13 Oct
2022
13 Oct
'22
11:37 a.m.
Hello NEST community,
I am at the outset of a large project and I would like to start off with
the right simulator. Perhaps someone here can advise on the viability of
NEST for my particular needs. I will be modeling neuromorphic hardware with
some unique design constraints, such as:
- A complex dendritic arbor
- Dendrite potentials may be coupled with their neighbors
- Several dendrites may feed into other dendrites (a several-layer-deep
dendritic tree)
- The arbor eventually feeds into the soma which has typical LIF
behavior, with a membrane potential determined by the sum of all directly
connected dendrites
- The solution for the leaky-integrator term of dendrites must be drawn
from a precalculated lookup table
The question here is how best to facilitate these dendritic nuances within
NEST? Any counsel on whether it makes sense to pursue modeling such a
system in NEST would be highly appreciated. We would like to eventually
scale to the maximum number of neurons our hardware/clusters and your
simulator can facilitate.
Thank you,
Ryan
26 Oct
26 Oct
8:24 a.m.
Dear Ryan,
Thanks for writing in! All of the things you mention are probably in principle possible
with NEST Simulator, but we would have to look in detail at some things, in particular the
dendrite-to-dendrite coupling. It is possible this could be handled by the "gap
junction" type connection in NEST. You would define as many input (or
"receive") ports ("rports") to your neuron as there are connection
sites between the two dendrites. Each of these would be connected pairwise with a gap
junction connection. See
https://nest-simulator.readthedocs.io/en/latest/models/gap_junction.html and reference [1]
there. Gap junctions in NEST use the waveform relaxation method, so please double-check
that that meets your needs.
Support for complex dendritic arbors was recently added; please see
https://nest-simulator.readthedocs.io/en/latest/models/cm_default.html for details. You
could probably customize this model with the specific set of ion channels and dynamics
that you need.
It would be interesting to tackle the remainder with NESTML
(https://nestml.readthedocs.org/). We would be happy to follow up on implementing a lookup
table in an efficient manner. Compartmental models support is also coming soon for NESTML
(you will be able to define your ion channels and dynamics in the NESTML syntax, rather
than in C++).
Hope these provide some starting points, please let us know how you fare!
With kind regards,
Charl
On Thu, Oct 13, 2022, at 17:37, Ryan O'Loughlin wrote:
Hello NEST community,
I am at the outset of a large project and I would like to start off with the right
simulator. Perhaps someone here can advise on the viability of NEST for my particular
needs. I will be modeling neuromorphic hardware with some unique design constraints, such
as:
* A complex dendritic arbor
* Dendrite potentials may be coupled with their neighbors
* Several dendrites may feed into other dendrites (a several-layer-deep dendritic tree)
* The arbor eventually feeds into the soma which has typical LIF behavior, with a
membrane potential determined by the sum of all directly connected dendrites
* The solution for the leaky-integrator term of dendrites must be drawn from a
precalculated lookup table
The question here is how best to facilitate these dendritic nuances within NEST? Any
counsel on whether it makes sense to pursue modeling such a system in NEST would be highly
appreciated. We would like to eventually scale to the maximum number of neurons our
hardware/clusters and your simulator can facilitate.
Thank you,
Ryan
_______________________________________________
NEST Users mailing list -- users(a)nest-simulator.org
To unsubscribe send an email to users-leave(a)nest-simulator.org
156
days inactive
169
days old
1 comments
2 participants
participants (2)
-
Charl Linssen -
Ryan O'Loughlin