Dear all,
In my working group, we are working on the creation of the spiking neural network of human CA1 (4.8 million neurons). We are using a specific function for defining the connection between the neurons. This function gives us 30 billion pairwise connections. We would like to know what the most efficient for the creating the network in NEST 3 using this pairwise connectivity.
Thank you in advance for your answer,
Kusch Lionel
Dear Lionel,
I assume by "specific function" you mean a function that receives lists of potential source and target neurons and then decides which neuron to connect to which in a mathematical or algorithmic way?
Could you share more information about how your connectivity is defined? If you do not want to share that information on this public mailing list, feel free to send a support request with more details to support@ebrains.eu and mention that it should be routed to me.
Best, Hans Ekkehard
--
Prof. Dr. Hans Ekkehard Plesser Head, Department of Data Science
Faculty of Science and Technology Norwegian University of Life Sciences PO Box 5003, 1432 Aas, Norway
Phone +47 6723 1560 Email hans.ekkehard.plesser@nmbu.no Home http://arken.nmbu.no/~plesser
On 18/10/2021, 12:09, "kusch lionel" lionel.kusch@univ-amu.fr wrote:
Dear all,
In my working group, we are working on the creation of the spiking neural network of human CA1 (4.8 million neurons). We are using a specific function for defining the connection between the neurons. This function gives us 30 billion pairwise connections. We would like to know what the most efficient for the creating the network in NEST 3 using this pairwise connectivity.
Thank you in advance for your answer,
Kusch Lionel
_______________________________________________ NEST Users mailing list -- users@nest-simulator.org To unsubscribe send an email to users-leave@nest-simulator.org
Dear Hans,
The "specific function" is a complex function written in Matlab using morphology of the neurons for defining the connectivity between neurons. The output of this function is a file with pre-synapsy and prost-synapsy GIDs. For the moment, there is not plan to integrate this algorithm in NEST.
My question is in the case where the connection of synapse is defined in a file (GID of input and GIDs of output), what is the best way to load it in NEST and connect the neurons?
Best regard,
Kusch Lionel
On 19/10/2021 22:16, Hans Ekkehard Plesser wrote:
Dear Lionel,
I assume by "specific function" you mean a function that receives lists of potential source and target neurons and then decides which neuron to connect to which in a mathematical or algorithmic way?
Could you share more information about how your connectivity is defined? If you do not want to share that information on this public mailing list, feel free to send a support request with more details to support@ebrains.eu and mention that it should be routed to me.
Best, Hans Ekkehard
--
Prof. Dr. Hans Ekkehard Plesser Head, Department of Data Science
Faculty of Science and Technology Norwegian University of Life Sciences PO Box 5003, 1432 Aas, Norway
Phone +47 6723 1560 Email hans.ekkehard.plesser@nmbu.no Home http://arken.nmbu.no/~plesser
On 18/10/2021, 12:09, "kusch lionel" lionel.kusch@univ-amu.fr wrote:
Dear all, In my working group, we are working on the creation of the spiking neural network of human CA1 (4.8 million neurons). We are using a specific function for defining the connection between the neurons. This function gives us 30 billion pairwise connections. We would like to know what the most efficient for the creating the network in NEST 3 using this pairwise connectivity. Thank you in advance for your answer, Kusch Lionel _______________________________________________ NEST Users mailing list -- users@nest-simulator.org To unsubscribe send an email to users-leave@nest-simulator.org
NEST Users mailing list -- users@nest-simulator.org To unsubscribe send an email to users-leave@nest-simulator.org
Dear Lionel,
Then indeed the numpy-array-based approach by Alberto is the best bet at the moment. You probably will need to do this block-wise to not overload memory. Through an EBRAINS infrastructure voucher, we currently develop a NEST backend for the SONATA format. So if your specification is stored in SONATA format, you could use that once we have a prototype ready.
Otherwise, we could into the possiblities of the conn_gen interface and libneurosim to add a backend that could read your files directly.
Best, Hans Ekkehard
--
Prof. Dr. Hans Ekkehard Plesser Head, Department of Data Science
Faculty of Science and Technology Norwegian University of Life Sciences PO Box 5003, 1432 Aas, Norway
Phone +47 6723 1560 Email hans.ekkehard.plesser@nmbu.no Home http://arken.nmbu.no/~plesser
On 20/10/2021, 10:26, "kusch lionel" lionel.kusch@univ-amu.fr wrote:
Dear Hans,
The "specific function" is a complex function written in Matlab using morphology of the neurons for defining the connectivity between neurons. The output of this function is a file with pre-synapsy and prost-synapsy GIDs. For the moment, there is not plan to integrate this algorithm in NEST.
My question is in the case where the connection of synapse is defined in a file (GID of input and GIDs of output), what is the best way to load it in NEST and connect the neurons?
Best regard,
Kusch Lionel
On 19/10/2021 22:16, Hans Ekkehard Plesser wrote: > Dear Lionel, > > I assume by "specific function" you mean a function that receives lists of potential source and target neurons and then decides which neuron to connect to which in a mathematical or algorithmic way? > > Could you share more information about how your connectivity is defined? If you do not want to share that information on this public mailing list, feel free to send a support request with more details to support@ebrains.eu and mention that it should be routed to me. > > Best, > Hans Ekkehard > > -- > > Prof. Dr. Hans Ekkehard Plesser > Head, Department of Data Science > > Faculty of Science and Technology > Norwegian University of Life Sciences > PO Box 5003, 1432 Aas, Norway > > Phone +47 6723 1560 > Email hans.ekkehard.plesser@nmbu.no > Home http://arken.nmbu.no/~plesser > > > > > On 18/10/2021, 12:09, "kusch lionel" lionel.kusch@univ-amu.fr wrote: > > Dear all, > > In my working group, we are working on the creation of the spiking > neural network of human CA1 (4.8 million neurons). We are using a > specific function for defining the connection between the neurons. This > function gives us 30 billion pairwise connections. > We would like to know what the most efficient for the creating the > network in NEST 3 using this pairwise connectivity. > > Thank you in advance for your answer, > > Kusch Lionel > > _______________________________________________ > NEST Users mailing list -- users@nest-simulator.org > To unsubscribe send an email to users-leave@nest-simulator.org > > _______________________________________________ > NEST Users mailing list -- users@nest-simulator.org > To unsubscribe send an email to users-leave@nest-simulator.org
-
Alberto Antonietti -
Hans Ekkehard Plesser -
kusch lionel