 
https://github.com/NeuroML/org.neuroml.model/blob/master/src/test/java/org/neuroml/model/test/NeuroML2Test.java

Installing the source for jNeuroML 
(https://github.com/NeuroML/jNeuroML/blob/master/README.md) should get 
you most of the way, and you'll have compiled jars in the target 
directories of all the Java repositories, and they'll be ready for use 
locally with Maven.

You'll have to decide whether to go with the stable NeuroML 2 beta1 
release on the master branches (which should be quite adequate for what 
you need), or try the most recent development version (use 'python 
getNeuroML.py development' in jNeuroML)

Padraig


On 11/11/13 14:59, Önder Gürcan wrote:
> Hello all,
>
> I am developing a Java-based neural network simulator and I am 
> planning to integrate NeuroML facility to my simulator. Basically, I 
> want to be able to load and save neural network models in the cellular 
> level (neurons and their synapses).
>
> In this sense I checked the following website: 
> http://www.neuroml.org/tool_support.php
>
> However I was not if I need to use jNeuroML or NeuroML Java API. Which 
> one shall I use? Which one is the latest and more stable?
>
> Best regards,
>
> Önder Gürcan
> http://das.ege.edu.tr/~ogurcan/ <http://das.ege.edu.tr/%7Eogurcan/>
>
>
> ------------------------------------------------------------------------------
> November Webinars for C, C++, Fortran Developers
> Accelerate application performance with scalable programming models. Explore
> techniques for threading, error checking, porting, and tuning. Get the most
> from the latest Intel processors and coprocessors. See abstracts and register
> http://pubads.g.doubleclick.net/gampad/clk?id=60136231&iu=/4140/ostg.clktrk
>
>
> _______________________________________________
> Neuroml-technology mailing list
> Neu...@li...
> https://lists.sourceforge.net/lists/listinfo/neuroml-technology


-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Using NeuroML with Java

From: Önder G. <ond...@gm...> - 2013-11-11 14:59:45

Hello all,

I am developing a Java-based neural network simulator and I am planning to
integrate NeuroML facility to my simulator. Basically, I want to be able to
load and save neural network models in the cellular level (neurons and
their synapses).

In this sense I checked the following website:
http://www.neuroml.org/tool_support.php

However I was not if I need to use jNeuroML or NeuroML Java API. Which one
shall I use? Which one is the latest and more stable?

Best regards,

Önder Gürcan
http://das.ege.edu.tr/~ogurcan/

[Neuroml-python] NeuroML 2 beta1 release

From: Padraig G. <p.g...@uc...> - 2013-09-03 13:19:22

Hi,

We are proud to announce the latest stable release of NeuroML 2: beta1.

This represents a simultaneous merging of the "development" branches of 
a number of NeuroML 2/LEMS repositories on GitHub (listed below) to the 
"master" branches, and an updated NeuroML 2 Schema 
<https://github.com/NeuroML/NeuroML2/blob/master/Schemas/NeuroML2/NeuroML_v2beta1.xsd> 
to reflect the latest specification.

A summary of the changes in this release is here: 
https://github.com/NeuroML/NeuroML2/blob/master/HISTORY.md

The online descriptions of model components defined for Cells 
<http://www.neuroml.org/NeuroML2CoreTypes/Cells.html>, Synapses 
<http://www.neuroml.org/NeuroML2CoreTypes/Synapsess.html>, Channels 
<http://www.neuroml.org/NeuroML2CoreTypes/Channels.html> etc. have also 
been updated.

This is the first of a number of stable releases of NeuroML 2 towards a 
"final" v2.0. We hope that this process (ongoing work happening in the 
development branches, with ~1-2 monthly release cycle) will provide the 
stability required for developers to adopt the new version, while making 
it clear where the bleeding edge development is taking place.

The main repositories for NeuroML 2 and LEMS are listed below:

Schemas & documentation:
https://github.com/NeuroML/NeuroML2 (Main NeuroML 2 repository, Schemas 
& CompType definitions)
https://github.com/LEMS/LEMS (Main LEMS repository with Schema)
Java libraries:
https://github.com/NeuroML/org.neuroml.model.injectingplugin (Required 
by the 2 libraries below)
https://github.com/NeuroML/org.neuroml.model                          
(Java API for NeuroML v2)
https://github.com/NeuroML/org.neuroml1.model (Java API for NeuroML v1.8.1)
https://github.com/NeuroML/org.neuroml.export                          
(Various export formats for NML2/LEMS)
https://github.com/NeuroML/org.neuroml.import (Importing other formats 
to LEMS)
https://github.com/NeuroML/jNeuroML (Bundles all above into one Jar file)
https://github.com/LEMS/jLEMS                             
                          (Reference implementation of LEMS in Java)
Python:
https://github.com/NeuralEnsemble/libNeuroML (Python API for NeuroML 2)
https://github.com/LEMS/pylems (Reference implementation in Python for LEMS)

Regards,
The NeuroML Editors & contributors

  

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Updated Open Source Brain website & INCF GSoC project

From: Padraig G. <p.g...@uc...> - 2013-04-23 15:14:08

Hi,

2 quick things, we've recently updated the Open Source Brain website 
(http://opensourcebrain.org), which in addition to a new look, will 
hopefully make it easier to browse between projects of similar brain 
regions and find cell and network models of interest. We've also added a 
grading system to indicate how well models are supported in different 
simulators and how well they're curated against published version. 
Comments very welcome.

There is also a NeuroML/OSB related project suggestion as part of the 
2013 Google Summer of Code with the INCF as mentoring organisation. This 
is a great opportunity for interested students to learn the latest 
NeuroML/PyNN technologies, get to know a detailed cortical network model 
really well, and _earn some extra cash_ for contributing to an 
interesting project over the summer. Deadline May 3.

*Open source, cross simulator, large scale cortical models*

*Description: *An increasing number of studies are using large scale 
network models incorporating realistic connectivity to understand 
information processing in cortical structures. High performance 
computational resources are becoming more widely available to 
computational neuroscientists for this type of modelling and general 
purpose, well tested simulation environments such as NEURON and NEST are 
widely used. In addition, hardware solutions (based on custom 
neuromorphic hardware or off the shelf FPGA/GPU hardware) are in 
development, promising ever larger, faster simulations. However, there 
is a lack of well tested, community maintained network model examples 
which can work across all of these simulation solutions, which both 
incorporate realistic cell and network properties and provide networks 
of sufficient complexity to test the performance and scalability of 
these platforms. This work will involve converting a number of published 
large scale network models into open, simulator independent formats such 
as PyNN, NeuroML and NineML, testing them across multiple simulator 
implementations and making them available to the community through the 
Open Source Brain repository.

For more details see: http://www.incf.org/gsoc. Contact Andrew Davison 
(and...@un...) or me if you're interested.

Padraig

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] jNeuroML: a Swiss army knife for (almost) all your NeuroML 2 & LEMS needs...

From: Padraig G. <p.g...@uc...> - 2013-04-10 09:47:35

Hi,

There are a number of repositories in active development under GitHub 
for handling NeuroML 2 and LEMS using Java. To make it easier to access 
all of this functionality, we've created a single package, jNeuroML, 
which allows access to most of this functionality through a simple 
command line interface and requires minimal installation. To get a 
precompiled binary for this, type:

     svn checkout svn://svn.code.sf.net/p/neuroml/code/jNeuroMLJar
     git clone git://github.com/NeuroML/NeuroML2.git
     cd jNeuroMLJar

Typing ./jnml (or jnml.bat) will list the options available. Some of the 
current options include:

     ./jnml -validate MyNeuroML.nml              (validate NeuroML 2 
document against the current schema)
     ./jnml -validatev1 MyNeuroML1.xml         (validate NeuroML v1 
document against the v1.8.1 schema)
     ./jnml MyLEMS.xml                                    (parse & 
simulate a LEMS model using jLEMS)
     ./jnml MyLEMS.xml -graph                        (generate png of 
structure of LEMS model using GraphViz)

Export and import features for NEURON, SBML, Brian etc. are in 
development (https://github.com/NeuroML/org.neuroml.export and 
https://github.com/NeuroML/org.neuroml.import) and this functionality 
will be included in the jnml utility as it is developed.

Points to note:

- Adding the environment variable JNML_HOME, pointing to the jNeuroMLJar 
folder, as well as adding this path to the PATH variable will let you 
use the jnml utility from any folder.

- Running svn update in the jNeuroMLJar folder will get the latest 
version of the binary. There are much better ways to distribute binaries 
than putting them in an SVN repo I know, but this is a rapidly changing 
application and this seems to best way to distribute the latest release 
at the moment with the minimum of hassle for users.


     --- Getting the source for jNeuroML ---

If you prefer to clone all of the individual repositories and build the 
jNeuroML application yourself, use the getNeuroML.py utility in the 
jNeuroML repo:

     git clone git://github.com/NeuroML/jNeuroML.git neuroml_dev/jNeuroML
     cd neuroml_dev/jNeuroML
     python getNeuroML.py

This will clone ~11 repos for NML2 & LEMS (including Python based 
libraries) into neuroml_dev/ and compile the Java based ones using 
Maven. The full process may take 5-10 mins on first installation, but 
subsequently running

     git pull
     python getNeuroML.py

in the jNeuroML folder will get the latest code for each repo & compile 
using Maven if necessary.

Use of Maven is a great way to manage versions of applications being 
developed in distributed repositories, and will make it easy to use 
selected parts of this for different Java applications. For example, 
these packages will be used in various ways to provide NeuroML/LEMS 
support in neuroConstruct and for handling NeuroML on the Open Source 
Brian website.

Enjoy,
Padraig


-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Mailing list for technical discussions about LEMS

From: Padraig G. <p.g...@uc...> - 2013-03-28 18:08:29

Hi,

I know how people are always complaining about how few mailing lists 
there are out there to sign up to, so here's another:

https://groups.google.com/forum/?fromgroups&hl=en-GB#!forum/lems-discuss

A mailing list specifically for discussing all things LEMS related. 
Major announcements about LEMS will still be made on the main NeuroML 
mailing lists.

Enjoy.
Padraig


-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Open Source Brain & NeuroML development workshop - Sardinia May 2013

From: Padraig G. <p.g...@uc...> - 2013-02-20 17:10:33

Dear all,

The kick off meeting for the Open Source Brain 
(http://www.opensourcebrain.org) initiative will take place in Alghero, 
Sardinia on 13-15th May this year. This initiative aims to encourage 
collaborative modelling in computational neuroscience, and create a 
repository of well tested, open source cell & network models from 
multiple brain regions & species, which can be used across simulators 
and other applications through the use of standardised languages like 
NeuroML & PyNN.

This first meeting will focus on modelling of the cerebellum, but 
modellers & experimentalists working on other brain regions are very 
welcome to join. Also, application/simulator developers who wish to test 
their software with real world models are encouraged to participate. 
There will be tutorials on the main technologies behind OSB and there 
will be plenty of time for hands on testing of models and short talks by 
modellers/software developers who have something to share with the 
community.

The models on OSB are shared via public, open source repositories (e.g. 
at GitHub <https://github.com/OpenSourceBrain>), and we plan that the 
majority of work from the meeting on annotation/model sharing/wiki 
editing will be immediately available for the wider community.

Developments with OSB are closely related to work on NeuroML & LEMS, and 
this meeting will serve as the NeuroML Development Workshop for this 
year. Further extensions to NeuroML/LEMS will be guided by the types of 
models most actively used in OSB, and models converted to NeuroML on OSB 
can benefit from the libraries and automated 
annotation/validation/visualisation tools being developed to support the 
platform.

The meeting is free to attend, but registration is required. Please see 
http://www.opensourcebrain.org/projects/osb/wiki/Meetings for more 
details. The OSB initiative is primarily supported by the Wellcome Trust.

Regards,
Padraig

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology & Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Update on LEMS & NeuroML 2 developments

From: Padraig G. <p.g...@uc...> - 2013-02-15 11:46:08

Hi,

We've been busy in a few different directions recently with developments 
for NeuroML 2 and LEMS so probably time for a quick update.

     LEMS

- The main specifications for LEMS (Low Entropy Model Specification 
language) has been moved here: https://github.com/LEMS/LEMS. This 
currently consists of the XML Schema for LEMS, but the documentation, 
core examples etc. will be moved here also from the Java & Python 
repositories below.

- jLEMS, or the Java reference implementation of LEMS by Robert Cannon 
has been moved here: https://github.com/LEMS/jLEMS. This includes a 
number of examples which can be read in & executed by jLEMS and has the 
most up to date documentation for LEMS elements 
(http://lems.github.com/jLEMS/elements.html).

- PyLEMS (you guessed it, a Python implementation of LEMS) developed by 
Gautham Ganapathy is here: https://github.com/LEMS/pylems, and uses the 
same updated version of LEMS as jLEMS and can parse/validate/simulate 
most of the same examples.

     NeuroML 2

- The main repository for NeuroML 2 specifications is here: 
https://github.com/NeuroML/NeuroML2. This includes the Schemas, 
including the NeuroML 2 beta schema 
<https://github.com/NeuroML/NeuroML2/blob/master/Schemas/NeuroML2/NeuroML_v2beta.xsd>, 
and the LEMS ComponentTypes 
<https://github.com/NeuroML/NeuroML2/tree/master/NeuroML2CoreTypes> 
which define the structure and behaviour of elements of NeuroML 2. These 
definitions can be viewed online for a (hopefully) clearer view of their 
structure, e.g. izhikevichCell 
<http://www.neuroml.org/NeuroML2CoreTypes/Cells.html#izhikevichCell>, 
nmdaSynapse 
<http://www.neuroml.org/NeuroML2CoreTypes/Synapses.html#nmdaSynapse>, 
sineGenerator 
<http://www.neuroml.org/NeuroML2CoreTypes/Inputs.html#sineGenerator>, etc.

- A number of Java packages modules for handling NeuroML have been 
created here <https://github.com/NeuroML/>. One of these, 
org.neuroml.model <https://github.com/NeuroML/org.neuroml.model>, is a 
Java API for reading, writing & validating NeuroML 2. Others include 
packages for exporting NeuroML & LEMS models to various formats. These 
packages use Maven <http://maven.apache.org/index.html> to make it 
easier to incorporate them into other Java applications. We plan to 
gather all these NeuroML/LEMS packages together into a single 
downloadable package <https://github.com/NeuroML/jNeuroML> with a simple 
command line interface for reading/writing/executing/converting models.

- The Python API for NeuroML 2, libNeuroML 
<https://github.com/NeuralEnsemble/libNeuroML>, is undergoing major 
refactoring, but the existing version should still work fine for 
reading/writing NML2. Please get in contact with Mike Vella or myself if 
you're keen to use this. This initiative is closely linked to a Python 
API for multicompartmental modelling, Pyramidal 
<https://github.com/vellamike/pyramidal>, which will further increase 
interoperability between NeuroML & PyNN.

Hope that provides a good overview of what's available at the moment. As 
suggested most of these initiatives are under active development, and 
offers of help or just feedback are more than welcome! These 
developments are also closely related to our work for the Open Source 
Brain repository (http://opensourcebrain.org). More exciting 
announcements about that coming soon!

Regards,
The NeuroML, LEMS & OSB teams

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Update on NeuroML 2, LEMS, libNeuroML, PyNN, Open Source Brain, OpenWorm, etc.

From: Padraig G. <p.g...@uc...> - 2012-09-10 14:05:26

Hi all,

There has been quite a bit of activity on NeuroML and related 
initiatives recently, so I thought it might be time for a general update 
on these and how they're all related.

     -- NeuroML 2 --

The latest version of the ComponentType definitions for NeuroML 2 can be 
browsed here <http://www.neuroml.org/NeuroML2CoreTypes/Cells.html> (e.g. 
izhikevichCell 
<http://www.neuroml.org/NeuroML2CoreTypes/Cells.html#izhikevichCell>, 
adExIaFCell 
<http://www.neuroml.org/NeuroML2CoreTypes/Cells.html#adExIaFCell>, 
nmdaSynapse 
<http://www.neuroml.org/NeuroML2CoreTypes/Synapses.html#nmdaSynapse>). 
These are generated from the LEMS definitions 
<http://sourceforge.net/apps/trac/neuroml/browser/NeuroML2/NeuroML2CoreTypes> 
of the dynamical elements in NeuroML 2. The hierarchical structure of 
these types (see first figure here 
<http://www.neuroml.org/neuroml2.php>) has been updated recently to 
prefix abstract component types with "base", e.g. all synapse types 
extend baseSynapse 
<http://www.neuroml.org/NeuroML2CoreTypes/Synapses.html#baseSynapse>, etc.

The NeuroML 2 Schema 
<http://sourceforge.net/apps/trac/neuroml/browser/NeuroML2/Schemas/NeuroML2/NeuroML_v2alpha.xsd> 
and the examples (NML2_*.nml files here 
<http://sourceforge.net/apps/trac/neuroml/browser/NeuroML2/examples>) 
are being kept up to date with these developing types.

Some of the LEMS definitions have been annotated with the corresponding 
entries in the emerging Computational Neuroscience Ontology 
<http://bioportal.bioontology.org/ontologies/46856/> (see example here 
<http://www.neuroml.org/NeuroML2CoreTypes/Synapses.html#stpMechanism>).


     -- LEMS --

The core set of elements <http://localhost/lems/elements.html> in LEMS 
for describing ComponentTypes and their associated Behaviors are quite 
stable. The Java Reference implementation 
<http://neuroml.org/lems/interpreter.html> which can parse, validate and 
execute models described in LEMS code had been under development on the 
Sourceforge repository 
<http://sourceforge.net/apps/trac/neuroml/browser/LEMS> has been moved 
to a GitHub repository and renamed jLEMS 
<https://github.com/robertcannon/jLEMS>. Some minor refactoring is 
taking place on the code (along with new features like automated 
flattening of hierarchical ComponentTypes and a new numerical 
integration mechanism) and we will let you know when this is complete.

The Java application which bundles LEMS and the NML2 core ComponentType 
definitions, and allows import and export to different formats 
(including NEURON, SBML and NineML) is still under active development 
(see here <http://www.neuroml.org/neuroml2.php#libNeuroML>). The current 
name for this, libNeuroML, is probably not appropriate given the usage 
of this type of name in other packages (libSBML etc.) and the developing 
Python API called libNeuroML. This package will be renamed in the near 
future (suggestions welcome; jNML2 perhaps?)

There is also a Python implementation of LEMS (which like the jLEMS will 
allow reading, writing, validating & execution of LEMS models) under 
development here <https://github.com/lisphacker/pylems>.


-- libNeuroML in Python --

Mike Vella has been doing great work developing libNeuroML as a pure 
Python API for handling NeuroML files as part of his INCF/Google Summer 
of Code project. More info here 
<http://libneuroml.readthedocs.org/en/latest/index.html>, or see the 
GitHub repository <https://github.com/NeuralEnsemble/libNeuroML>.

Mike is also working on Pyramidal 
<http://pyramidal.readthedocs.org/en/latest/index.html>, which is a 
Python API for multicompartmental modelling of neurons (GitHub repo 
<https://github.com/vellamike/pyramidal>). He's had some good initial 
success with creating scripts in Python which can generate 
multicompartmental simulations to run on NEURON or MOOSE.

Both of these initiatives will help increase the interoperability 
between PyNN and NeuroML <http://www.neuroml.org/pynn.php>, allowing 
modellers to move seamlessly between declarative model definitions in 
NeuroML and procedural model creation, modification and analysis with 
Python and PyNN.


-- Open Source Brain ---

This initiative <http://www.opensourcebrain.org/> which has been 
presented at the CNS <http://www.cnsorg.org/cns-2012-atlantadecatur> and 
COMBINE <http://co.mbine.org/events> meetings this summer aims to 
facilitate collaborative development of models in computational 
neuroscience. The initial version of this repository hosts many cortical 
<http://www.opensourcebrain.org/embedded/osb/index.html>, cerebellar and 
other detailed cell and network models. The intention is to gradually 
translate these models to simulator independent formats (mainly using 
NeuroML, but also PyNN, SED-ML, etc. where appropriate) so they can be 
simulated/analysed by multiple applications. Rather than be an archive 
for static, published models like ModelDB 
<http://senselab.med.yale.edu/modeldb/>, the models will develop over 
time, cells, synapses and channels can be reused, and automated tests 
will ensure best practices in model specification and reproducibility of 
behaviour across simulators.

The majority of models are hosted in individual GitHub repositories 
<https://github.com/OpenSourceBrain>, and the main website will provide 
a place to discover models for different brain regions 
<http://www.opensourcebrain.org/themes/>, view more readable 
descriptions 
<http://www.opensourcebrain.org/embedded/cerebellarnucleusneuron/index.html> 
of projects and their components, provide an infrastructure for project 
specific wikis 
<http://www.opensourcebrain.org/projects/grancellsolinasetal10/wiki> and 
issue tracking <http://www.opensourcebrain.org/issues/8>, and provide 
guides <http://www.opensourcebrain.org/guides> to the latest tools for 
working with the models in the repository.

You are invited to register 
<http://www.opensourcebrain.org/account/register> on the web site, 
clone/fork projects, comment on existing models and provide feedback on 
this developing initiative.

Matteo Cantarelli who has been active in the OpenWorm project has 
recently joined the Silver lab and will be helping out on the 
development of this site.


-- Open Worm --

This project <http://openworm.org/> has the ambitious aim of simulating 
a simple life form, the roundworm caenorhabditis elegans 
<http://en.wikipedia.org/wiki/Caenorhabditis_elegans>, in a 
computational model, incorporating electrical and chemical cellular 
activity and interaction with the physical environment. An initial step 
in this project is gathering information on the connectome of c. elegans 
which has been translated into a very nice online 3D browser 
<http://browser.openworm.org/>, and has also been converted to NeuroML. 
This can be downloaded as part of a neuroConstruct project 
<http://code.google.com/p/openworm/wiki/neuroConstructCElegans>.

Matteo has also been working on the OpenWorm Simulation Engine 
<http://code.google.com/p/openworm/wiki/SimulationFrameworkConcepts>, a 
new simulation framework which will have the task of integrating the 
electrophysiological and physical aspects of the final model. The 
neuronal solver in this framework will be closely integrated with code 
being developed for NeuroML 2 & LEMS, creating a new simulator which 
will have the ability to run generic NeuroML models on desktops, GPUs 
and in the cloud.


Hope this update is enough to keep you going for a while. Please get in 
contact if you have any feedback, or join the various mailing lists this 
update has been sent to to stay up to date with the relevant initiatives.

Regards,
Padraig

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

[Neuroml-python] Minutes of recent Python API meeting

From: Padraig G. <p.g...@uc...> - 2012-07-06 14:55:40

Hi all,

Minutes of the recent meeting in London for the Python API for 
multicompartmental simulations (libNeuroML/Mike Vella's Google Summer of 
Code project) is available here: 
http://libneuroml.readthedocs.org/en/latest/meeting_june_2012.html.

Thanks again to the INCF for supporting this meeting.

Regards,
Padraig

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

Re: [Neuroml-python] Specification of a model

From: Padraig G. <p.g...@uc...> - 2012-06-27 18:25:32

Hi,

More answers below...


On 27/06/12 09:20, Mike Vella wrote:
> Hi Stephan,
>
> Answered inline
>
> On 27 June 2012 08:45, Stephan Gerhard <uni...@gm...> wrote:
>> Hi,
>>
>> I had a brief look at the envisioned specification of a model
>> https://github.com/NeuralEnsemble/libNeuroML/blob/master/hhExample/hh_NEUROML2.py
>>
>> A few comments/improvements:
>>
>> * I like the separation between morphology definition and channel
>> density specification.
>> If you specify the segmentGroup, and later add more segments, it would have to
>> retroactively set the channel densities for this added segment. This
>> could become a
>> bit unpredictable.
This is not an issue for the serialisation in XML, 
segments/segmentGroups are always set before biophysics. It could be a 
problem though in building the cells through the API, but if you think 
about it, if a segment is specified as part of dendrite_group and 
channels have previously been put on this group, then the intention 
probably is to have the channels on that new segment too.

This might be an argument for (at least initially...) building the 
object model description up before instantiating it on the simulator (as 
is the case in morphforge). All the parts of the model would have to be 
specified and then an instance of the corresponding objects is created 
in the simulator.
>>
>> * It seems as if you add segments later to the morphology, they do not
>> have to be topologically
>> connected to the already existing segments. (like on line 67+). I'd
>> rather enforce connected
>> topology here.
> Connected topology is enforced. If you add a segment it does have to
> be topologically connected, the add_segment() method in this example
> (we decided to use attach instead) topologically connects the segment
> to some default node on a default segment (probably root). As a side
> note, there was some discussion about implementing "virtual
> connections" vs "physical connections", this is (mostly) for treating
> edge cases in modelling where people wish to treat spatially separated
> segments as though they were electrically connected.
Yes, we've discussed this about whether to allow "floating dendrites" as 
is the case in many reconstructed cells. This seems to be a requirement 
is we want to use this API for real multicomp simulations, but if an 
application like CATMAID only deals with fully connected cells it can 
ignore this option.
>
>> * The network specification (line 93) is clearly targeted at
>> population level modeling. This is
>> not very general, as people would want to model networks with sets of
>> singly identified neurons.
>> Of course, you could define a population with only one cell, but this
>> would not be the goal of
>> a good API design.
It's a bit of overhead but it's not too bad if you think of "Network" as 
"All my instantiated model elements" and the creation of "Population" as 
"Add N cells to my model". Cells defined before this (as well as 
channels, synapses, etc.) are really just templates and not instantiated 
until put into the network in some way.

As I say, this may seem like overhead, but it's a way to treat networks 
of multiple cells & single instance cells in the same way without 
introducing 2 ways to instantiate them.

Of course you can have an "unwrapped" cell in an XML file or create it 
in a Python script and do whatever you want with it, but the Network 
element is a holder for instantiated cells & connections between them (& 
inputs).

>>
>> * Also, the specification of the population of cells is without their
>> spatial locations, which would
>> look like necessary information for a simulation.
Yes, good point. This is very much supported in NMLv1.x, and the 
equivalent entities from there need to be brought over to v2.0. What's 
delaying this is making sure LEMS will parse & understand the use of 
<instance> elements inside <population>, but until now size="x" has been 
adequate. Consider it bumped up the TODO list...
>>
>> * In an ideal world, I would want to specify inputs at a defined
>> location, e.g. at a junction of two
>> segments, or even at a point along a segment.
Yes, specifying a connection to a point a fraction along a segment will 
be supported.
>>
>> * Synapse specification: What I feel generally missing is the
>> specification of the location of
>> a synapse. Specification of synapse location on a neuronal morphology
>> has a big impact on
>> membrane voltage dynamics. Thus it would seem sensible to me to be
>> able to specify this,
>> and have a generic mechanism to distribute the synapse on the tree if
>> location is unknown or
>> not specified. Adding synaptic connectivity on a population level
>> might only make sense
>> in a very particular modeling scenario, but it's not very generic.
This was possible in v1.x (and should be possible in v2) with adding 
explicit synaptic inputs to segment locations (e.g. line 146 here: 
http://sourceforge.net/apps/trac/neuroml/browser/NeuroML2/examples/NMLv1.x/CompleteNetwork.xml). 
This was more influenced by the modelling world, where the synapse was a 
spiking process, there wasn't a need to just say "there's a synapse here"...

The synaptic input was always specified outside the morphology, but 
there may be an argument to have some flagged locations for synapses 
specified inside the <cell> element/object and maybe make connections 
later based on these synapse ids.

>>
>> Synapse connectivity could be from a proximal/distal end of a segment
>> in one cell to the
>> proximal/distal end of a segment in another cell. Keep in mind that
>> creating objects for
>> each synapse will be very inefficient for a large number of synapses.
Flagging that a synapse is along a segment shouldn't always require 
creation of a new object.

>>
>> I'm looking forward to test the API with a neural circuit
>> reconstructed in CATMAID! :)
Would it be posible to create a short Python script that parses your 
example at 
https://github.com/NeuralEnsemble/libNeuroML/blob/master/hdf5Examples/neurohdf_microcircuit.hdf 
and serialises this into a simple indented text file (grouped by 
neurons, with lists of segments & synapses)? This could be your ideal 
format for serialisation (you don't have to make XML out of it though). 
I think this could influence an updated form for the NeuroML XML element 
model, which would be needed for introducing this feature consistently 
into the API.

Regards,
Padraig

>>
>> Stephan
>>
>> ------------------------------------------------------------------------------
>> Live Security Virtual Conference
>> Exclusive live event will cover all the ways today's security and
>> threat landscape has changed and how IT managers can respond. Discussions
>> will include endpoint security, mobile security and the latest in malware
>> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>> _______________________________________________
>> Neuroml-python mailing list
>> Neu...@li...
>> https://lists.sourceforge.net/lists/listinfo/neuroml-python
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Neuroml-python mailing list
> Neu...@li...
> https://lists.sourceforge.net/lists/listinfo/neuroml-python
>


-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

Re: [Neuroml-python] [Neuroml-technology] support for spiketime files as synaptic input

From: Padraig G. <p.g...@uc...> - 2012-06-27 10:37:31

Hi,

There isn't support for this in NeuroML v1.x (and it's not planned at 
this stage), but there is an example of how to do this in v2.0:
http://sourceforge.net/apps/trac/neuroml/browser/NeuroML2/examples/LEMS_NML2_Ex12_Net2.xml

Anything which produces spikes (i.e. has an outward EventPort e.g. a 
spikeArray 
http://www.neuroml.org/NeuroML2CoreTypes/Inputs.html#spikeArray, or just 
a cell http://www.neuroml.org/NeuroML2CoreTypes/Cells.html#iafTauCell in 
a population) can be the endpoint of a synapticConnection (or 
synapticConnectionWD with weight & delay). There isn't support for 
loading the spikes in from a simpler text file, but there should be a 
simple transformation to convert something like that into a XML element 
like <spikeArray id="spikes2"><spike ...>

Alternatively these arrays could be created in Python with libNeuroML by 
adding Spikes to a SpikeArray object. Haven't got an example of this 
yet, but you could maybe see how it might be done from this example for 
building a network:
http://sourceforge.net/apps/trac/neuroml/browser/NeuroML2/python/examples/run_network.py

Note that this example is based on the Python API generated from the XML 
Schema & is not too pretty. This will eventually be cleaned up & 
integrated with the libNeuroML on GitHub & made very PyNN friendly.

Regards,
Padraig

On 26/06/12 14:40, Aditya Gilra wrote:
> Is there support for providing spiketime files as synaptic input?
> Almost all cells in my olfactory bulb model in PyMOOSE need 
> spiketrains as input.
>
> My present hack is to specify pre_cell_id="file" and a file number as 
> pre_segment_id in <connection> tags.
> An alternative could be have a tag under <inputs>.
>
> (Forgot to mention this in the libNeuroML meet, but sufficiently 
> important to merit independent attention.)
>
> Best,
> Aditya.
>
>
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats.http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
>
>
> _______________________________________________
> Neuroml-technology mailing list
> Neu...@li...
> https://lists.sourceforge.net/lists/listinfo/neuroml-technology

-- 

-----------------------------------------------------
Padraig Gleeson
Room 321, Anatomy Building
Department of Neuroscience, Physiology&  Pharmacology
University College London
Gower Street
London WC1E 6BT
United Kingdom

+44 207 679 3214
p.g...@uc...
-----------------------------------------------------

Re: [Neuroml-python] Specification of a model

From: Mike V. <vel...@gm...> - 2012-06-27 08:20:35

Hi Stephan,

Answered inline

On 27 June 2012 08:45, Stephan Gerhard <uni...@gm...> wrote:
> Hi,
>
> I had a brief look at the envisioned specification of a model
> https://github.com/NeuralEnsemble/libNeuroML/blob/master/hhExample/hh_NEUROML2.py
>
> A few comments/improvements:
>
> * I like the separation between morphology definition and channel
> density specification.
> If you specify the segmentGroup, and later add more segments, it would have to
> retroactively set the channel densities for this added segment. This
> could become a
> bit unpredictable.
>
> * It seems as if you add segments later to the morphology, they do not
> have to be topologically
> connected to the already existing segments. (like on line 67+). I'd
> rather enforce connected
> topology here.

Connected topology is enforced. If you add a segment it does have to
be topologically connected, the add_segment() method in this example
(we decided to use attach instead) topologically connects the segment
to some default node on a default segment (probably root). As a side
note, there was some discussion about implementing "virtual
connections" vs "physical connections", this is (mostly) for treating
edge cases in modelling where people wish to treat spatially separated
segments as though they were electrically connected.

>
> * The network specification (line 93) is clearly targeted at
> population level modeling. This is
> not very general, as people would want to model networks with sets of
> singly identified neurons.
> Of course, you could define a population with only one cell, but this
> would not be the goal of
> a good API design.
>
> * Also, the specification of the population of cells is without their
> spatial locations, which would
> look like necessary information for a simulation.
>
> * In an ideal world, I would want to specify inputs at a defined
> location, e.g. at a junction of two
> segments, or even at a point along a segment.
>
> * Synapse specification: What I feel generally missing is the
> specification of the location of
> a synapse. Specification of synapse location on a neuronal morphology
> has a big impact on
> membrane voltage dynamics. Thus it would seem sensible to me to be
> able to specify this,
> and have a generic mechanism to distribute the synapse on the tree if
> location is unknown or
> not specified. Adding synaptic connectivity on a population level
> might only make sense
> in a very particular modeling scenario, but it's not very generic.
>
> Synapse connectivity could be from a proximal/distal end of a segment
> in one cell to the
> proximal/distal end of a segment in another cell. Keep in mind that
> creating objects for
> each synapse will be very inefficient for a large number of synapses.
>
> I'm looking forward to test the API with a neural circuit
> reconstructed in CATMAID! :)
>
> Stephan
>
> ------------------------------------------------------------------------------
> Live Security Virtual Conference
> Exclusive live event will cover all the ways today's security and
> threat landscape has changed and how IT managers can respond. Discussions
> will include endpoint security, mobile security and the latest in malware
> threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
> _______________________________________________
> Neuroml-python mailing list
> Neu...@li...
> https://lists.sourceforge.net/lists/listinfo/neuroml-python

[Neuroml-python] Specification of a model

From: Stephan G. <uni...@gm...> - 2012-06-27 07:46:08

Hi,

I had a brief look at the envisioned specification of a model
https://github.com/NeuralEnsemble/libNeuroML/blob/master/hhExample/hh_NEUROML2.py

A few comments/improvements:

* I like the separation between morphology definition and channel
density specification.
If you specify the segmentGroup, and later add more segments, it would have to
retroactively set the channel densities for this added segment. This
could become a
bit unpredictable.

* It seems as if you add segments later to the morphology, they do not
have to be topologically
connected to the already existing segments. (like on line 67+). I'd
rather enforce connected
topology here.

* The network specification (line 93) is clearly targeted at
population level modeling. This is
not very general, as people would want to model networks with sets of
singly identified neurons.
Of course, you could define a population with only one cell, but this
would not be the goal of
a good API design.

* Also, the specification of the population of cells is without their
spatial locations, which would
look like necessary information for a simulation.

* In an ideal world, I would want to specify inputs at a defined
location, e.g. at a junction of two
segments, or even at a point along a segment.

* Synapse specification: What I feel generally missing is the
specification of the location of
a synapse. Specification of synapse location on a neuronal morphology
has a big impact on
membrane voltage dynamics. Thus it would seem sensible to me to be
able to specify this,
and have a generic mechanism to distribute the synapse on the tree if
location is unknown or
not specified. Adding synaptic connectivity on a population level
might only make sense
in a very particular modeling scenario, but it's not very generic.

Synapse connectivity could be from a proximal/distal end of a segment
in one cell to the
proximal/distal end of a segment in another cell. Keep in mind that
creating objects for
each synapse will be very inefficient for a large number of synapses.

I'm looking forward to test the API with a neural circuit
reconstructed in CATMAID! :)

Stephan

Re: [Neuroml-python] Agenda for libNeuroML meeting, 25th, 26th May

From: Andrew D. <and...@gm...> - 2012-06-22 14:08:58

On 22 juin 2012, at 15:54, Padraig Gleeson wrote:
> Hi,
> 
> I've put 9am to 6pm to the agenda, but some people are arriving later 
> (Andrew is just lunchtime Mon to lunchtime Tues).

No, I'll be there for 9am on Monday provided the Eurostar is on time.

> We have the room till 
> 6 but will see how long we last... I'll be there from about 8:30 and 
> I'll tell the security guy on the door to let you sign in & just walk up 
> to the 3rd floor. Any problems, call me on +44 797 9970283.

1 message has been excluded from this view by a project administrator.

Flat | Threaded

1 2 3 > >> (Page 1 of 3)

2012	Jan	Feb	Mar	Apr (1)	May (21)	Jun (11)	Jul (1)	Aug	Sep (1)	Oct	Nov	Dec
2013	Jan	Feb (2)	Mar (1)	Apr (2)	May	Jun	Jul	Aug	Sep (1)	Oct	Nov (9)	Dec
2016	Jan	Feb (1)	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct	Nov	Dec
2017	Jan	Feb	Mar	Apr	May	Jun	Jul	Aug	Sep	Oct (3)	Nov	Dec