Access the full text.
Sign up today, get DeepDyve free for 14 days.
(Plesser, H., Diesmann, M., Gewaltig, M., & Morrison, A. (2015). Nest: The neural simulation tool. In D. Jaeger & R. Jung (Eds.), Encyclopedia of computational neuroscience (pp. 1849–1852). Springer New York.)
Plesser, H., Diesmann, M., Gewaltig, M., & Morrison, A. (2015). Nest: The neural simulation tool. In D. Jaeger & R. Jung (Eds.), Encyclopedia of computational neuroscience (pp. 1849–1852). Springer New York.Plesser, H., Diesmann, M., Gewaltig, M., & Morrison, A. (2015). Nest: The neural simulation tool. In D. Jaeger & R. Jung (Eds.), Encyclopedia of computational neuroscience (pp. 1849–1852). Springer New York., Plesser, H., Diesmann, M., Gewaltig, M., & Morrison, A. (2015). Nest: The neural simulation tool. In D. Jaeger & R. Jung (Eds.), Encyclopedia of computational neuroscience (pp. 1849–1852). Springer New York.
I. Wichert, S. Jee, E. Schutter, Sungho Hong (2020)
Pycabnn: Efficient and Extensible Software to Construct an Anatomical Basis for a Physiologically Realistic Neural Network ModelFrontiers in Neuroinformatics, 14
Stefano Masoli, Marialuisa Tognolina, U. Laforenza, F. Moccia, E. D’Angelo (2020)
Parameter tuning differentiates granule cell subtypes enriching transmission properties at the cerebellum input stageCommunications Biology, 3
N. Carnevale, M. Hines (2006)
The NEURON Book: Index
Mikael Lindahl, Jeanette Kotaleski (2016)
Untangling Basal Ganglia Network Dynamics and Function: Role of Dopamine Depletion and Inhibition Investigated in a Spiking Network ModeleNeuro, 3
(2009)
Front Neuroinform. 2009;3:1
Yazan Billeh, Binghuang Cai, Sergey Gratiy, Kael Dai, Ramakrishnan Iyer, N. Gouwens, R. Abbasi-Asl, Xiaoxuan Jia, J. Siegle, Shawn Olsen, C. Koch, Stefan Mihalas, A. Arkhipov (2019)
Systematic Integration of Structural and Functional Data into Multi-scale Models of Mouse Primary Visual CortexNeuron, 106
B. Hellwig (2000)
A quantitative analysis of the local connectivity between pyramidal neurons in layers 2/3 of the rat visual cortexBiological Cybernetics, 82
Allan Jones, C. Overly, S. Sunkin (2009)
The Allen Brain Atlas: 5 years and beyondNature Reviews Neuroscience, 10
Rajnish Ranjan, G. Khazen, L. Gambazzi, Srikanth Ramaswamy, Sean Hill, F. Schürmann, H. Markram (2011)
Channelpedia: An Integrative and Interactive Database for Ion ChannelsFrontiers in Neuroinformatics, 5
C. Gerfen, M. Economo, J. Chandrashekar (2018)
Long distance projections of cortical pyramidal neuronsJournal of Neuroscience Research, 96
A. Santuy, L. Tomas-Roca, José-Rodrigo Rodríguez, J. González-Soriano, F. Zhu, Zhen Qiu, S. Grant, J. DeFelipe, Á. Merchán-Pérez (2020)
Estimation of the number of synapses in the hippocampus and brain-wide by volume electron microscopy and genetic labelingScientific Reports, 10
(Anwar, H., Riachi, I., Schürmann, F., & Markram H. (2009). An approach to capturing neuron morphological diversity. In Computational neuroscience: Realistic modeling for experimentalists. De Schutter E., editor. (Cambridge: The MIT Press) 211–232. 10.7551/mitpress/9780262013277.003.0010.)
Anwar, H., Riachi, I., Schürmann, F., & Markram H. (2009). An approach to capturing neuron morphological diversity. In Computational neuroscience: Realistic modeling for experimentalists. De Schutter E., editor. (Cambridge: The MIT Press) 211–232. 10.7551/mitpress/9780262013277.003.0010.Anwar, H., Riachi, I., Schürmann, F., & Markram H. (2009). An approach to capturing neuron morphological diversity. In Computational neuroscience: Realistic modeling for experimentalists. De Schutter E., editor. (Cambridge: The MIT Press) 211–232. 10.7551/mitpress/9780262013277.003.0010., Anwar, H., Riachi, I., Schürmann, F., & Markram H. (2009). An approach to capturing neuron morphological diversity. In Computational neuroscience: Realistic modeling for experimentalists. De Schutter E., editor. (Cambridge: The MIT Press) 211–232. 10.7551/mitpress/9780262013277.003.0010.
Kael Dai, J. Hernando, Yazan Billeh, Sergey Gratiy, Judit Planas, Andrew Davison, S. Dura-Bernal, P. Gleeson, A. Devresse, B. Dichter, Michael Gevaert, J. King, Werner Geit, Arseny Povolotsky, E. Muller, J. Courcol, A. Arkhipov (2019)
The SONATA data format for efficient description of large-scale network modelsPLoS Computational Biology, 16
Werner Geit, Michael Gevaert, G. Chindemi, Christian Rössert, J. Courcol, Eilif Müller, F. Schürmann, Idan Segev, H. Markram (2016)
BluePyOpt: Leveraging Open Source Software and Cloud Infrastructure to Optimise Model Parameters in NeuroscienceFrontiers in Neuroinformatics, 10
Ruggero Scorcioni, Sridevi Polavaram, G. Ascoli (2008)
L-Measure: a web-accessible tool for the analysis, comparison and search of digital reconstructions of neuronal morphologiesNature Protocols, 3
Jakob Jordan, M. Helias, M. Diesmann, Susanne Kunkel (2020)
Efficient Communication in Distributed Simulations of Spiking Neuronal Networks With Gap JunctionsFrontiers in Neuroinformatics, 14
Hermann Cuntz, Friedrich Forstner, A. Borst, M. Häusser (2010)
One Rule to Grow Them All: A General Theory of Neuronal Branching and Its Practical ApplicationPLoS Computational Biology, 6
Iain Hepburn, Weiliang Chen, Stefan Wils, E. Schutter (2012)
STEPS: efficient simulation of stochastic reaction–diffusion models in realistic morphologiesBMC Systems Biology, 6
P. Gleeson, S. Crook, R. Cannon, M. Hines, Guy Billings, Matteo Farinella, Thomas Morse, Andrew Davison, Subhasis Ray, U. Bhalla, S. Barnes, Yoana Dimitrova, Angus Silver (2010)
NeuroML: A Language for Describing Data Driven Models of Neurons and Networks with a High Degree of Biological DetailPLoS Computational Biology, 6
J. Hjorth, Alexander Kozlov, Alexander Kozlov, Ilaria Carannante, Johanna Nylén, R. Lindroos, Yvonne Johansson, A. Tokarska, Matthijs Dorst, Shreyas Suryanarayana, G. Silberberg, J. Kotaleski, J. Kotaleski, S. Grillner (2020)
The microcircuits of striatum in silicoProceedings of the National Academy of Sciences of the United States of America, 117
S. Sudhakar, Sungho Hong, Ivan Raikov, Rodrigo Publio, Claus Lang, T. Close, Daqing Guo, Mario Negrello, E. Schutter (2017)
Spatiotemporal network coding of physiological mossy fiber inputs by the cerebellar granular layerPLoS Computational Biology, 13
Sergey Gratiy, Yazan Billeh, Kael Dai, C. Mitelut, David Feng, N. Gouwens, Nicholas Cain, C. Koch, C. Anastassiou, A. Arkhipov (2018)
BioNet: A Python interface to NEURON for modeling large-scale networksPLoS ONE, 13
J. Fuzik, Amit Zeisel, Z. Máté, D. Calvigioni, Y. Yanagawa, G. Szabó, S. Linnarsson, T. Harkany (2015)
Integration of electrophysiological recordings with single-cell RNA-seq data identifies novel neuronal subtypesNature biotechnology, 34
K. Amunts, A.C. Knoll, T. Lippert, C. Pennartz, P. Ryvlin, A. Destexhe, Viktor Jirsa, E. D’Angelo, J. Bjaalie (2019)
The Human Brain Project—Synergy between neuroscience, computing, informatics, and brain-inspired technologiesPLoS Biology, 17
A. Muñoz-Manchado, Carolina Gonzales, Amit Zeisel, Hermany Munguba, Bo Bekkouche, N. Skene, P. Lönnerberg, J. Ryge, K. Harris, S. Linnarsson, J. Hjerling-Leffler (2018)
Diversity of Interneurons in the Dorsal Striatum Revealed by Single-Cell RNA Sequencing and PatchSeqCell Reports, 24
Haroon Anwar, Imad Riachi, Sean Hill, F. Schürmann, H. Markram (2009)
An Approach to Capturing Neuron Morphological Diversity
(Akar, N. A. et al. (2019). Arbor — A morphologically-detailed neural network simulation library for contemporary high-performance computing architectures. 2019 27th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP), Pavia, Italy, 2019, pp. 274–282. 10.1109/EMPDP.2019.8671560.)
Akar, N. A. et al. (2019). Arbor — A morphologically-detailed neural network simulation library for contemporary high-performance computing architectures. 2019 27th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP), Pavia, Italy, 2019, pp. 274–282. 10.1109/EMPDP.2019.8671560.Akar, N. A. et al. (2019). Arbor — A morphologically-detailed neural network simulation library for contemporary high-performance computing architectures. 2019 27th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP), Pavia, Italy, 2019, pp. 274–282. 10.1109/EMPDP.2019.8671560., Akar, N. A. et al. (2019). Arbor — A morphologically-detailed neural network simulation library for contemporary high-performance computing architectures. 2019 27th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP), Pavia, Italy, 2019, pp. 274–282. 10.1109/EMPDP.2019.8671560.
M. Humphries, R. Wood, K. Gurney (2009)
Dopamine-modulated dynamic cell assemblies generated by the GABAergic striatal microcircuitNeural networks : the official journal of the International Neural Network Society, 22 8
M. Hines, Andrew Davison, Eilif Müller (2008)
NEURON and PythonFrontiers in Neuroinformatics, 3
R. Cannon, D. Turner, G. Pyapali, H. Wheal (1998)
An on-line archive of reconstructed hippocampal neuronsJournal of Neuroscience Methods, 84
P. Berthet, Mikael Lindahl, Philip Tully, J. Hellgren-kotaleski, A. Lansner (2016)
Functional Relevance of Different Basal Ganglia Pathways Investigated in a Spiking Model with Reward Dependent PlasticityFrontiers in Neural Circuits, 10
M. Economo, Johan Winnubst, Erhan Bas, Tiago Ferreira, J. Chandrashekar (2019)
Single‐neuron axonal reconstruction: The search for a wiring diagram of the brainJournal of Comparative Neurology, 527
S. Grillner, N. Ip, C. Koch, W. Koroshetz, H. Okano, Miri Polachek, M. Poo, T. Sejnowski (2016)
Worldwide initiatives to advance brain researchNature Neuroscience, 19
Mélissa Cizeron, Zhen Qiu, Babis Koniaris, R. Gokhale, N. Komiyama, E. Fransén, S. Grant (2020)
A brainwide atlas of synapses across the mouse life spanScience, 369
(Carnevale, T., & Hines, M. (2006). The NEURON book (p. 2006). Cambridge University Press. 10.1017/CBO9780511541612.)
Carnevale, T., & Hines, M. (2006). The NEURON book (p. 2006). Cambridge University Press. 10.1017/CBO9780511541612.Carnevale, T., & Hines, M. (2006). The NEURON book (p. 2006). Cambridge University Press. 10.1017/CBO9780511541612., Carnevale, T., & Hines, M. (2006). The NEURON book (p. 2006). Cambridge University Press. 10.1017/CBO9780511541612.
(Williams, T. J., Balakrishnan, R., Delalondre, F., Schuermann, F., Muller, E., &Gewaltig, M. O. (2018). Large-Scale Simulation of Brain Tissue, Blue Brain Project, EPFL. United States: N. p., 2018. Web. 10.2172/1483995.)
Williams, T. J., Balakrishnan, R., Delalondre, F., Schuermann, F., Muller, E., &Gewaltig, M. O. (2018). Large-Scale Simulation of Brain Tissue, Blue Brain Project, EPFL. United States: N. p., 2018. Web. 10.2172/1483995.Williams, T. J., Balakrishnan, R., Delalondre, F., Schuermann, F., Muller, E., &Gewaltig, M. O. (2018). Large-Scale Simulation of Brain Tissue, Blue Brain Project, EPFL. United States: N. p., 2018. Web. 10.2172/1483995., Williams, T. J., Balakrishnan, R., Delalondre, F., Schuermann, F., Muller, E., &Gewaltig, M. O. (2018). Large-Scale Simulation of Brain Tissue, Blue Brain Project, EPFL. United States: N. p., 2018. Web. 10.2172/1483995.
(Lindahl, M., & Hellgren Kotaleski, J. (2017). Untangling basal ganglia network dynamics and function: role of dopamine depletion and inhibition investigated in a spiking network model. eNeuro, 3(6), ENEURO.0156-16.2016. 10.1523/ENEURO.0156-16.2016.)
Lindahl, M., & Hellgren Kotaleski, J. (2017). Untangling basal ganglia network dynamics and function: role of dopamine depletion and inhibition investigated in a spiking network model. eNeuro, 3(6), ENEURO.0156-16.2016. 10.1523/ENEURO.0156-16.2016.Lindahl, M., & Hellgren Kotaleski, J. (2017). Untangling basal ganglia network dynamics and function: role of dopamine depletion and inhibition investigated in a spiking network model. eNeuro, 3(6), ENEURO.0156-16.2016. 10.1523/ENEURO.0156-16.2016., Lindahl, M., & Hellgren Kotaleski, J. (2017). Untangling basal ganglia network dynamics and function: role of dopamine depletion and inhibition investigated in a spiking network model. eNeuro, 3(6), ENEURO.0156-16.2016. 10.1523/ENEURO.0156-16.2016.
Michael Reimann, J. King, Eilif Müller, Srikanth Ramaswamy, H. Markram (2015)
An algorithm to predict the connectome of neural microcircuitsFrontiers in Computational Neuroscience, 9
M. Gewaltig, M. Diesmann (2007)
NEST (NEural Simulation Tool)Scholarpedia, 2
B. Torben-Nielsen (2014)
An Efficient and Extendable Python Library to Analyze Neuronal MorphologiesNeuroinformatics, 12
N. Akar, Benjamin Cumming, V. Karakasis, Anne Küsters, W. Klijn, A. Peyser, Stuart Yates (2019)
Arbor — A Morphologically-Detailed Neural Network Simulation Library for Contemporary High-Performance Computing Architectures2019 27th Euromicro International Conference on Parallel, Distributed and Network-Based Processing (PDP)
P. Kumbhar, M. Hines, Jeremy Fouriaux, A. Ovcharenko, J. King, F. Delalondre, F. Schürmann (2019)
CoreNEURON : An Optimized Compute Engine for the NEURON SimulatorFrontiers in Neuroinformatics, 13
O. Gokce, O. Gokce, Geoffrey Stanley, B. Treutlein, N. Neff, J. Camp, R. Malenka, Patrick Rothwell, M. Fuccillo, T. Südhof, T. Südhof, S. Quake, S. Quake (2016)
Cellular Taxonomy of the Mouse Striatum as Revealed by Single-Cell RNA-Seq.Cell reports, 16 4
K. Gurney, T. Prescott, P. Redgrave (2001)
A computational model of action selection in the basal ganglia. II. Analysis and simulation of behaviourBiological Cybernetics, 84
Kerry Brown, G. Barrionuevo, A. Canty, V. Paola, J. Hirsch, G. Jefferis, Ju Lu, M. Snippe, I. Sugihara, G. Ascoli (2011)
The DIADEM Data Sets: Representative Light Microscopy Images of Neuronal Morphology to Advance Automation of Digital ReconstructionsNeuroinformatics, 9
Arpiar Saunders, Evan Macosko, Alec Wysoker, Melissa Goldman, Fenna Krienen, H. Rivera, Elizabeth Bien, Matthew Baum, L. Bortolin, Shuyu Wang, A. Goeva, J. Nemesh, Nolan Kamitaki, Sara Brumbaugh, D. Kulp, S. Mccarroll (2018)
Molecular Diversity and Specializations among the Cells of the Adult Mouse BrainCell, 174
T. Williams, R. Balakrishnan, F. Delalondre, F. Schuermann, E. Muller, M. Gewaltig (2018)
Large-Scale Simulation of Brain Tissue, Blue Brain Project, EPFL
William Podlaski, Alexander Seeholzer, Lukas Groschner, G. Miesenböck, Rajnish Ranjan, T. Vogels (2017)
Mapping the function of neuronal ion channels in model and experimenteLife, 6
R. McDougal, Thomas Morse, N. Carnevale, Luis Marenco, Rixin Wang, M. Migliore, P. Miller, G. Shepherd, M. Hines (2016)
Twenty years of ModelDB and beyond: building essential modeling tools for the future of neuroscienceJournal of Computational Neuroscience, 42
Rosanna Migliore, C. Lupascu, L. Bologna, Armando Romani, J. Courcol, S. Antonel, Werner Geit, A. Thomson, A. Mercer, S. Lange, Joanne Falck, Christian Rössert, Ying Shi, Olivier Hagens, Maurizio Pezzoli, T. Freund, S. Káli, Eilif Müller, F. Schürmann, H. Markram, M. Migliore (2018)
The physiological variability of channel density in hippocampal CA1 pyramidal cells and interneurons explored using a unified data-driven modeling workflowPLoS Computational Biology, 14
(2020)
Construct an Anatomical Basis for a Physiologically
G. Rosen, Robert Williams (2001)
Complex trait analysis of the mouse striatum: independent QTLs modulate volume and neuron numberBMC Neuroscience, 2
Cristina Colangelo, P. Shichkova, D. Keller, H. Markram, Srikanth Ramaswamy (2019)
Cellular, Synaptic and Network Effects of Acetylcholine in the NeocortexFrontiers in Neural Circuits, 13
H. Okano, A. Miyawaki, K. Kasai (2015)
Brain/MINDS: brain-mapping project in JapanPhilosophical Transactions of the Royal Society B: Biological Sciences, 370
Hanson Ho, M. Both, A. Siniard, Sasha Sharma, James Notwell, M. Wallace, Dino Leone, Amy Nguyen, Eric Zhao, Hannah Lee, D. Zwilling, K. Thompson, S. Braithwaite, M. Huentelman, T. Portmann (2018)
A Guide to Single-Cell Transcriptomics in Adult Rodent Brain: The Medium Spiny Neuron Transcriptome RevisitedFrontiers in Cellular Neuroscience, 12
Masood Akram, S. Nanda, Patricia Maraver, R. Armañanzas, G. Ascoli (2018)
An open repository for single-cell reconstructions of the brain forestScientific Data, 5
T. Insel, Story Landis, F. Collins (2013)
The NIH BRAIN InitiativeScience, 340
J. Hjorth, K. Blackwell, Jeanette Kotaleski (2009)
Gap Junctions between Striatal Fast-Spiking Interneurons Regulate Spiking Activity and Synchronization as a Function of Cortical ActivityThe Journal of Neuroscience, 29
H. Markram, E. Muller, Srikanth Ramaswamy, Michael Reimann, M. Abdellah, Carlos Sanchez, A. Ailamaki, L. Alonso-Nanclares, N. Antille, Selim Arsever, Guy Kahou, Thomas Berger, A. Bilgili, Nenad Buncic, Athanassia Chalimourda, G. Chindemi, J. Courcol, F. Delalondre, V. Delattre, S. Druckmann, Raphael Dumusc, J. Dynes, S. Eilemann, Eyal Gal, Michael Gevaert, Jean-Pierre Ghobril, A. Gidon, Joe Graham, Anirudh Gupta, V. Haenel, Etay Hay, T. Heinis, J. Hernando, M. Hines, Lida Kanari, D. Keller, J. Kenyon, G. Khazen, Yihwa Kim, J. King, Z. Kisvárday, P. Kumbhar, S. Lasserre, Jean-Vincent Bé, Bruno Magalhães, Á. Merchán-Pérez, J. Meystre, Benjamin Morrice, J. Muller, Alberto Muñoz-Céspedes, S. Muralidhar, Keerthan Muthurasa, D. Nachbaur, T. Newton, Max Nolte, A. Ovcharenko, Juan Palacios, L. Pastor, R. Perin, Rajnish Ranjan, Imad Riachi, José-Rodrigo Rodríguez, J. Riquelme, Christian Rössert, K. Sfyrakis, Ying Shi, J. Shillcock, G. Silberberg, R. Silva, F. Tauheed, Martin Telefont, Maria Toledo-Rodriguez, Thomas Tränkler, Werner Geit, Jafet Díaz, Richard Walker, Yun Wang, Stefano Zaninetta, J. DeFelipe, Sean Hill, Idan Segev, F. Schürmann (2015)
Reconstruction and Simulation of Neocortical MicrocircuitryCell, 163
Stefano Casali, E. Marenzi, Chaitanya Medini, C. Casellato, E. D’Angelo (2019)
Reconstruction and Simulation of a Scaffold Model of the Cerebellar NetworkFrontiers in Neuroinformatics, 13
Subhasis Ray, U. Bhalla (2008)
PyMOOSE: Interoperable Scripting in Python for MOOSEFrontiers in Neuroinformatics, 2
Mark Wilkinson, M. Dumontier, I. Aalbersberg, Gabrielle Appleton, M. Axton, A. Baak, N. Blomberg, J. Boiten, Luiz Santos, P. Bourne, J. Bouwman, A. Brookes, Tim Clark, M. Crosas, I. Dillo, Olivier Dumon, S. Edmunds, C. Evelo, R. Finkers, Alejandra González-Beltrán, A. Gray, Paul Groth, C. Goble, J. Grethe, J. Heringa, P. Hoen, R. Hooft, Tobias Kuhn, Ruben Kok, J. Kok, S. Lusher, M. Martone, Albert Mons, A. Packer, Bengt Persson, P. Rocca-Serra, M. Roos, Rene Schaik, Susanna-Assunta Sansone, E. Schultes, T. Sengstag, Ted Slater, George Strawn, M. Swertz, Mark Thompson, J. Lei, E. Mulligen, Jan Velterop, A. Waagmeester, P. Wittenburg, K. Wolstencroft, Jun Zhao, B. Mons (2016)
The FAIR Guiding Principles for scientific data management and stewardshipScientific Data, 3
G. Einevoll, A. Destexhe, M. Diesmann, S. Grün, Viktor Jirsa, M. Kamps, M. Migliore, T. Ness, H. Plesser, F. Schürmann (2019)
The Scientific Case for Brain SimulationsNeuron, 102
Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations
Lida Kanari, Srikanth Ramaswamy, Ying Shi, S. Morand, J. Meystre, R. Perin, M. Abdellah, Yun Wang, K. Hess, H. Markram (2019)
Objective Morphological Classification of Neocortical Pyramidal CellsCerebral Cortex (New York, NY), 29
Simulation of large-scale networks of neurons is an important approach to understanding and interpreting experimental data from healthy and diseased brains. Owing to the rapid development of simulation software and the accumulation of quantitative data of different neuronal types, it is possible to predict both computational and dynamical properties of local microcircuits in a ‘bottom- up’ manner. Simulated data from these models can be compared with experiments and ‘top-down’ modelling approaches, successively bridging the scales. Here we describe an open source pipeline, using the software Snudda, for predicting microcir- cuit connectivity and for setting up simulations using the NEURON simulation environment in a reproducible way. We also illustrate how to further ‘curate’ data on single neuron morphologies acquired from public databases. This model building pipeline was used to set up a first version of a full-scale cellular level model of mouse dorsal striatum. Model components from that work are here used to illustrate the different steps that are needed when modelling subcortical nuclei, such as the basal ganglia. . . . . . Keywords Large-scale simulations Striatum Basal ganglia Brain microcircuits Synaptic connectivity Model building pipeline Introduction 2019; Billeh et al., 2020;Hjorthetal., 2020). Collecting data from the brain at multiple biological scales from mouse, non- Neuroscientists are producing data at an ever growing rate, and human primates, and human, are important goals of several of sharing the data in public databases. Within the computational the big brain initiatives (Insel et al., 2013; Amunts et al., 2019; neuroscience field, hypothesis-driven modelling has over many Okano et al., 2015; Grillner et al., 2016), and will further facil- decades generated new ideas that in turn have been tested via itate and speed up this modelling process. In parallel, various experiments. Recently a data-driven mechanistic modelling ap- brain simulation tools have been optimized to capitalize on proach has also gained ground thanks to new technologies supercomputers (Hepburn et al., 2012; Plesser et al., 2015; allowing the collection of large quantities of useful data. In Carnevale & Hines, 2006; Hines et al., 2009; Kumbhar et al., particular, large-scale spiking neural network models have 2019; Ray & Bhalla, 2008; Gleeson et al., 2010;Jordanet al., been reconstructed in a data-driven manner and simulated 2020;Akaret al., 2019). In this respect, the principles of FAIR (Markram et al., 2015; Gratiy et al., 2018; Migliore et al., – Findable, Accessible, Interoperable, Reusable (Wilkinson 2018;Casalietal., 2019;Einevolletal., 2019;Kanarietal., et al., 2016) – are important for catalysing this process, both with regard to the experimental data, the data-driven models as well as the software used during the modelling and simulation process. We also believe that to be able to reproduce the actual * J. J. Johannes Hjorth model reconstruction process, given the same or new additional hjorth@kth.se data, is one important aspect of the FAIR criteria when making the modelling process transparent, repeatable, reusable and Science for Life Laboratory, School of Electrical Engineering and comparable. Computer Science, KTH Royal Institute of Technology, Here we present our open source modelling pipeline that SE-10044 Stockholm, Sweden facilitates a reproducible, data-driven reconstruction of cellu- Department of Neuroscience, Karolinska Institute, lar level network/microcircuit models. This pipeline inspired SE-17172 Stockholm, Sweden 686 Neuroinform (2021) 19:685–701 by the cortical column microcircuit (Markram et al., 2015)has open source toolchain for building and simulating ana- been applied to predict a full-scale microcircuit model of the tomically constrained biologically detailed neural net- mouse dorsal striatum (Hjorth et al., 2020). Snudda is a soft- works including morphology-based neuron touch detec- ware to create a detailed network of connected neurons, where tion; (ii) we illustrate the use of this platform on the ex- the connectivity is derived from reconstructed neuronal mor- ample of the striatal microcircuit, implemented at a very phologies as well as from more qualitative experimental detailed level and accuracy; (iii) we include all tools and knowledge (see also Reimann et al., 2015). ‘Snudda’ means parameters in the source code repository, enabling other ‘touch’ in Swedish, and it supports the creation of a network labs to reproduce as well as reconstruct our striatal model with connectivity based on touch detection. If detailed mor- with new data when it becomes available. phological data exist, the algorithm looks for close appositions between axons and dendrites, which are locations for putative synapses. Thus the morphology restricts where connections Getting Started with Snudda can be positioned. Snudda can also define the axon using a probability cloud if a reconstructed axon is missing. This is an Snudda is available for download from GitHub (https://github. extension of the method where the connection probability is com/hjorthmedh/Snudda) or from PyPi through pip3 install proportional to the overlap of two spheres representing axons snudda. Both source code and the data files necessary to set and dendrites (Humphries et al., 2009). Based on a set of rules, up a striatal network are provided, Table 1 provides an as described below, the putative synapses are then pruned to overview of the Snudda directory structure. Snudda is match the connectivity seen from pairwise experimental re- compatible with Linux, Mac and Windows 10. cordings, or other types of connectivity experiments. The In the directory snudda/data/neurons/<region> there are same technique can be applied to also place gap junctions. separate subdirectories for each neuron type (<region> in The generated network can then be simulated using parallel our use case is striatum). Each of those directories contains NEURON (Carnevale & Hines, 2006). Similar approaches multiple subdirectories, one for each unique morphology from have been used to build the somato-sensory cortex microcir- that neuron type. The neuron directories include the morphol- cuit (Markram et al., 2015; Colangelo et al., 2019), visual ogy in SWC format, a JSON parameter file with one or more cortex model (Billeh et al., 2020; Dai et al., 2020), cerebellar sets of optimised neuron parameters from BluePyOpt, a JSON network (Sudhakar et al., 2017; Casali et al., 2019;Wichert mechanism file specifying which mechanisms are present in et al., 2020) and hippocampal neurons (Migliore et al., 2018). each compartment, and a JSON modulation file which spec- The reconstruction of a local microcircuit model (such as ifies the neuron modulation of the neuron. The JSON file striatum) consists of the following steps: a) experimental data format was chosen as it is a standardised and human readable acquisition of the electrophysiological and morphological prop- way to store structured data. The neurons folder also has a erties of neuronal types, and also characterisation of synapses, b) mechanisms folder containing the NEURON model descrip- optimization of neuron and synapse models, c) placement of the tion language .mod files with definitions of ionic mechanisms. model neurons in the brain volume to be modelled, d) prediction To keep the networks separate, each generated network has of microcircuit connectivity in silico, e) constraining and emulat- its own directory which contains a network.json file that links ing inputs for the model, and finally f) simulating the microcir- together all the different components that make up the network. cuitry. Our software Snudda is used for steps c)-f). Software The network.json file can be manually created, or in the case of Treem for improving the morphological reconstructions in pre- the striatal network there is a way to automatically generate a paratory step b) is described at the end. The code is publicly network.json file of user specified size. The script init.py can be available on GitHub (https://github.com/Hjorthmedh/Snudda/) extended to create networks of other brain structures. A Jupyter and (https://github.com/a1eko/treem). Below we will go notebook in examples/notebooks shows an alternative example through the different steps and provide code to set up an for how to define brain slices and other structures. example network using Snudda, followed by explanation of the In the main Snudda directory there is an examples folder configuration files, network building and simulation process, as with useful scripts and Jupyter notebooks for generating and well as some preprocessing options. The network example running networks. The directory snudda/plotting contains corresponds to a 0.5 mm cube within the mouse striatum. We scripts to plot simulation results as well as visualise the net- will assume that we already have a set of electrophysiologically work or parts of it using Blender (https://www.blender.org/). optimized neurons and synapses, e.g. using the optimization tool BluePyOpt (Van Geit et al., 2016). Examples of neuron and synapse models relevant for striatum are provided on GitHub Use Case: Striatal Microcircuit (in the snudda/examples folder with scripts and notebooks). Our approach offers novel contributions in the follow- First we create an example striatal network, then further down ing aspects: (i) we design and present a complete, free and we go through all configuration details. The network- Neuroinform (2021) 19:685–701 687 Table 1 Snudda directory snudda Snudda directory, code usually executed from this folder structure snudda/data Snudda data folder snudda/data/mesh 3D meshes for brain structures snudda/data/neurons/striatum morphology and parameters for striatal neurons in subfolders snudda/data/neurons/mechanisms NEURON mechanisms folder with mod-files snudda/data/synapses synapse model parameters snudda/data/input_config input configuration files snudda/input_tuning scripts for synaptic input tuning snudda/plotting scripts for plotting networks, includes subfolder with blender scripts snudda/utils scripts for small tasks examples examples on how to run snudda, contains shell scripts and notebooks tests contains unit and regression tests tests/networks networks used or created for testing tests/validation morphologies and mechanisms used for testing config.json can be generated using the snudda init shell com- configuration file network.json specifying e.g. 10,062 neurons mand. In the below example a homogeneous cube with 0.5 mm in a directory called smallSim. side length in the mouse striatum is generated (Fig. 1a). This corresponds to 10,062 neurons (Fig. 1b) using the estimated average density of striatal neurons (Rosen & Williams, 2001), Next we need to place the neurons in a specified volume. but the number of neurons can be varied depending on compu- For large simulations the neurons are placed inside a volume tational resources and research questions. The neurons are tak- representing the mouse striatum (Fig. 1a), while smaller net- en from the data/neurons/striatum directory, where every neu- works use a simple cube (Fig. 1b, c). This is done to preserve ron type has its own directory, e.g. dspn or ispn. A neuron type physiological neuron densities in the simulations. The mesh is represented by one or more single-cell models, each in its definition of the striatal volume, or other structures, can be own subdirectory as described above. When the network is extracted from databases such as the Allen Brain Atlas of initialised, the init.py code will look in the folders of the differ- the mouse brain. The command to place the neurons inside ent neuron types and instantiate single-cell models in random the volume defined by the mesh is: order. No modifications of the neuron models other than rota- tions are applied at runtime. In order to improve the cell diver- sity, one should populate the neuron types directories with a The next step is the touch detection (Fig. 2a). Here the sufficient number of different single-cell models. algorithm voxelizes the space, and looks for overlaps (within The following commands can be used in a terminal win- a certain predefined distance) between axons and dendrites dow to generate the example network that we have used for from different neurons (Fig. 3a, b)(Hellwig, 2000). the figures in this article. We go into more detail in the sec- tions later in the article. There are additional Jupyter notebook examples in examples/notebooks and examples/ The touch detection will create a putative set of synapses at Neuroinformatics2021. The first step is to create a all the close appositions between axons and dendrites. Fig. 1 Example of the volume definition. a Selection of the volume of somas placed within the red cube. c Illustration of 100 neurons showing interest (red cube, size of the side 500 μm) inside the left part of the dorsal the complexity of axons and dendrites striatum (blue shells beneath the cerebral cortex). b The 10,062 neuron 688 Neuroinform (2021) 19:685–701 Fig. 2 Example of the synaptic pruning procedure when connecting fraction f1 of all synapses is removed. The soft max (SM) synapse filter neurons within the microcircuit. a Putative synapses detected between does not disconnect any connected pairs as it only reduces the number of iSPN and dSPN shown on top, and remaining synapses after pruning synapses of pairs that are connected by a large number of synapses. shown below. b Connection probability as a function of distance after Finally mu2 filters neuron pairs with few synapses, and leads to a large each of the pruning steps. Distance dependent pruning (DP) filters syn- reduction in connectivity. c Number of synapses between connected apses based on the distance to the soma on the postsynaptic neuron. A pairs. d Number of connected neighbours each post synaptic neuron has Fig. 3 Touch detection and pruning. a Illustration of ball and stick neurons with soma and dendrites marked in black, axon in red and synapse in green. However, not all close appositions correspond to real synap- b Corresponding two neurons in the hypervoxel representation. The ses, as explained in detail further down. The next step prunes neurites are traced by taking small steps along Δx, Δy, Δz corresponding the set of putative synapses to match the connectivity seen in to the direction of the neurite. Where axon and dendrites occupy the same experimental pairwise recordings (Fig. 2b, distance dependent voxel a putative synapse is detected, here marked by a green dot. The volume of the soma is also voxelized. c Ball-and-stick neurons arranged connectivity). The rules used for pruning are qualitatively so their synapses are on a grid, with four synapses connecting every neuron similar to what Markram et al. (2015) created for their cortical pair. Pruning with parameter f1 = 1, 0.5 and 0.25 keeping 100%, 50% and network. The parameters for the pruning (Fig. 3c-g) are spec- 25% of all synapses, respectively. Number of synapses retained shown ified in the network.json file, explained more in detail below. under each network (including a numeric estimate of mean and standard deviation, n = 1000). d Combination of f1 and mu2 pruning. Here P = f1 · The command to perform the pruning is: 1.0 / (1.0 + exp(−8.0 / mu2 ·(nSynapses - mu2))), where P is the probability to keep a synapse, nSynapses is the total number of synapses connecting the neuron pre-post pair. e softMax pruning step. If there are more than softMax Code to generate figures analysing the connectivity (Fig. synapses then the probability of keeping synapses between that pair is P =2 2c, d) (distance dependent connection probability, histogram · softMax / ((1 + exp(−(nSynapses - softMax)/5))· nSynapses). f Pruning a3 = 1, 0.5 and 0.25 removing all synapses between a connected pair in 0%, showing the number of synapses between connected neigh- 50% and 75% of the cases, respectively. g Distance dependent pruning of bours, histogram showing the number of connected neigh- proximal, medial and distal synapses. Jupyter Notebooks to generate this bours) is in snudda/analyse_striatum.py, also see examples/ figure are available on Snudda GitHub in the examples/ Neuroinformatics2021 folder) Neuroinformatics2021 for Jupyter notebooks. Neuroinform (2021) 19:685–701 689 690 Neuroinform (2021) 19:685–701 Next we need to generate external synaptic input for the 469 million intrastriatal GABAergic synapses which corre- network simulation. Here we specify how much time we want sponds to 75.6 million synapses per mm , well within the ex- to generate inputs for. The parameters for the synaptic input perimental range (18–285 million). This also agrees with an- from cortex and thalamus are defined in a separate JSON file: other independent study by Cizeron et al. (2020). Snudda Configuration Explained The last step involves compiling the .mod files, and then running the simulation. Connectivity Configuration Running the command snudda init mysimulation –size N gen- There are two functions in the plotting directory that allow the erates the Snudda network configuration file network- user to plot either the spike raster or the voltage traces, plotting/ config.json where N is the size of the network to create. The plot_spike_raster.py (Fig. 4) and plotting/plot_traces.py, the lat- network-config.json contains the blueprints for the striatal net- ter requires the user to have run the simulation with the –voltOut work hierarchically organised into the blocks RandomSeed, parameter to also save the voltage traces. The simulation output Volume, Units, Connectivity and Neurons. All parameters in the configuration file are specified in SI units. Below we go files are stored in the $simName/simulations directory. into more detail how Snudda is configured. The RandomSeed block specifies the random seeds used for the different steps of the network creation. Validation Against Anatomical Data The network building results in Fig. 2 have been validated in our previous study. Figures 8, S4-S7 in Hjorth et al. (2020) showed experimental pair-wise connection probability be- tween neuron types and how the Snudda generated network matched the data, and matched estimations of number of syn- apses between connected pairs. To validate the mean synaptic density in the simulated net- work, we use the data from the recent studies with genetic labeling and electron microscopy techniques. Santuy et al. (2020) estimate a density of 1.41 synapses/μm in the striatum, The Volume block can contain several named volumes, with 4.4% of the symmetric synapses (range 1.29–20.23%), such as for example Striatum, Cortex, Thalamus. For each which corresponds to 62 million GABAergic synapses per volume block we define type (e.g. mesh), dMin (minimum 3 3 mm . In our striatal model 500,000 neurons (6.2mm ) has distance between somas in the volume), meshFile (this Fig. 4 Simulation of 10,062 striatal neurons (4872 dSPN, 4872 iSPN, 133 FS, 113 ChIN, 72 LTS) receiving cortical and thalamic input. The cortical drive is increased for half a second at 1.0 s. The histograms above the spike raster show the total number of spikes in the respective neuron populations. (See the striatum_ example* notebooks in the example/notebooks folder, note that the size of the network was changed) Neuroinform (2021) 19:685–701 691 specifies the Wavefront OBJ file that defines the mesh enclosing the volume) and meshBinWidth (voxelization size of the mesh for determining what is inside and outside the mesh during cell placement). Future versions of Snudda will allow for density variations within the volume and directional gradients for the neurons. FG An example block for PopulationUnits looks as follows. Here two units are defined: UnitID 1with 20% of dSPN and iSPN neurons in Striatum,and UnitID 2with 30%. In the Connectivity block we define the rules guiding how the different neuron populations are connected togeth- er. Each connection pair has its own block (e.g. “iSPN,dSPN”). In the example below this is illustrated with the iSPN-dSPN pair (indirect-pathway and direct-pathway striatal projection neurons, respectively), which are con- nected by GABA synapses. Each connection has a conductance parameter, which spec- ifies the mean and standard deviation of the conductance. The channelParameters gives flexibility by specifying a dictio- nary with the channel specific parameters, in this case it is tau1, tau2, failRate (the synapse failure rate) that are passed directly to the NEURON channel model. Next parameterFile (JSON file with additional channel parameters) and modFile (NEURON channel .mod file). The final two blocks pruning and pruningOther specify the pruning parameters for neurons 692 Neuroinform (2021) 19:685–701 within the same population unit, and neurons in different pop- ulation units. The pruning parameters help shape the connec- P ¼ f1 P P a3 P keep mu SM dist tivity by parameterising the rules that define which putative The examples given in Fig. 3c-g are synthetic, their pur- connections should be removed, and which ones should be pose is to illustrate pruning rules. The f1 parameter defines kept. The probability to keep a synapse is equal to the product how large a fraction of the putative synapses should be of the individual pruning steps: kept, a value of 1.0 or None means that this pruning step Neuroinform (2021) 19:685–701 693 is not used (Fig. 3c). For f1 = 0.5 we would expect on av- 0.5 we have two parts to the pruning, first f1 where half the erage 0.5 · 400 = 200 synapses kept, and for f1 =0.25 we synapses are removed, then mu2 which operates on all the expect 0.25 · 400 = 100 synapses (c.f. 203 and 104 synapses synapses between a coupled pair. We thus look at the 100 randomly selected for f1 =0.5 and f1 =0.25, respectively, possible neuron pairs. The expected number of synapses in Fig. 3c). E ¼ 100 ∑ P P ðÞ n n,with P the probability of a syn n mu n The mu2 defines a sigmoid curve used to decide whether to neuron pair having n synapses after the f1 pruning. We then keep or remove all synapses between a coupled pair of neu- get rons (Fig. 3d): 4 4 4 4 4 4 0:935 þ 3 0:5 þ 2 0:065 þ 1 0:005 0:5 100 0 1 2 3 P ¼ 1=ðÞ 1 þ expðÞ −8=mu2ðÞ n−mu2 mu ¼ 65:9 With mu2 =3, we have P (n = 4) = 93.5%, P (n = mu mu synapses, and for f1 = 0.25 we expect on average 3) = 50%, P (n =2)= 6.5%, P (n = 1) = 0.5%. Thus for mu mu f1 = 1 we expect 400 · 0.935 = 374 synapses left. For f1 = 4 4 4 3 1 4 2 2 4 1 3 4 0:935 0:25 þ 3 0:5 0:25 0:75 þ 2 0:065 0:25 0:75 þ 1 0:005 0:25 0:75 100 ¼ 11:4 0 1 2 3 synapses. neuron populations. Here each neuron template has its The softMax specifies at which value we start applying own block. For each neuron template we specify four a soft cap to the total number of synapses between the files: morphology (a SWC file defining the soma, axon pair: and dendrites), parameters (neuron parameters optimised using BluePyOpt), mechanisms (NEURON mechanisms), modulation (a JSON file defining the neuromodulation). P ¼ 2softMax=ðÞ ðÞ 1 þ expðÞ −ðÞ n−softMax =5 n SM The template can be used to define multiple neurons, the number defined by num.The hoc parameter is op- tional, and intended to be used in the future when where P is probability to keep a synapse, and n is the SM exporting to SONATA format for use with initial number of synapses between the pair of neurons Neurodamus (Williams et al., 2018). The neuronType (Fig. 3e). For n =4 and softMax =3 yields P =82.5% SM can be either neuron or virtualNeuron, the latter can resulting in on average 400 · 0.825 = 330 synapses left. be used to define axons from other structures providing For softMax = 2 around 239 synapses will remain, and for input to the striatum. The rotationMode lets us specify softMax = 1 on average 129 synapses are kept. if the neurons should be left unrotated, or rotated in The a3 parameter specifies which fraction of all con- some manner. The volumeID defines which volume the nected pairs to keep, e.g. 0.8 means that 20% of all con- neurons belong to. nected pairs will have all their synapses removed (Fig. 3f). Here a3 =1,0.5and0.25resultinonaverage400, 200 and 100 synapses left, correspondingly. The distPruning defines a distance d dependent function P : d → [0,1], where d is the distance from the soma dist along the dendrites (Fig. 3g). The expected number of synapses for the distance dependent pruning is given by E ¼ 20 ∑ PdðÞ where P(d) is one of the equations syn k k¼1 in Fig. 3g. In this example the distances to the soma are d = 56, 65, 86, 95, 116, 125, 146, 155, 176, 185, 206, 215, 236, 245, 266, 275, 296, 305, 326 and 335 μm(with The $DATA keyword is a shorthand for the snudda/data 20 putative synapses at each distance). The expected folder. number of synapses in the three cases are thus 107, 134 and 91, respectively. Continuing our look at the network configuration file structure, the final block Neurons defines the different 694 Neuroinform (2021) 19:685–701 Configuring External Synaptic Input The input spikes to the network are generated as prescribed in the input configuration file input.json in JSON format. Below we will give a simple example of how to set up input, and there are more examples available on Github in the examples/ notebooks directory. In this example the dSPN will each receive 200 inputs, with 1 Hz Poisson random spikes. The configuration also specifies the conductance and the tmGlut mod file that is used by NEURON to simulate the input synapses. To complement the cortical (Ctx) input with thalamic, add a second input block with parameters inside the dSPN target block and give it a name e.g. Thalamic. The entire dSPN population will then receive both cortical and thalamic inputs. Only one target block is applied to each neuron. When Snudda generates input for the network it iterates through all the different neurons in the simulation and picks the most specific target block that matches that neuron. In the example below a dSPN with neuron ID 5 and morphology dSPN_0 will match all three blocks, but the neuron ID is most specific so the “5” block will be used. A dSPN_1 morphology neuron will only match the dSPN block and will use that. In the dSPN target configuration the start, end and frequency are specified as vectors. Here the input is 4 Hz at 2–3 s, and2Hz at 5–7 s. We can also use population units to specify heterogenous external input to the target volume, see examples/notebooks on Github. To create advanced inputs not supported by Snudda the custom spike times can be read from a CSV file (with one spike train per row) by using “generator”: “csv” and “csvFile”:“path/to/your/csvfile”. A more complex example using additional input generation functionality is given below. The type defines what sort of input the synapses form, e.g. AMPA_NMDA or GABA.The number of inputs to each neuron can either be defined directly using nInputs or indirectly by specifying the density of inputs Neuroinform (2021) 19:685–701 695 synapseDensity along the dendrites. If both parameters are no other neurons within a distance dMin from it, the position is given, the code will use the density but scale it so that the accepted. To avoid an artificial increase of neuron density at nInputs are created. There is also an optional parameterFile the border, the neuron positions placed outside the mesh are that can be used to define a set of parameters for the synaptic also tracked. These padding positions are not counted towards channel. the total, and are discarded afterwards. Orientation of the neu- The populationCorrelation describes how correlated the rons in the striatum is completely random, but it is possible to Poisson input is that is generated by mixing a shared mother specify other ways to sample the orientation. process (each spike is included with probability P =sqrt(C)) For touch detection the space is divided into voxels of 3 μm and a number of independent child processes (inclusion prob- side length (Fig. 3b). Synapses are only detected when axon ability 1 - P) to get the resulting input trains (Hjorth et al., and dendrite are present in the same voxel. The maximal in- 2009). teraction distance is thus decided by the voxel size. To parallelise the touch detection the voxels are grouped into hypervoxels,containing 100 voxels each. The mouse dorsal striatum occupies about 26.2 mm (https://mouse.brain-map. org/), and contains almost 2 million neurons (Rosen & Williams, 2001). The first step is to identify which neurons belong to which hypervoxels. For a large portion of the neu- rons they will be present in more than one hypervoxel. This procedure is done in parallel where the worker nodes of the parallel computer get allocated a subset of the neurons and based on the vertex coordinates of the neurites calculate which hypervoxel the neurons are in. The results are gathered, creat- ing a list of neurons for each hypervoxel. The hypervoxels are then sorted based on the number of neurons inside, and those with most neurons are processed first for better load balance. To perform the touch detection, a line parsing algorithm takes small steps Δx, Δy, Δz along all line segments of the den- drites, marking the voxels they intersect. The voxels contained within the soma are also marked. It then repeats the procedure for the axon line segments of the morphologies. Voxels that We also include the functionality of virtual neurons, which contain both axons and dendrites are considered to have a are neurons that are not simulated, instead their activity is putative synapse if the two neuron types are allowed to have driven by a predefined spike train. This can be used to model, a connection between them (Fig. 3a, b). for example, the activation of reconstructed cortical axons in The purpose of the touch detection is to find the potential the striatum, which after touch detection will drive the neurons locations where neurons can connect to each other based on they connect to. the restrictions set by the morphologies. The result of the When using synapse density to place excitatory input onto above touch detection is a set of putative synapses, which is the neurons, larger neurons will receive more input than small- larger than the set of actual synapses. In the pruning step, the er neurons of the same type. However, size does not necessar- set of putative synapses is reduced to match the connectivity ily correlate with excitability of the neuron, or steepness of the statistics from experimental pairwise recordings. I-V curve which depends on intrinsic channels. To handle this In experiments it is common to report only the binning size variation of excitability Snudda allows the user to scale the and the number of connected neuron pairs, and the total num- number of synapses reaching a neuron, a process which in a ber of pairs. It would be beneficial for circuit modelling if the real network might be regulated by neuronal homeostatic distance for each pair was also recorded, we could then extract processes. distance dependent connectivity profiles and compare those to what the computer models predict. The pruning is divided into multiple steps, described above. The touch detection for a What Happens under the Hood? cubic millimeter can be run in a couple of hours on a desktop, and the whole striatum can be created in a couple of hours on a For each volume modelled the cell placement is restricted to supercomputer (Fig. 5). As an example, creating a striatal be inside the mesh specified. The neurons are placed one by network with 10,000 neurons (6.4 million synapses and one, with coordinates randomly sampled from a uniform dis- 1468 gap junctions) on a desktop Intel Xeon W-2133 CPU @ 3.60GHz with 6 cores and 64GB RAM took: init ~1 s, tribution. If a neuron position is inside the mesh, and there are 696 Neuroinform (2021) 19:685–701 Fig. 5 Snudda benchmarking on Tegner cluster at PDC/KTH. Each node with 500,000 neurons, around 469 million synapses and a hundred thou- has Intel E5-2690v3 Haswell with 2 × 12 cores and 512 GB RAM. a sand gap junctions. Place takes very little time compared to the other two Runtime on one node (24 CPU cores) for different network sizes. b phases and is barely visible at the bottom of the two graphs Runtime as a function of the number of CPUs when creating a network place 10 s, detect 8 min and prune 6 min. For 20,000 (50,000) et al., 2009; https://github.com/BlueBrain/NeuroM; https:// neurons the corresponding times are ~1 s (~1 s), 16 s (32 s), github.com/BlueBrain/NeuroR). Here we will illustrate 15 min (38 min), 12 min (36 min) to detect 14.3 million (39.9 typical use cases of manipulating morphological data on the million) synapses and 3462 (9395) gap junctions. example of a small Python module treem (https://github.com/ In addition to JSON configuration files, the resulting net- a1eko/treem), developed by the authors in conjunction with work data is stored in HDF5 files. Snudda as a complementary instrument to above mentioned packages. Module Treem provides data structure and command-line tools for accessing and manipulating the digital reconstruc- The Challenge of Limited Morphology Data tions of the neuron morphology in Stockley-Wheal-Cannon format, SWC (Cannon et al., 1998). Access to morphological A big challenge of the biologically detailed anatomically data from the source code is supported by several Python constrained simulations of the neural microcircuits is avail- classes. Common operations with SWC files are possible from ability of the high quality morphological reconstructions of the user-written scripts or via the command-line tool swc. For the main neuron types, in sufficient numbers and variability. the detailed description of the user interface, see API and CLI Open public morphometric repositories, similar to ModelDB references in the online documentation (https://treem. for models present in the world-wide web since 1996 readthedocs.io). (McDougal et al., 2017), were pioneered in 2006 by G. A common reconstruction error is so called “z-jump” Ascoli with NeuroMorpho.Org (Akram et al., 2018; http:// (Brown et al., 2011) when a part of the neurite gets shifted neuromorpho.org/) and continued by other research centers along the z-axis by a few micrometers as shown in Fig. 6a (top like Allen Brain Institute (Jones et al., 2009; https://portal. panel). These can result from an accumulated error during the brain-map.org), Janelia Research Campus (Gerfen et al., manual reconstruction or as a mistake in automatic procedure. 2018; Economo et al., 2019; http://mouselight.janelia.org/), Possible z-jumps can be eliminated in Treem by the repair eBRAINS Knowledge Graph (https://kg.ebrains.eu/), to command using one of the four methods, align, split, tilt or name a few, have become increasingly popular among join, as illustrated in Fig. 6a. Choice of the repair method as computational neuroscientists. well as the assessment of the result should ideally be left to the Single-cell morphological reconstructions vary in quality author of the reconstructed data; if this is not possible the due to the difference in experimental procedures leading to preference is given to the method which better preserves cell varying degrees of physical integrity of the neurites, spatial symmetry. resolution, tissue shrinkage, slicing, etc. Need for consistent Since neuronal tissue can shrink due to dehydration during quality assurance of morphological reconstructions facilitated histological preparation, correction factors are required before development of morphology processing tools for morphomet- a reconstructed morphology can enter the simulation pipeline. ric measurements, data processing and error correction, such Shrinkage correction involves scaling of the entire reconstruc- as L-measure (Scorcioni et al., 2008; http://cng.gmu.edu: tion in (x, y)-plane, expansion in z-direction, as well as de- 8080/Lm), TREES toolbox (Cuntz et al., 2010; https://www. creasing contraction of selected neurites, e.g. dendrites, by treestoolbox.org/), btmorph (Torben-Nielsen, 2014; https:// stretching along their principal axes (termed “unravelling” in bitbucket.org/btorb/btmorph) and NeuroM/NeuroR (Anwar Neuroinform (2021) 19:685–701 697 Fig. 6 Repairing digital reconstruction of the neuron morphology. a (align, split, tilt and join). b Repairing the dendrites cut at the slice border. Correcting “z-jump” reconstruction errors (top panel). Dots illustrate Orange dots in a 3D plot show the cut points of the dendrites. Red lines in reconstructed points, soma is black, dendrites are blue, the orange dot 2D projections show “repaired” dendrites, i.e. extended neurites using labels the node at the point of presumed discontinuity. Four correction undamaged reconstructions of the same topological order as the cut methods implemented in Treem (Python module treem) are shown below branches Markram et al., 2015) or length-preserving spatial filtering spread of morpho-electric characteristics in live neurons. To (not shown, see online documentation). enforce variability in the simulation based on the limited num- Another important omission is that the neurons located close ber of reconstructed neurons, we apply random manipulations to the slice surface often have their neurites cut and thus miss- to the morphological reconstructions. Examples of the length- ing in the digital reconstruction. Cut neurites can be replaced preserving modifications implemented in Treem are shown in using the intact branches of the same topological order from the Fig. 7a. Methods jitter, twist and rotate do not change the inner part of the slice, assuming spherical or axial symmetry of length of the dendritic branches and thus do not affect electro- the neuron morphology as shown in Fig. 6b. In Treem this is physiological features of the optimized models but help to achieved with the repair command (see online documentation recover spatial symmetry of the morphological reconstruc- for the example commands to reproduce Fig. 6b). tions as shown in Fig. 7b. To distribute excitability of the One of the aspects of the large-scale simulations is realistic single-cell models, digital reconstructions can be scaled ran- variability of the model parameters mimicking the natural domly in 3D, as was done in the large-scale simulation by Fig. 7 Adding variability to the morphological reconstructions. a repaired reconstructions (n =3) and “jittered”, i.e. randomly manipulated Examples of modification methods used in Treem (modifications reconstructions, nine variants per each cell (n = 27). Here, only random preserving the total length are shown). b Distribution of the coordinates twisting of the dendritic branches at the bifurcation points was applied of the dendritic terminations of the fast-spiking interneurons at different which proved to be sufficient to restore the symmetry stages of morphology processing - original reconstructions (n =3), 698 Neuroinform (2021) 19:685–701 Hjorth et al. (2020) and illustrated in the online documentation above. Data on electrophysiological recordings published of Treem (https://treem.readthedocs.io). might be incomplete in such a way that only a few selected traces, as shown in the published manuscripts, exist. Also transcriptional, electrophysiological and morphological data Discussion might come from different experiments. Ideally, however, it would be best to have recordings from the neurons that were In this paper we show how to use our open source modelling morphologically reconstructed, such as in patch-seq technique pipeline to build microcircuit models. An important goal is (e.g. Fuzik et al., 2016). Although the electrophysiological that the model building process should be transparent and properties are well studied, one might lack the knowledge of possible to reproduce by other labs, and the model should be which ion conductances are expressed in the neurons. extendable when new data accumulate. The pipeline is devel- Fortunately, such data are starting to emerge, and for example, oped for setting up large-scale simulations of subcortical nu- for striatum transcriptomics data already exist (Muñoz- clei, such as striatum. In our current pipeline, based on the Manchado et al., 2018;Ho et al., 2018; Gokce et al., 2016; software Snudda, neurons can be placed in a defined volume, Saunders et al., 2018). If one has a good hypothesis of which and then prediction of the location of synapses (as well as gap channels are expressed, characterisation as well as models are junctions) can be made using neuron morphologies and the starting to be collected at resources such as the Channelpedia specified pruning rules. Also synaptic data for short-term syn- (Ranjan et al., 2011; http://channelpedia.net) and Ion Channel aptic plasticity or failure rates can be represented. Finally a Genealogy (Podlaski et al., 2017; https://icg.neurotheory.ox. simulation using the NEURON simulation environment can ac.uk/). Although still not trivial, if one has both the be launched. A challenge when using cellular level data from morphology and electrophysiological data of a particular public databases is that sometimes the data for the reconstruct- neuron type, workflows have already been developed for ed neuronal morphologies only include soma and dendrites, optimizing neuron models (Van Geit et al., 2016; Migliore missing the axon entirely. Therefore Snudda supports the pre- et al., 2018;Masoli etal., 2020). diction of synapses using different approaches. If the detailed A natural future goal would, however, be to link microcir- morphology is available, the reconstructed axons and den- cuit models built in different labs, e.g. a cortical microcircuit drites can be used to constrain which neurons are within reach connected to a striatal microcircuit. Then interoperability be- of one another. If the axon is missing, the user can instead tween models as well as model specification, such as specify an axonal density which is then used for the synapse SONATA, would be crucial. We have on our road map for detection. Also our pipeline provides the opportunity to ‘re- Snudda to support the SONATA standard (Dai et al., 2020) pair’ the dendritic morphologies, and we have illustrated ways and work has already started on it to become interoperable with the EBRAINS infrastructure (https://ebrains.eu/). to do this using the software Treem. In addition, Treem can provide jittering of morphological parameters to increase the variability in the modelled population of neurons, which is useful to avoid artefacts when there are too few available Information Sharing Statement morphologies for each neuron type. The current version of Snudda does not treat spines separately from the rest of the The presented software Snudda (version 1.1; dendrites, a future improvement would be to allow the user to RRID:SCR_021210) and Treem (version 1.0.0, DOI:https:// specify requirements to target spines separately, e.g. if spines doi.org/10.5281/zenodo.4890845) are available on GitHub are already specified in the reconstruction data. and PyPI: Models of neocortical microcircuits have been built with Snudda - https://github.com/Hjorthmedh/Snudda, https:// similar approaches (see above), however, not all elements of pypi.org/project/snudda/ their workflow were available or open-source at the time of Treem - https://github.com/a1eko/Treem, https://pypi.org/ Snudda development. We believe that our open source pipe- project/Treem/ line might become useful when building biophysically de- tailed microcircuit models of other subcortical brain regions, Acknowledgements The simulations were performed on resources pro- such as the other basal ganglia nuclei. vided by the Swedish National Infrastructure for Computing at PDC When using the current modelling workflow, it is assumed (Center for Parallel Computing). We acknowledge the use of Fenix that one has a collection of quantitatively detailed neuron Research Infrastructure resources, which are partially funded from the models for each neuron type to be used in the modelled mi- European Union’s Horizon 2020 research and innovation programme through the ICEI project under the grant agreement No. 800858. The crocircuit. Such neuron models might come from public data- authors wish to thank Sten Grillner, Johanna Frost-Nylén, Robert bases (see above). But most likely several of the neuron types Lindroos and Ilaria Carannante for helpful discussions. We also thank in the selected microcircuit to be modelled might have to be Robin de Schepper, Kadri Pajo, and Wilhelm Thunberg for help with built from scratch. Here the challenges are several as described software compatibility. Neuroinform (2021) 19:685–701 699 Code Availability (see Information Sharing Statement) Mihalas, S., & Arkhipov, A. (2020). Systematic integration of struc- tural and functional data into multi-scale models of mouse primary visual cortex. Neuron, 106(3), 388–403.e18. https://doi.org/10. Funding Open access funding provided by Royal Institute of 1016/j.neuron.2020.01.040. Technology. Horizon 2020 Framework Programme (785907, HBP Brown, K. M., Barrionuevo, G., Canty, A. J., De Paola, V., Hirsch, J. A., SGA2); Horizon 2020 Framework Programme (945539, HBP SGA3); Jefferis, G. S., Lu, J., Snippe, M., Sugihara, I., & Ascoli, G. A. Vetenskapsrådet (VR-M-2017-02806, VR-M-2020-01652); Swedish e- (2011). The DIADEM data sets: Representative light microscopy science Research Center (SeRC); KTH Digital Futures. images of neuronal morphology to advance automation of digital reconstructions. Neuroinformatics, 9(2–3), 143–157. https://doi.org/ Data Availability Data used for single-cell neuron models as well as 10.1007/s12021-010-9095-5. synaptic connectivity within the example network of the striatal micro- Cannon, R. C., Turner, D. A., Pyapali, G. K., & Wheal, H. V. (1998). An circuit is available on GitHub (in the snudda/data and examples folders at on-line archive of reconstructed hippocampal neurons. Journal of https://github.com/hjorthmedh/Snudda). Neuroscience Methods, 84(1–2), 49–54. https://doi.org/10.1016/ s0165-0270(98)00091-0. Carnevale, T., & Hines, M. (2006). The NEURON book (p. 2006). Declarations Cambridge University Press. https://doi.org/10.1017/ CBO9780511541612. Conflicts of interest/Competing Interests The authors have no relevant Casali, S., Marenzi, E., Medini, C., Casellato, C., & D’Angelo, E. (2019). financial or non-financial interests to disclose. Reconstruction and simulation of a scaffold model of the cerebellar network. Frontiers in Neuroinformatics, 13,37. https://doi.org/10. Open Access This article is licensed under a Creative Commons 3389/fninf.2019.00037. Attribution 4.0 International License, which permits use, sharing, adap- Cizeron, M., Qiu, Z., Koniaris, B., Gokhale, R., Komiyama, N. H., tation, distribution and reproduction in any medium or format, as long as Fransén, E., & Grant, S. G. N. (2020). A brainwide atlas of synapses you give appropriate credit to the original author(s) and the source, pro- across the mouse life span. Science., 369(6501), 270–275. https:// vide a link to the Creative Commons licence, and indicate if changes were doi.org/10.1126/science.aba3163. made. The images or other third party material in this article are included Colangelo, C., Shichkova, P., Keller, D., Markram, H., & Ramaswamy, in the article's Creative Commons licence, unless indicated otherwise in a S. (2019). Cellular, synaptic and network effects of acetylcholine in credit line to the material. If material is not included in the article's the neocortex. Frontiers in Neural Circuits, 13,24. https://doi.org/ Creative Commons licence and your intended use is not permitted by 10.3389/fncir.2019.00024. statutory regulation or exceeds the permitted use, you will need to obtain Cuntz, H., Forstner, F., Borst, A., & Häusser, M. (2010). One rule to grow permission directly from the copyright holder. To view a copy of this them all: A general theory of neuronal branching and its practical licence, visit http://creativecommons.org/licenses/by/4.0/. application. PLoS Computational Biology, 6(8), e1000877. https:// doi.org/10.1371/journal.pcbi.1000877. Dai, K., Hernando, J., Billeh, Y. N., Gratiy, S. L., Planas, J., Davison, A. P., Dura-Bernal, S., Gleeson, P., Devresse, A., Dichter, B. K., Gevaert, M., King, J. G., van Geit, W. A. H., Povolotsky, A. V., References Muller, E., Courcol, J. D., & Arkhipov, A. (2020). The SONATA data format for efficient description of large-scale network models. Akar, N. A. et al. (2019). Arbor — A morphologically-detailed neural PLoS Computational Biology, 16(2), e1007696. https://doi.org/10. network simulation library for contemporary high-performance 1371/journal.pcbi.1007696. computing architectures. 2019 27th Euromicro International Economo, M. N., Winnubst, J., Bas, E., Ferreira, T. A., & Chandrashekar, Conference on Parallel, Distributed and Network-Based J. (2019). Single-neuron axonal reconstruction: The search for a Processing (PDP), Pavia, Italy, 2019, pp. 274–282. https://doi.org/ wiring diagram of the brain. The Journal of Comparative 10.1109/EMPDP.2019.8671560. Neurology, 527(13), 2190–2199. https://doi.org/10.1002/cne. Akram, M., Nanda, S., Maraver, P., Armañanzas, R., & Ascoli, G. A. (2018). An open repository for single-cell reconstructions of the Einevoll, G. T., Destexhe, A., Diesmann, M., Grün, S., Jirsa, V., de brain forest. Sci Data, 5, 180006. https://doi.org/10.1038/sdata. Kamps,M.,Migliore, M., Ness,T.V.,Plesser,H.E.,& 2018.6. Schürmann, F. (2019). The scientific case for brain simulations. Amunts, K., Knoll, A. C., Lippert, T., Pennartz, C. M. A., Ryvlin, P., Neuron., 102(4), 735–744. https://doi.org/10.1016/j.neuron.2019. Destexhe, A., Jirsa, V. K., D’Angelo, E., & Bjaalie, J. G. (2019). 03.027. The human brain project-synergy between neuroscience, comput- Fuzik, J., Zeisel, A., Máté, Z., Calvigioni, D., Yanagawa, Y., Szabó, G., ing, informatics, and brain-inspired technologies. PLoS Biology, Linnarsson, S., & Harkany, T. (2016). Integration of electrophysio- 17(7), e3000344. https://doi.org/10.1371/journal.pbio.3000344. logical recordings with single-cell RNA-seq data identifies neuronal Anwar, H., Riachi, I., Schürmann, F., & Markram H. (2009). An ap- subtypes. Nature Biotechnology, 34(2), 175–183. https://doi.org/10. proach to capturing neuron morphological diversity. In 1038/nbt.3443. Computational neuroscience: Realistic modeling for Gerfen, C. R., Economo, M. N., & Chandrashekar, J. (2018). Long dis- experimentalists. De Schutter E., editor. (Cambridge: The MIT tance projections of cortical pyramidal neurons. Journal of Press) 211–232. https://doi.org/10.7551/mitpress/9780262013277. Neuroscience Research, 96(9), 1467–1475. https://doi.org/10. 003.0010. 1002/jnr.23978. Berthet, P., Lindahl, M., Tully, P. J., Hellgren-Kotaleski, J., & Lansner, Gleeson, P., Crook, S., Cannon, R. C., Hines, M. L., Billings, G. O., A. (2016). Functional relevance of different basal ganglia pathways Farinella, M., Morse, T. M., Davison, A. P., Ray, S., Bhalla, U. S., investigated in a spiking model with reward dependent plasticity. Barnes, S. R., Dimitrova, Y. D., & Silver, R. A. (2010). NeuroML: Frontiers in Neural Circuits., 10,53. https://doi.org/10.3389/fncir. A language for describing data driven models of neurons and net- 2016.00053. works with a high degree of biological detail. PLoS Computational Biology, 6(6), e1000815. https://doi.org/10.1371/journal.pcbi. Billeh, Y. N., Cai, B., Gratiy, S. L., Dai, K., Iyer, R., Gouwens, N. W., Abbasi-Asl,R.,Jia,X.,Siegle,J.H.,Olsen,S.R.,Koch,C., 1000815. 700 Neuroinform (2021) 19:685–701 Gokce, O., Stanley, G. M., Treutlein, B., Neff, N. F., Camp, J. G., Kanari, L., Ramaswamy, S., Shi, Y., Morand, S., Meystre, J., Perin, R., Abdellah, M., Wang, Y., Hess, K., & Markram, H. (2019). Malenka, R. C., Rothwell, P. E., Fuccillo, M. V., Sudhof, T. C., & Quake, S. R. (2016). Cellular taxonomy of the mouse striatum as Objective morphological classification of neocortical pyramidal revealed by single-cell RNA-Seq. Cell Reports, 16(4), 1126–1137. cells. Cerebral Cortex, 29(4), 1719–1735. https://doi.org/10.1093/ https://doi.org/10.1016/j.celrep.2016.06.059. cercor/bhy339. Gratiy, S. L., Billeh, Y. N., Dai, K., Mitelut, C., Feng, D., Gouwens, N. Kumbhar, P., Hines, M., Fouriaux, J., Ovcharenko, A., King, J., Delalondre, F., & Schürmann, F. (2019). CoreNEURON : An opti- W., Cain, N., Koch, C., Anastassiou, C. A., & Arkhipov, A. (2018). BioNet: A Python interface to NEURON for modeling large-scale mized compute engine for the NEURON simulator. Frontiers in networks. PLoS ONE, 13(8), e0201630. https://doi.org/10.1371/ Neuroinformatics, 13, 63. https://doi.org/10.3389/fninf.2019. journal.pone.0201630. 00063. Grillner, S., Ip, N., Koch, C., Koroshetz, W., Okano, H., Polachek, M., Lindahl, M., & Hellgren Kotaleski, J. (2017). Untangling basal ganglia Poo, M. M., & Sejnowski, T. J. (2016). Worldwide initiatives to network dynamics and function: role of dopamine depletion and advance brain research. Nature Neuroscience, 19(9), 1118–1122. inhibition investigated in a spiking network model. eNeuro, 3(6), https://doi.org/10.1038/nn.4371. ENEURO.0156-16.2016. https://doi.org/10.1523/ENEURO.0156- Gurney, K., Prescott, T. J., & Redgrave, P. (2001). A computational 16.2016. model of action selection in the basal ganglia. II Analysis and sim- Markram, H., Muller, E., Ramaswamy, S., Reimann, M. W., Abdellah, ulation of behaviour. Biological Cybernetics, 84(6), 411–423. M., Sanchez, C. A., Ailamaki, A., Alonso-Nanclares, L., Antille, N., https://doi.org/10.1007/PL00007985. Arsever, S., Kahou, G. A. A., Berger, T. K., Bilgili, A., Buncic, N., Hellwig, B. (2000). A quantitative analysis of the local connectivity be- Chalimourda, A., Chindemi, G., Courcol, J. D., Delalondre, F., tween pyramidal neurons in layers 2/3 of the rat visual cortex. Delattre, V., Druckmann, S., Dumusc, R., Dynes, J., Eilemann, S., Biological Cybernetics, 82(2), 111–121. https://doi.org/10.1007/ Gal, E., Gevaert, M. E., Ghobril, J. P., Gidon, A., Graham, J. W., pl00007964. Gupta, A., Haenel, V., Hay, E., Heinis, T., Hernando, J. B., Hines, Hepburn, I., Chen, W., Wils, S., & De Schutter, E. (2012). STEPS: M., Kanari, L., Keller, D., Kenyon, J., Khazen, G., Kim, Y., King, J. Efficient simulation of stochastic reaction-diffusion models in real- G., Kisvarday, Z., Kumbhar, P., Lasserre, S., le Bé, J. V., istic morphologies. BMC Systems Biology, 6,36. https://doi.org/10. Magalhães, B. R. C., Merchán-Pérez, A., Meystre, J., Morrice, B. 1186/1752-0509-6-36. R., Muller, J., Muñoz-Céspedes, A., Muralidhar, S., Muthurasa, K., Hines, M. L., Davison, A. P., & Muller, E. (2009). NEURON and Nachbaur, D., Newton, T. H., Nolte, M., Ovcharenko, A., Palacios, Python. Frontiers in Neuroinformatics, 3,1. https://doi.org/10. J., Pastor, L., Perin, R., Ranjan, R., Riachi, I., Rodríguez, J. R., Riquelme, J. L., Rössert, C., Sfyrakis, K., Shi, Y., Shillcock, J. C., 3389/neuro.11.001.2009. Hjorth, J., Blackwell, K. T., & Kotaleski, J. H. (2009). Gap junctions Silberberg, G., Silva, R., Tauheed, F., Telefont, M., Toledo- Rodriguez, M., Tränkler, T., van Geit, W., Díaz, J. V., Walker, R., between striatal fast-spiking interneurons regulate spiking activity Wang, Y., Zaninetta, S. M., DeFelipe, J., Hill, S. L., Segev, I., & and synchronization as a function of cortical activity. The Journal of Schürmann, F. (2015). Reconstruction and simulation of neocortical Neuroscience, 29(16), 5276–5286. https://doi.org/10.1523/ microcircuitry. Cell, 163(2), 456–492. https://doi.org/10.1016/j.cell. jneurosci.6031-08.2009. 2015.09.029. Hjorth, J. J. J., Kozlov, A., Carannante, I., Frost Nylén, J., Lindroos, R., Johansson, Y., Tokarska, A., Dorst, M. C., Suryanarayana, S. M., Masoli, S., Tognolina, M., Laforenza, U., Moccia, F., & D'Angelo, E. Silberberg, G., Hellgren Kotaleski, J., & Grillner, S. (2020). The (2020). Parameter tuning differentiates granule cell subtypes microcircuits of striatum in silico. Proceedings of the National enriching transmission properties at the cerebellum input stage. Academy of Sciences of the United States of America, 117(17), Communications Biology, 3(1), 222. https://doi.org/10.1038/ 9554–9565. https://doi.org/10.1073/pnas.2000671117. s42003-020-0953-x. Ho, H., Both, M. D., Siniard, A., Sharma, S., Notwell, J. H., Wallace, M., McDougal, R. A., Morse, T. M., Carnevale, T., Marenco, L., Wang, R., Leone, D. P., Nguyen, A., Zhao, E., Lee, H., Zwilling, D., Migliore, M., Miller, P. L., Shepherd, G. M., & Hines, M. L. (2017). Thompson,K.R.,Braithwaite,S.P.,Huentelman,M., & Twenty years of ModelDB and beyond: Building essential modeling Portmann, T. (2018). A guide to single-cell transcriptomics in adult tools for the future of neuroscience. Journal of Computational rodent brain: The medium spiny neuron transcriptome revisited. Neuroscience, 42(1), 1–10. https://doi.org/10.1007/s10827-016- Frontiers in Cellular Neuroscience, 12, 159. https://doi.org/10. 0623-7. 3389/fncel.2018.00159. Migliore, R., Lupascu, C. A., Bologna, L. L., Romani, A., Courcol, J.-D., Humphries, M. D., Wood, R., & Gurney, K. (2009). Dopamine- Antonel, S., van Geit, W. A. H., Thomson, A. M., Mercer, A., modulated dynamic cell assemblies generated by the GABAergic Lange, S., Falck, J., Rössert, C. A., Shi, Y., Hagens, O., Pezzoli, striatal microcircuit. Neural Networks, 22(8), 1174–1188. https:// M., Freund, T. F., Kali, S., Muller, E. B., Schürmann, F., Markram, doi.org/10.1016/j.neunet.2009.07.018. H., & Migliore, M. (2018). The physiological variability of channel Insel, T. R., Landis, S. C., & Collins, F. S. (2013). Research priorities. density in hippocampal CA1 pyramidal cells and interneurons ex- The NIH BRAIN initiative. Science., 340(6133), 687–688. https:// plored using a unified data-driven modeling workflow. PLoS doi.org/10.1126/science.1239276. Computational Biology, 14(9), e1006423. https://doi.org/10.1371/ Jones, A. R., Overly, C. C., & Sunkin, S. M. (2009). The Allen brain journal.pcbi.1006423. atlas: 5 years and beyond. Nature Reviews. Neuroscience, 10(11), Muñoz-Manchado, A. B., Bengtsson Gonzales, C., Zeisel, A., Munguba, 821–828. https://doi.org/10.1038/nrn2722. H., Bekkouche, B., Skene, N. G., Lönnerberg, P., Ryge, J., Harris, Jordan, J., Helias, M., Diesmann, M., & Kunkel, S. (2020). Efficient K. D., Linnarsson, S., & Hjerling-Leffler, J. (2018). Diversity of communication in distributed simulations of spiking neuronal net- interneurons in the dorsal striatum revealed by single-cell RNA se- works with gap junctions. Frontiers in Neuroinformatics, 14,12. quencing and PatchSeq. Cell Rep, 24(8), 2179–2190.e7. https://doi. https://doi.org/10.3389/fninf.2020.00012. org/10.1016/j.celrep.2018.07.053. Neuroinform (2021) 19:685–701 701 Okano, H., Miyawaki, A., & Kasai, K. (2015). Brain/MINDS: Brain- Sudhakar, S. K., Hong, S., Raikov, I., Publio, R., Lang, C., Close, T., Guo, D., Negrello, M., & De Schutter, E. (2017). Spatiotemporal mapping project in Japan. Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences, 370(1668), network coding of physiological mossy fiber inputs by the cerebellar 20140310. https://doi.org/10.1098/rstb.2014.0310. granular layer. PLoS Computational Biology, 13, e1005754. https:// Plesser, H., Diesmann, M., Gewaltig, M., & Morrison, A. (2015). Nest: doi.org/10.1371/journal.pcbi.1005754. The neural simulation tool. In D. Jaeger & R. Jung (Eds.), Torben-Nielsen, B. (2014). An efficient and extendable Python library to Encyclopedia of computational neuroscience (pp. 1849–1852). analyze neuronal morphologies. Neuroinformatics, 12(4), 619–622. Springer New York. https://doi.org/10.1007/s12021-014-9232-7. Podlaski, W. F., Seeholzer, A., Groschner, L. N., Miesenböck, G., Van Geit, W., Gevaert, M., Chindemi, G., Rössert, C., Courcol, J., Ranjan, R., & Vogels, T. P. (2017). Mapping the function of neuro- Muller, E. B., Schürmann, F., Segev, I., & Markram, H. (2016). nal ion channels in model and experiment. Elife., 6, e22152. https:// BluePyOpt: Leveraging open source software and cloud infrastruc- doi.org/10.7554/elife.22152. ture to optimise model parameters in neuroscience. Frontiers in Ranjan, R., Khazen, G., Gambazzi, L., Ramaswamy, S., Hill, S. L., Neuroinformatics, 10, 17. https://doi.org/10.3389/fninf.2016. Schürmann, F., & Markram, H. (2011). Channelpedia: An integra- tive and interactive database for ion channels. Frontiers in Wichert, I., Jee, S., De Schutter, E., & Hong, S. (2020). Pycabnn: Neuroinformatics, 5,36. https://doi.org/10.3389/fninf.2011.00036. Efficient and extensible software to construct an anatomical basis Ray, S., & Bhalla, U. S. (2008). PyMOOSE: Interoperable scripting in for a physiologically realistic neural network model. Frontiers in Python for MOOSE. Frontiers in Neuroinformatics, 2,6. https://doi. Neuroinformatics, 14, 31. https://doi.org/10.3389/fninf.2020. org/10.3389/neuro.11.006.2008. Reimann, M.W.,King,J. G.,Muller, E.B.,Ramaswamy,S.,& Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J., Appleton, G., Markram, H. (2015). An algorithm to predict the connectome of Axton,M.,Baak,A.,Blomberg,N.,Boiten,J.W.,daSilva neural microcircuits. Frontiers in Computational Neuroscience, 9, Santos, L. B., Bourne, P. E., Bouwman, J., Brookes, A. J., Clark, 120. https://doi.org/10.3389/fncom.2015.00120. T., Crosas, M., Dillo, I., Dumon, O., Edmunds, S., Evelo, C. T., Rosen, G. D., & Williams, R. W. (2001). Complex trait analysis of the Finkers,R.,Gonzalez-Beltran,A.,Gray,A. J.G.,Groth,P., mouse striatum: Independent QTLs modulate volume and neuron Goble, C., Grethe, J. S., Heringa, J., ’t Hoen, P. A. C., Hooft, R., number. BMC Neuroscience, 2,5. https://doi.org/10.1186/1471- Kuhn, T., Kok, R., Kok, J., Lusher, S. J., Martone, M. E., Mons, A., 2202-2-5. Packer, A. L., Persson, B., Rocca-Serra, P., Roos, M., van Schaik, Santuy, A., Tomás-Roca, L., Rodríguez, J. R., González-Soriano, J., Zhu, R., Sansone, S. A., Schultes, E., Sengstag, T., Slater, T., Strawn, G., F., Qiu, Z., Grant, S. G. N., DeFelipe, J., & Merchan-Perez, A. Swertz, M. A., Thompson, M., van der Lei, J., van Mulligen, E., (2020). Estimation of the number of synapses in the hippocampus Velterop, J., Waagmeester, A., Wittenburg, P., Wolstencroft, K., and brain-wide by volume electron microscopy and genetic labeling. Zhao, J., & Mons, B. (2016). The FAIR Guiding Principles for Scientific Reports, 10, 14014. https://doi.org/10.1038/s41598-020- scientific data management and stewardship [published correction 70859-5. appears in Sci Data. 2019 Mar 19;6(1):6]. Sci Data, 3, 160018. Saunders, A., Macosko, E. Z., Wysoker, A., Goldman, M., Krienen, F. https://doi.org/10.1038/sdata.2016.18. M., de Rivera, H., Bien, E., Baum, M., Bortolin, L., Wang, S., Williams, T. J., Balakrishnan, R., Delalondre, F., Schuermann, F., Muller, Goeva, A., Nemesh, J., Kamitaki, N., Brumbaugh, S., Kulp, D., & E., &Gewaltig, M. O. (2018). Large-Scale Simulation of Brain McCarroll, S. A. (2018). Molecular diversity and specializations Tissue, Blue Brain Project, EPFL. United States: N. p., 2018. among the cells of the adult mouse brain. Cell, 174(4), 1015– Web. https://doi.org/10.2172/1483995. 1030.e16. https://doi.org/10.1016/j.cell.2018.07.028. Scorcioni, R., Polavaram, S., & Ascoli, G. (2008). L-measure: A web- Publisher’sNote Springer Nature remains neutral with regard to jurisdic- accessible tool for the analysis, comparison and search of digital tional claims in published maps and institutional affiliations. reconstructions of neuronal morphologies. Nature Protocols, 3, 866–876. https://doi.org/10.1038/nprot.2008.51.
Neuroinformatics – Springer Journals
Published: Oct 1, 2021
Keywords: Large-scale simulations; Striatum; Basal ganglia; Brain microcircuits; Synaptic connectivity; Model building pipeline
You can share this free article with as many people as you like with the url below! We hope you enjoy this feature!
Read and print from thousands of top scholarly journals.
Already have an account? Log in
Bookmark this article. You can see your Bookmarks on your DeepDyve Library.
To save an article, log in first, or sign up for a DeepDyve account if you don’t already have one.
Copy and paste the desired citation format or use the link below to download a file formatted for EndNote
Access the full text.
Sign up today, get DeepDyve free for 14 days.
All DeepDyve websites use cookies to improve your online experience. They were placed on your computer when you launched this website. You can change your cookie settings through your browser.