Abstract
The basal ganglia of the laboratory rat consist of a few core regions that are specifically interconnected by efferents and afferents of the central nervous system. In nearly 800 reports of tract-tracing investigations the connectivity of the basal ganglia is documented. The readout of connectivity data and the collation of all the connections of these reports in a database allows to generate a connectome. The collation, curation and analysis of such a huge amount of connectivity data is a great challenge and has not been performed before (Bohland et al. PloS One 4:e7200, 2009) in large connectomics projects based on meta-analysis of tract-tracing studies. Here, the basal ganglia connectome of the rat has been generated and analyzed using the consistent cross-platform and generic framework neuroVIISAS. Several advances of this connectome meta-study have been made: the collation of laterality data, the network-analysis of connectivity strengths and the assignment of regions to a hierarchically organized terminology. The basal ganglia connectome offers differences in contralateral connectivity of motoric regions in contrast to other regions. A modularity analysis of the weighted and directed connectome produced a specific grouping of regions. This result indicates a correlation of structural and functional subsystems. As a new finding, significant reciprocal connections of specific network motifs in this connectome were detected. All three principal basal ganglia pathways (direct, indirect, hyperdirect) could be determined in the connectome. By identifying these pathways it was found that there exist many further equivalent pathways possessing the same length and mean connectivity weight as the principal pathways. Based on the connectome data it is unknown why an excitation pattern may prefer principal rather than other equivalent pathways. In addition to these new findings the local graph-theoretical features of regions of the connectome have been determined. By performing graph theoretical analyses it turns out that beside the caudate putamen further regions like the mesencephalic reticular formation, amygdaloid complex and ventral tegmental area are important nodes in the basal ganglia connectome. The connectome data of this meta-study of tract-tracing reports of the basal ganglia are available for further network studies, the integration into neocortical connectomes and further extensive investigations of the basal ganglia dynamics in population simulations.
Similar content being viewed by others
Abbreviations
- A:
-
All (all inputs and outputs)
- AD:
-
Average degree
- Ac:
-
Accumbens nucleus
- AC:
-
Amygdaloid complex
- AGl:
-
Lateral agranular prefrontal cortex
- AGm:
-
Medial agranular prefrontal cortex
- Aut:
-
Authoritativeness
- AW:
-
Average weight
- BC:
-
Betweenness centrality
- BG:
-
Basal ganglia
- C:
-
Circle
- CC:
-
Cluster-coefficient
- CE:
-
Closeness centrality
- chain:
-
Chain pattern of a motif
- CL:
-
Centrolateral thalamic nucleus
- CM:
-
Central medial thalamic nucleus
- CNS:
-
Central nervous system
- CPu:
-
Caudate putamen
- DG:
-
Degree
- Dic:
-
Direct input from contralateral
- Dii:
-
Direct input from ipsilateral
- Dis:
-
Direct input from ipsi- and contralateral
- DNN:
-
Direct neighbor network
- Doc:
-
Direct output to contralateral
- Doi:
-
Direct output to ipsilateral
- Dos:
-
Direct output to ipsi- and contralateral
- DR:
-
Dorsal raphe nucleus
- EC:
-
Eigenvector centrality
- Ent:
-
Entorhinal cortex
- HIPP:
-
Hippocampus
- Hub:
-
Hubness
- I:
-
In (input to a region; used in tables)
- in:
-
Symmetric input connection to a central node of a motif
- INN:
-
Indirect neighbor network
- L:
-
Laterality
- LGP:
-
Lateral globus pallidus
- LHb:
-
Lateral habenular nucleus
- MDL:
-
Mediodorsal thalamic nucleus lateral part
- MDM:
-
Mediodorsal thalamic nucleus medial part
- MDS:
-
Metric multidimensional scaling
- MGP:
-
Medial globus pallidus
- MRF:
-
Mesencephalic reticular formation
- O:
-
Out (Output of region; used in tables only)
- out:
-
Symmetric output connection from a central node of a motif
- PC:
-
Paracentral thalamic nucleus
- PCA:
-
Principal component analysis
- PF:
-
Parafascicular thalamic nucleus
- Pir:
-
Piriform cortex
- PL:
-
Path length
- PRC:
-
Page rank centrality
- Pub:
-
Number of articles
- RADin:
-
Radiality of the input
- RADout:
-
Radiality of the output
- Rec:
-
Reciprocal
- Rel:
-
Reliability
- Sic:
-
Subtree input from contralateral
- Sii:
-
Subtree input from ipsilateral
- Sis:
-
Subtree input from ipsi- and contralateral
- SG:
-
Subgraph centrality
- SNC:
-
Substantia nigra compact part
- SNR:
-
Substantia nigra reticular part
- Soc:
-
Subtree output to contralateral
- Soi:
-
Subtree output to ipsilateral
- Sos:
-
Subtree output to ipsi- and contralateral
- SP:
-
Length of shortest path
- SPN:
-
Spiny neurons of the CPu
- STh:
-
Subthalamic nucleus
- VA:
-
Ventro anterior thalamic nucleus
- VL:
-
Ventrolateral thalamic nucleus
- VM:
-
Ventromedial thalamic nucleus
- VTA:
-
Ventral tegmental area A10
References
Albert R, Barabasi AL (2002) Statistical mechanics of complex networks. Rev Mod Phys 74:47–97
Albin RL, Young AB, Penney JB (1989) The functional anatomy of basal ganglia disorders. Trends Neurosci 12:366–375
Alexander GE, Crutcher MD (1990) Functional architecture of basal ganglia circuits: neural substrates of parallel processing. Trends Neurosci 13:266–271
Archambault D, Munzner T, Auber D (2008) Grouseflocks: steerable exploration of graph hierarchy space. IEEE Trans Vis Comput Graph 14:900–913
Bai X, Yu L, Liu Q, Zhang J, Li A, Han D, Luo Q, Gong H (2006) A high-resolution anatomical rat atlas. J Anat 209:707–708
Bakker R, Wachtler T, Diesmann M (2012) CoCoMac 2.0 and the future of tract-tracing databases. Front Neuroinform 6:1–6
Barbas H, Hilgetag CC, Saha S, Dermon CR, Suski JL (2005) Parallel organization of contralateral and ipsilateral prefrontal cortical projections in the rhesus monkey. BMC Neurosci 6:32
Bassett DS, Bullmore E (2006) Small-world brain networks. Neuroscientist 12:512–523
Bennett BD, Bolam JP (1994) Synaptic input and output of parvalbumin-immunoreactive neurons in the neostriatum of the rat. Neuroscience 62:707–719
Bevan MD, Booth PA, Eaton SA, Bolam JP (1998) Selective innervation of neostriatal interneurons by a subclass of neuron in the globus pallidus of the rat. J Neurosci 18:9438–9452
Bjaalie JG (2002) Localization in the brain: new solutions emerging. Nat Rev Neurosci 3:322–325
Bohland J, Bokil H, Allen C, Mitra P (2009a) The brain atlas concordance problem: quantitative comparison of anatomical parcellations. PloS One 4:e7200
Bohland JW, Wu C, Barbas H, Bokil H, Bota M, Breiter HC, Cline HT, Doyle JC, Freed PJ, Greenspan RJ, Haber SN, Hawrylycz M, Herrera DG, Hilgetag CC, Huang JZ, Jones A, Jones EG, Karten HJ, Kleinfeld D, Kötter R, Lester HA, Lin JM, Mensh BD, Mikula S, Panksepp J, Price JL, Safdieh J, Saper CB, Schiff ND, Schmahmann JD, Stillman BW, Svoboda K, Swanson LW, Toga AW, Van Essen DC, Watson JD, Mitra PP (2009b) A proposal for a coordinated effort for the determination of brainwide neuroanatomical connectivity in model organisms at a mesoscopic scale. PLoS Comput Biol 5:1–9
Bosch C, Mailly P, Degos B, Deniau JM, Venance L (2012) Preservation of the hyperdirect pathway of basal ganglia in a rodent brain slice. Neuroscience 215:31–41
Bota M, Swanson L (2007a) The neuron classification problem. Brain Res Rev 56:79–88
Bota M, Swanson L (2007b) Online workbenches for neural network connections. J Comp Neurol 500:807–814
Bota M, Swanson L (2010) Collating and curating neuroanatomical nomenclatures: principles and use of the brain architecture knowledge management system (BAMS). Front Neuroinf 4:1–16
Bounova G, de Weck O (2012) Overview of metrics and their correlation patterns for multiple-metric topology analysis on heterogeneous graph ensembles. Phys Rev E 85:1–11
Brandes U, Erlebach T (2005) Network analysis. Springer, Berlin
Bullmore E, Sporns O (2009) Complex brain networks: graph theoretical analysis of structural and functional systems. Nat Rev Neurosci 10:186–198
Burns G (1997) Neural connectivity of the rat: theory, methods and applications. PhD thesis, University of Oxford
Burns G, Cheng WC (2006) Tools for knowledge acquisition within the neuroscholar system and their application to anatomical tract-tracing data. J Biomed Discov Collab 1:10–16
Burns G, Young M (2000) Analysis of the connectional organization of neural systems associated with the hippocampus in rats. Phil Trans R Soc Lond Ser B 355:55–70
Burns G, Cheng WC, Thompson R, Swanson L (2006) The NeuARt II system: a viewing tool for neuroanatomical data based on published neuroanatomical atlases. BMC Bioinf 7:531–550
Burns G, Cheng WC, Thompson R, Swanson L (2008a) The NeuARt II system: a viewing tool for neuroanatomical data based on published neuroanatomical atlases. LNCS 5151:9–18
Burns G, Feng D, Hovy E (2008b) Intelligent approaches to mining the primary research literature: techniques, systems, and examples. In: Computational intelligence in medical informatics. Studies in computational intelligence, vol 85. Springer, Berlin, pp 17–50
Chida Y, Toyosawa K (1994) Laterality of theta waves recorded from the bilateral hippocampi in freely moving male rats. J Vet Med Sci 56:407–410
Colizza V, Flammini A, Serrano MA, Vespignani A (2006) Detecting rich-club ordering in complex networks. Nat Phys 2:110–115
Costa LdF, Sporns O (2006) Correlating thalamocortical connectivity and activity. Appl Phy Lett 89:1–3
DeLong MR (1990) Primate models of movement disorders of basal ganglia origin. Trends Neurosci 13:281–285
Diesmann M, Gewaltig MO, Aertsen A (1999) Stable propagation of synchronous spiking in cortical neural networks. Nature 402:529
Echtermeyer C, Costa L, Rodrigues F, Kaiser M (2011) Automatic network fingerprinting through single-node motifs. Plos One 6:1–9
Ercsey-Ravasz M, Markov NT, Lamy C, Van Essen DC, Knoblauch K, Toroczkai Z, Kennedy H (2013) A predictive network model of cerebral cortical connectivity based on a distance rule. Neuron 80:184–197
Estrada E (2010) Quantifying network heterogeneity. Phys Rev E 82(066):102
Estrada E, Hatano N (2008) Communicability in complex networks. Phys Rev E 77(036):111
Fagiolo G (2007) Clustering in complex directed networks. Phys Rev E 76(026):107
Felleman D, Essen DV (1991) Distributed hierarchical processing in the primate cerebral cortex. Cereb Cortex 1:1–47
Feng G, Burton N, Hill B, Davidson D, Kerwin J, Scott M, Lindsay S, Baldock R (2005) JAtlasView: a Java atlas-viewer for browsing biomedical 3D images and atlases. BMC Bioinform 6:1–7
Foster KC, Muth SQ, Potterat JJ, Rothenberg RB (2001) A faster Katz status score algorithm. Comput Math Organ Theory 7:275–285
Freeman LC (1977) A set of measures of centrality based on betweenness. Sociometry 40:35–41
Gerfen C (2004) Basal ganglia. In: The rat nervous system. Elsevier Academic Press, San Diego, pp 455–508
Gerfen C, Bolam J (2010) The neuroanatomical organization of the basal ganglia. In: Handbook of basal ganglia structure and function: a decade of progress. Elsevier Academic Press, San Diego, pp 3–28
Gewaltig MO, Diesmann M (2007) Nest (neural simulation tool). Scholarpedia 2:1430
Gouws A, Woods W, Millman R, Morland A, Green G (2009) DataViewer3D: an open-source, cross-platform multi-modal neuroimaging data visualization tool. Front Neuroinf 3:1–18
Gustafson C, Tretiak O, Bertrand L, Nissanov J (2004) Design and implementation of software for assembly and browsing of 3D brain atlases. Comput Methods Programs Biomed 74:53–61
Gustafson C, Bug W, Nissanov J (2007) Neuroterrain—a client–server system for browsing 3d biomedical image data sets. BMC Bioinform 8
Hagmann P, Cammoun L, Gigandet X, Meuli R, Honey CJ, Wedeen VJ, Sporns O (2008) Mapping the structural core of human cerebral cortex. PLoS Biol 6:1–15
Harriger L, van den Heuvel MP, Sporns O (2012) Rich club organization of macaque cerebral cortex and its role in network communication. PLoS One 7(e46):497
Hilgetag CC, Grant S (2000) Uniformity, specificity and variability of corticocortical connectivity. Philos Trans R Soc Lond B Biol Sci 355:7–20
Hjornevik T, Leergaard T, Darine D, Moldestad O, Dale A, Willoch F, Bjaalie J (2007) Three-dimensional atlas system for mouse and rat brain imaging data. Front Neuroinf 1:1–11
Honey C, Kötter R, Breakspear M, Sporns O (2007) Network structure of cerebral cortex shapes functional connectivity on multiple time scales. Proc Natl Acad Sci USA 104:10240–10245
Humphries M, Gurney K, Prescott T (2006) The brainstem reticular formation is a small-world, not scale-free, network. Proc R Soc B 273:503–511
Izhikevich EM, Edelman GM (2008) Large-scale model of mammalian thalamocortical systems. Proc Natl Acad Sci USA 105:3593–3598
Ju T, Warren J, Carson J, Bello M, Kakadiaris I, Chiu W, Thaller C, Eichele G (2006) 3d volume reconstruction of a mouse brain from histological sections using warp filtering. J Neurosci Methods 156:84–100
Keinan A, Hilgetag CC, Meilijson I, Ruppin E (2004) Causal localization of neural function: the shapley value method. Neurocomputing 58–60:215–222
Kita H (1993) GABAergic circuits of the striatum. Prog Brain Res 99:51–72
Kita H, Kitai ST (1994) The morphology of globus pallidus projection neurons in the rat: an intracellular staining study. Brain Res 636:308–319
Kleinberg JM (1999) Authoritative sources in a hyperlinked environment. J ACM 46:604–632
Koos T, Tepper JM (1999) Inhibitory control of neostriatal projection neurons by GABAergic interneurons. Nat Neurosci 2:467–472
Koslow S (1997) Neuroinformatic: an overview of the human brain project. Lawrence Erlbaum Associates, New Jersey
Koslow S (2005) Databasing the brain: from data to knowledge. Wiley, New York
Kötter R (2003) Neuroscience databases: a practical guide. Kluwer Academic Publishers, Dordrecht
Kötter R (2004) Online retrieval, processing, and visualization of primate connectivity data from the CoCoMac database. Neuroinformatics 2:127–144
Kötter R (2007) Anatomical concepts of brain connectivity. Springer, Berlin
Kötter R, Reid AT, Krumnack A, Wanke E, Sporns O (2007) Shapley ratings in brain networks. Front Neuroinf 1:1–9
Lanciego JL, Wouterlood FG (2011) A half century of experimental neuroanatomical tracing. J Chem Neuroanat 42:157–183
Lee J, Munch K, Carlis J, Pardo J (2008) Internet image viewer (iiv). BMC Med Imaging 29:1–20
Lima-Mendez G, van Helden J (2009) The powerful law of the power law and other myths in network biology. Mol BioSyst 5:1482–1493
Looi JC, Walterfang M, Nilsson C, Power BD, van Westen D, Velakoulis D, Wahlund LO, Thompson PM (2013) The subcortical connectome: hubs, spokes and the space between—a vision for further research in neurodegenerative disease. Aust N Z J Psychiatry EPub:1–4
Markov NT, Misery P, Falchier A, Lamy C, Vezoli J, Quilodran R, Gariel MA, Giroud P, Ercsey-Ravasz M, Pilaz LJ, Huissoud C, Barone P, Dehay C, Toroczkai Z, Van Essen DC, Kennedy H, Knoblauch K (2011) Weight consistency specifies regularities of macaque cortical networks. Cereb Cortex 21:1254–1272
Markov NT, Vezoli J, Chameau P, Falchier A, Quilodran R, Huissoud C, Lamy C, Misery P, Giroud P, Ullman S, Barone P, Dehay C, Knoblauch K, Kennedy H (2014) Anatomy of hierarchy: feedforward and feedback pathways in macaque visual cortex. J Comp Neurol 522(1):225–259
Martinez-Gonzalez C, van Andel J, Bolam JP, Mena-Segovia J (2013) Divergent motor projections from the pedunculopontine nucleus are differentially regulated in Parkinsonism. Brain Struct Funct 218 (epub ahead of print)
Martone ME, Gupta A, Ellisman MH (2004) E-neuroscience: challenges and triumphs in integrating distributed data from molecules to brains. Nat Neurosci 7:467–472
Mascaro MB, Bittencourt JC, Casatti CA, Elias CF (2005) Alternative pathways for catecholamine action in oral motor control. Neurosci Lett 386:34–39
Modha D, Singh R (2010) Network architecture of the long-distance pathways in the macaque brain. PNAS 107:13485–13490
Moore E, Poliakov A, Lincoln P, Brinkley J (2007) MindSeer: a portable and extensible tool for visualization of structural and functional neuroimaging data. BMC Bioinform 8(389):1–12
Nambu A, Tokuno H, Takada M (2002) Functional significance of the cortico-subthalamo-pallidal ’hyperdirect’ pathway. Neurosci Res 43:111–117
Newman M (2010) Networks: an introduction. Oxford University Press Inc., New York
Newman M, Girvan M (2004) Finding and evaluating community structure in networks. Phys Rev 69:1–15
Newman M, Barabasi AL, Watts DJ (eds) (2006) The structure and dynamics of networks. Princeton University Press, New Jersey
Niggemann J, Gebert A, Schulz S (2008) Model formulation: modeling functional neuroanatomy for an anatomy information system. JAMIA 15:671–678
Oh SW, Harris JA, Ng L, Winslow B, Cain N, Mihalas S, Wang Q, Lau C, Kuan L, Henry AM, Mortrud MT, Ouellette B, Nguyen TN, Sorensen SA, Slaughterbeck CR, Wakeman W, Li Y, Feng D, Ho A, Nicholas E, Hirokawa KE, Bohn P, Joines KM, Peng H, Hawrylycz MJ, Phillips JW, Hohmann JG, Wohnoutka P, Gerfen CR, Koch C, Bernard A, Dang C, Jones AR, Zeng H (2014) A mesoscale connectome of the mouse brain. Nature 508:207–214
Ozik J, Hunt B, Ott E (2004) Growing networks with geographical attachment preference: emergence of small worlds. Phys Rev E 69(026):108
Palombi O, Shin JW, Watson C, Paxinos G (2006) Neuroanatomical affiliation visualization-interface system. Neuroinformatics 4:299–317
Parent A (1986) Comparative neurobiology of the basal ganglia. Wiley, New York
Parthasarathy K, Bhalla US (2013) Laterality and symmetry in rat olfactory behavior and in physiology of olfactory input. J Neurosci 33:5750–5760
Passingham RE, Stephan KE, Kotter R (2002) The anatomical basis of functional localization in the cortex. Nat Rev Neurosci 3:606–616
Paxinos G, Watson C (2004) The rat brain in stereotaxic coordinates, 5th edn. Elsevier Academic Press, London
Paxinos G, Watson C (2007) The rat brain in stereotaxic coordinates, 6th edn. Elsevier Academic Press, Amsterdam
Paxinos G, Watson C (2009) BrainNavigator. Academic Press Inc., San Diego
Press W, Olshausen B, Van Essen D (2001) A graphical anatomical database of neural connections. Phil Trans R Soc Lond B 356:1147–1157
Riddle DR, Purves D (1995) Individual variation and lateral asymmetry of the rat primary somatosensory cortex. J Neurosci 15:4184–4195
Rubinov M, Sporns O (2010) Complex network measures of brain connectivity: uses and interpretations. Neuroimage 52:1059–1069
Sato F, Lavallee P, Levesque M, Parent A (2000) Single-axon tracing study of neurons of the external segment of the globus pallidus in primate. J Comp Neurol 417:17–31
Schmitt O, Eipert P (2012) Neuroviisas: approaching multiscale simulation of the rat connectome. Neuroinformatics 10:243–267
Schmitt O, Eipert P, Philipp K, Kettlitz R, Fuellen G, Wree A (2012) The intrinsic connectome of the rat amygdala. Front Neural Circuits 6:81
Schreiber F, Schwöbbermeyer H (2004) Towards motif detection in networks: frequency concepts and flexible search. In: Proceedings of the international workshop on network tools and applications in biology NETTAB04, pp 91–102
Schreiber F, Schwöbbermeyer H (2005) Mavisto: a tool for the exploration of network motifs. Bioinformatics 21:3572–3574
Shanahan M, Wildie M (2012) Knotty-centrality: finding the connective core of a complex network. PLoS One 7(e36):579
Shanahan M, Bingman VP, Shimizu T, Wild M, Güntürkün O (2013) Large-scale network organization in the avian forebrain: a connectivity matrix and theoretical analysis. Front Comp Neurosci 7(89):1–17
Smith Y, Bevan MD, Shink E, Bolam JP (1998) Microcircuitry of the direct and indirect pathways of the basal ganglia. Neuroscience 86:353–387
Sporns O (2002) Graph theory methods for the analysis of neural connectivity patterns. In: Kötter R (ed) Neuroscience databases. Kluwer Academic, Boston, pp 171–185
Sporns O (2006) Small-world connectivity, motif composition, and complexity of fractal neuronal connections. Biosystems 85:55–64
Sporns O (2011) Networks of the brain. The MIT Press, Cambridge
Sporns O, Kötter R (2004) Motifs in brain networks. PLoS Biol 2:1910–1918
Sporns O, Tononi G, Edelman GM (2000a) Connectivity and complexity: the relationship between neuroanatomy and brain dynamics. Neural Netw 13:909–922
Sporns O, Tononi G, Edelman GM (2000b) Theoretical neuroanatomy: relating anatomical and functional connectivity in graphs and cortical connection matrices. Cereb Cortex 10:127–141
Sporns O, Tononi G, Edelman GM (2002) Theoretical neuroanatomy and the connectivity of the cerebral cortex. Behav Brain Res 135:69–74
Sporns O, Chialvo DR, Kaiser M, Hilgetag CC (2004) Organization, development and function of complex brain networks. Trends Cogn Sci 8:418–425
Sporns O, Honey CJ, Kötter R (2007) Identification and classification of hubs in brain networks. PLoS One 2:1–14
Stam CJ, Reijneveld JC (2007) Graph theoretical analysis of complex networks in the brain. Nonlinear Biomed Phys 1:1–19
Stephan K, Kamper L, Bozkurt A, Burns G, Young M, Kötter R (2001a) Advanced database methodology for the collation of connectivity data on the macaque brain (CoCoMac). Philos Trans R Soc Lond B Biol Sci 356:1159–1186
Stephan KE, Kamper L, Bozkurt A, Burns GAPC, Young MP, Kötter R (2001b) Advanced database methodology for the Collation of Connectivity data on the Macaque brain (CoCoMac). Philos Trans R Soc Lond Ser B 356:1159–1186
Steriade M (2004) Neocortical cell classes are flexible entities. Nat Rev Neurosci 5:121–134
Sugar J, Witter M, van Strien N, Cappaert N (2011) The retrosplenial cortex: intrinsic connectivity and connections with the (para)hippocampal region in the rat. An interactive connectome. Front Neuroinf 5:1–13
Swanson L (2003) Brain maps: structure of the rat brain, vol 3. Elsevier, Amsterdam
Tepper JM, Abercrombie ED, Bolam JP (2007) Basal ganglia macrocircuits. Prog Brain Res 160:3–7
Thibeault CM, Srinivasa N (2013) Using a hybrid neuron in physiologically inspired models of the basal ganglia. Front Comput Neurosci 7:88. (eCollection)
Tiesinga P, Fellous JM, Sejnowski TJ (2008) Regulation of spike timing in visual cortical circuits. Nat Rev Neurosci 9:97–107
Towlson EK, Vértes PE, Ahnert SE, Schafer WR, Bullmore ET (2013) The rich club of the C. elegans neuronal connectome. J Neurosci 33:6380–6387
van den Heuvel MP, Sporns O (2011) Rich-club organization of the human connectome. J Neurosci 31:15775–15786
Van Horn JD, Gazzaniga MS (2013) Why share data? Lessons learned from the fMRIDC. Neuroimage 82:677–682
Vlachos I, Aertsen A, Kumar A (2012) Beyond statistical significance: implications of network structure on neuronal activity. PLoS Comput Biol 8(e1002):311
Wang Q, Sporns O, Burkhalter A (2012) Network analysis of corticocortical connections reveals ventral and dorsal processing streams in mouse visual cortex. J Neurosci 32:4386–4399
Ware C (2013) Information visualization: perception for design. Morgan Kaufmann Publishers Inc., San Francisco
Wong PC, Foote H, Chin G, Mackey P, Perrine K (2006a) Graph signatures for visual analytics. IEEE Trans Vis Comput Graph 12:1399–1413
Wong PC, Foote H, Mackey P, Perrine K, Chin G (2006b) Generating graphs for visual analytics through interactive sketching. IEEE Trans Vis Comput Graph 12:1386–1398
Yip AM, Horvath S (2007) Gene network interconnectedness and the generalized topological overlap measure. BMC Bioinform 8:22
Young M (1992a) Objective analysis of the topological organization of the primate cortical visual system. Nature 358:152–155
Young MP (1992b) Objective analysis of the topological organization of the primate cortical visual system. Nature 358:152–155
Young M (1993) The organization of neural systems in the primate cerebral cortex. Proc Biol Sci 252:13–18
Young M, Scannell J, Burns G, Blakemore C (1994) Analysis of connectivity: neural systems in the cerebral cortex. Rev Neurosci 5:227–250
Young M, Scannell J, Burns G (1996) The analysis of cortical connectivity. Springer, Berlin
Zhou S, Mondragn RJ (2004) The rich-club phenomenon in the internet topology. IEEE Commun Lett 8:180–182
Acknowledgments
The authors thank Klaus-Peter Schmitz (Department of Biomedical Engineering, University of Rostock) for the support of the neuroVIISAS project. We thank Frauke Winzer, Susanne Lehmann, Antje Schümann, Jennifer Meinhardt, Ann-Christin Klünker for their faithful work on extending the database and mappings. All work was supported by the Faculty of Mathematics and Natural Sciences and of the Faculty of Medicine of the University of Rostock.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix A
The appendix contains in the first part figures, in the second part tables and the third part formal definitions of matrices, graph-theoretical parameters and randomization models.
See Figs. 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26.
Appendix B
Tables 7, 8, 9, 10, 11, 12, 13, 14, 15.
Appendix C: Formal definitions of parameters, matrices and concepts
The definitions of expressions, parameters, matrices and simulation models (random graph models) used in this article are summarized in the following. More detailed descriptions, algorithms and proofs are provided elsewhere (Newman 2010; Rubinov and Sporns 2010; Newman et al. 2006; Kötter 2007; Brandes and Erlebach 2005; Kötter 2003).
Basic definitions
Node, vertex: The smallest subunit of a network. With regard to connectomes a node is a circumscribed or disjunctive region that contains neuron perikarya (sources of physiological action potential) and/or axonal terminals (targets of physiological action potentials).
Set of indexed nodes: The set of all indices of nodes is
Number of nodes: The number of nodes (regions, vertices) is
Edge: A directed edge \((i, j) \in N \times N\) is the line that connects vertices i and j with source i and target j. The set of directed edges E is
Edges: The number of edges (connections, links) \(\epsilon \) is
Set of edges:
The set of all not self-referencing edges is
Graph:
Adjacency matrix: The adjacency matrix (connectivity matrix) A is
Weighted matrix: The weighted matrix W is
whereas \(w_{ij}\) is the weight of the edge \((i, j)\) that connects i and j.
\(0 \le w_{ij} \le 1\).
Path: A sequence of vertices (\(v_1,\ldots v_k\)) is a path from (\(v_1\) to \(v_k\)) if \(\forall _i \in \{1,\ldots ,k-1\}:(v_i, v_{i+1}) \in E\). The length of a path \(v_1,\ldots ,v_k\) is \(k-1\).
Distance matrix: The distance matrix D is
where
Generalized topology matrices (GTOM): Let \(N_m(i)\) be the \(m-step\) neighborhood of node i:
then the GTOM-matrix of step m is defined as
The definitions of \({\rm GTOM}^{\rm in}_{(m)}\) and \({\rm GTOM}^{\rm out}_{(m)}\) with the directed \(m-step\) neighborhoods \(N^{\rm in}_m(i)\) and \(N^{\rm out}_m(i)\) are analog.
Degree all (\({\rm deg}_{{\rm all}}\), \({\rm DG}_{{\rm All}}\)): Self-references of nodes are not considered for all three degree measures.
\({\rm deg} (i) = {\rm deg}_{{\rm all}}(i)\)
Degree out:
Degree in:
Reciprocal edge count \({\rm Rec}(i)\): \({\rm Rec}(i)\) is the number of reciprocal edges adjacent to a node i.
Laterality: Let \(N^{\rm IPSI} \subseteq N\) be a subset of nodes and \({\rm deg} ^{\rm IPSI}(i)\) the degree of the node i in the subset \(N^{\rm IPSI}\). Then the lateralities are defined as follows:
Laterality of the reciprocal edge count \({\rm Lat}_{{\rm Rec}}(i)\):
The laterality of the reciprocal edge count is the fraction of ipsilateral reciprocal edges. For a node \(i \in N^{\rm IPSI} \subseteq N\).
Neighborhoods: Out-neighbors of i:
In-neighbors of i:
All neighbors of i:
Network parameters
Communicability matrix G:
Modularity measure: Let \(M=\{M_1,\ldots ,M_m\}\) be a partition of N. \(M_i\) is a group, module or cluster of vertices. With
the fraction of edges that fall within group \(M_i \subseteq N\) and
the fraction of ends of edges that are attached to vertices in group \(M_i\), the Modularity
whereas \(a^2_i\) is the fraction of edges that would connect vertices within group \(M_i\) if they were connected at random. A large modularity implies that the fraction of edges that fall within groups is larger than expected in the random case. The partition is generated by a “Greedy” optimization algorithm. Starting with a partition where every single node has its own group, stepwise those two groups are joined that increase Q most. The algorithm ends if there are no more such groups. The weighted case is similar, only the \(a_{ij}\) are replaced by \(w_{ij}\) and \(\ell \) is replaced by the sum of the edge weights
The method of Newman and Girvan (2004) was used.
Global efficiency GE:
\({\rm GE}^\rightarrow \) and \({\rm GE}^{\overrightarrow{w}}\) analog with \(d^\rightarrow (i,j)\) and \(d^{\overrightarrow{w}}(i,j)\)
Directed global efficiency:
Harmonic mean HM:
The directed and weighted versions use the directed and weighted global efficiencies.
Local efficiency: The local efficiency indicates how strong neighbors of nodes are interconnected. For each node i the inverse lengths of the shortest paths of the neighbors of i that are passing i are added. The local efficiency is this sum divided by the maximal possible sum of paths between neighbors that are connected with i. The efficiency of the network (global efficiency) is the average local efficiency of all nodes.
Directed local efficiency:
Weighted directed local efficiency:
whereby \(n_i=|N_i|\) and \(d_{jk}(N_i)\), respectively, \(d^{\overrightarrow{w}}_{jk}(N_i)\) is the length of the shortest path between j and k that contains only neighbors of i.
Directed assortativity coefficient \(r^{\rightarrow }\):
Directed and weighted assortativity coefficient \(r^{\overrightarrow{w}}\):
The correlation of the degrees of nodes that are connected: \(-1 \le r \le 1\). Large positive values imply that nodes are mainly connected to nodes with similar degrees. Large negative values imply that nodes with a large degree are mainly to nodes that have a small degree. If \(r \approx 0\) there is no relation detectable.
Average path length = characteristic path length \((\overline{d})\): With \(P=\{(i,j) \in N \times N | d(i,j) < \infty \}\), the set of paths.
In the weighted case the distances \(d(i,j)\) are replaced by the weighted distances.
Average directed degree \(\overline{{\rm deg} }\):
Heterogeneity VC: Coefficient of variation (VC) of the \({\rm degree}_{\rm all}\) parameter.
If \(H_{\rm VC}=0\), all nodes have the same degree. The larger \(H_{\rm VC}\) the more diverse are the node degrees. In the weighted case the versions of the degrees are used. The heterogeneity measure of Estrada (2010) was not implemented because it is not defined for directed and weighted graphs.
Line density Ld:
Without self-referencing edges.
Rich-club coefficient \(\phi (k)\): With \(N_k=\{i \in N | {\rm deg} (i) > k\}\) and \(E_k=\{(i,j) \in N_k \times N_k | (i,j) \in E\}\) \(G_k=(N_k, E_k)\) is the subgraph of \(G=(N,E)\) containing all vertices with a degree greater than k. The rich-club coefficient \(\phi (k)\) of a graph G is defined as the line density of the subgraph \(G_k\):
Diameter Diam:
Katz index: The Katz index (Katz status index, Katz centrality) is a measure for the direct and indirect input of a node (Foster et al. 2001).
The attenuation factor \(\alpha \) has to be smaller than the reciprocal of the absolute value of the largest eigenvalue of A. For a better readability and comparability of the results, in neuroVIISAS the Katz centrality is multiplied by the mean of the quotient \(\frac{{\rm deg} _{\rm in}(i)}{C_{{\rm Katz}}(i)}\) of all nodes with \(C_{{\rm Katz}}(i) > 0\). Hence, the values lie in the same range as the indegrees.
Number of triangles:
The maximum number of possible triangles that can be deviated from a complete reciprocal triangle is 8.
Weighted number of triangles:
Instead of the sum of triangles (\(t^{\rightarrow }(i)\)) the sum of geometric means of edge weights of each triangle is calculated. The following example provides \((w_{ij} \cdot w_{jk} \cdot w_{ik})^{\frac{1}{3}}\) as the summand:
Directed transitivity: The general definition of transitivity (T) is the sum of number of triangles around all nodes divided by the maximum possible sum of triangles around all nodes.
Directed and weighted transitivity:
whereby \(t_{\max }(i)= {\rm deg} (i) \cdot ({\rm deg} (i)-1) - 2 \cdot \hbox {rec}(i)\) with \({\rm deg} (i)=\) number of adjacent edges of i and \(\hbox {rec}(i)=\) number of reciprocal edges of i (the two directions of one reciprocal edge are considered as one reciprocal edge).
The degree \({\rm deg} \) and the reciprocity \(\hbox {rec}\) are defined as:
For the directed and weighted case:
Cluster coefficient (triangle based): The triangle based cluster coefficient (Fagiolo 2007) of a node n is the number of triangles around n divided by the maximum possible number. In this version of the cluster coefficient reciprocal edges to a neighbor of a node n can affect the cluster coefficient of node n. In the other version only edges between neighbors of n have an influence to the cluster coefficient of node n.
Cluster coefficient: Number of edges between the neighbors of a node divided by the maximum possible number. \(C^{\rightarrow }(i)\) refers to all neighbors of i.
\(C^{\rightarrow }_{\rm out}(i)\) refers to the out-neighbors of i.
\(C^{\rightarrow }_{\rm in}(i)\) refers to the in-neighbors of i.
In the weighted case the \(a_{ij}\) are replaced by the \(w_{ij}\).
Average cluster coefficient:
and
Small worldness S:
Centrality:
This centrality (degree centrality) is defined for an undirected network based on undirected degrees. A directed or weighted version is not available yet. For the calculation the directed network is transferred to an undirected one.
Circle length LC:
Eccentricity out: Eccentricity out, the output eccentricity of the vertex i is the maximum distance from i to any vertex.
Eccentricity in: Eccentricity in, the input eccentricity of the vertex i is the maximum distance from i to any vertex.
Cluster-coefficient of second neighbors: The cluster-coefficient of second neighbors (Hierarchical directed cluster coefficient of second (indirect) neighbors) \(C_2(i)\) is the number of edges between the 2nd neighbors of node i, divided by the maximum possible number of edges. In the weighted case it is the sum of weights of the edges between the 2nd neighbors of node i, divided by the maximum possible sum. With
the set of second neighbors of node i is:
In the weighted case the \(a_{ij}\) are replaced by \(w_{ij}\).
Average neighbor degree: The non-weighted average neighbor degree \({\rm NB}(i)\) of node i is
Weighted average neighbor degree: The weighted average neighbor degree \({\rm NB}(i)\) of node i is
Variation coefficient of neighbor degree:
The weighted case is analogue.
Locality index of node i \(({\rm Loc}(i))\): The locality index of node i is the fraction of edges adjacent to nodes in \(N^{+}_i\) whose source and target lie in \(N^{+}_i\).
The weighted case is analogue. A value of 0 means that the node is isolated. The larger the value, the less edges connect the neighborhood of i to outside node. The maximum of one is reached if the neighborhood of i is not connected to outside nodes.
Closeness centrality out \({\rm CC}^{\rm out}(i)\): The closeness centrality out with indices of nodes from which node i can be reached (\({\rm RN}^{\rm OUT}(i)\))\({\rm RN}^{\rm OUT}(i)=\lbrace j \in N \backslash \{i\} | d (i,j) < \infty \rbrace \)
Closeness centrality in \({\rm CC}^{\rm in}(i)\): The closeness centrality in with indices of nodes which can be reached from node i (\({\rm RN}^{\rm IN}(i)\))\({\rm RN}^{\rm IN}(i)=\lbrace j \in N \backslash \{i\} | d (j,i) < \infty \rbrace \)
Betweenness centrality BC:
where \(\rho _{j,k}\) is the number of shortest paths from j to k and \(\rho _{j,k}(i)\) is the number of shortest paths from j to k that pass through i. The directed and weighted definitions are the same.
Knotty centrality KC: Let \(\hat{N} \subseteq N\), \(|\hat{N}| > 1\) be a subset of N. Then
with the line density \({\rm Ld}(\hat{N})\) of the subgraph \(\hat{N}\). The knotty center of a graph G is a subset \(N_{\rm KC}\) of nodes with
\({\rm KC}(N_{\rm KC})\) is called the knotty-centerdness of the graph G.
The knotty-centrality of a node i is defined as
Stress S:
The directed and weighted definitions are the same.
Central point distance CPD:
where \( {\rm BC}_{\max }=\max\nolimits_{i \in N} \lbrace {\rm BC}(i) \rbrace \) is the maximum Betweenness centrality. The directed and weighted versions use the directed and weighted Betweenness centralities.
Participation coefficient: The partition \(M=\{M_1,\ldots M_m\}\) is generated as described in the definition of modularity.
with \(x \in \{\rm in, out, all\}\) and
(Number of edges from vertices of \(M_j\) to i).
(Number of edges from i to vertices of \(M_j\)).
(Number of edges between i and vertices of \(M_j\)).
with the same x and weighted definitions of degrees. One has \(0 \le {\rm PC}(i) \le 1\). If \({\rm PC}(i)=1\), the node i has no edges (in, out, all). If \({\rm PC}(i)=0\) all edges (in, out all) come from, go to or stay in the same cluster. The larger \({\rm PC}(i)\) the more clusters are involved in the edges of node i.
Z score/within module degree: Let \(M_i\) be the module containing node i. \({\rm deg} _x(i, M_i)\) \(x\in \{\rm in, out, all\}\) is defined in the participation coefficient.
is the mean and
the standard deviation of the within module \(M_i\) degree distribution. Then the Z score is defined as
and analogous
with the weighted versions of the mean and standard deviation. A value above one or below minus one implies that a node has significantly more or less edges from, to or from and to nodes in its cluster than the average node in its cluster has.
Eigenvector centrality: The eigenvector centrality \({\rm EC}(i)\) is the i-th component of the eigenvector with the largest corresponding eigenvalue of the adjacency matrix resp. weight matrix.
Shapley rating \(\phi \): The Shapley rating is a measure that provides information about the loss of connectivity following the removal of a node.
where \({\rm SCC}(\hat{N})\) is the set of strongly connected components of \(\hat{N}\). The smaller the value is, the more important is the node in the sense of connectivity of the graph. Because of the exponential number of subsets, this parameter can be approximated for large networks, only.
Radiality: The radiality of a node Rad is a measure of the distance of a node to all other nodes. Nodes that have a small radiality have larger distances to other nodes than those with a greater radiality.
Input radiality \({\rm Rad}_{\rm in}\): The input radiality of a node \({\rm Rad}_{\rm in}\) is
In the weighted case the weighted distances are used.
Output radiality \({\rm Rad}_{\rm out}\): The output radiality of a node \({\rm Rad}_{\rm out}\) is
In the weighted case the weighted distances are used.
Centroid value Cen: With \(g_{\rm out}(i,j)= |\{k \in N | d(i,k) < d(j,k) < \infty \}|\) and \(g_{\rm in}(i,j)= |\{k \in N | d(k,i) < d(k,j) < \infty \}|\) which are the number of nodes closer to node i than to node j with regard to In- and Out-distance, the centroid value is defined in the following.
Output centroid value \({\rm Cen}_{\rm out}\):
Input centroid value \({\rm Cen}_{\rm in}\):
In the weighted case the weighted distances are used. A value \({<}0\) implies, that there exists a node that is closer to most other nodes. A value \({\ge }0\) implies, that this node is most central in the network. A value \({=}0\) implies, that there are more than one most central nodes.
Page rank centrality PRC: \({\rm PRC}(i)= r_i\) where r is the solution of the linear system
with the damping factor \(\alpha =0.85\), the identity matrix I and the diagonal matrix B, whereby
In the weighted case the weight matrix W is used instead of A and the weighted version \({\rm deg} ^{\overrightarrow{w}}_{\rm out}(i)\) of the outdegree.
Flow coefficient FC: Number of paths of length 2 between neighbors of a node i that pass node i divided by the maximum possible numbers of sub paths.
In the weighted case we define the flow coefficient as the sum of weights of paths of length 2 between neighbors of a node i that pass node i divided by the maximum possible sum.
Average flow coefficient FC: Number of paths of length 2 between neighbors of a node i that pass node i divided by the maximum possible numbers of sub paths.
Subgraph centrality SC:
The subgraph centrality of the network is the average subgraph centrality of its nodes.
Undirected cyclic coefficient CyclC: The undirected cyclic coefficient as published by (Kim et al. 2005) Cyclic topology in complex networks.
With
Directed cyclic coefficient \({\rm CyclC}^\rightarrow \): A publication about the directed cyclic coefficient is unknown. The directed cyclic coefficient is implemented here as follows:
Directed weighted cyclic coefficient \({\rm CyclC}^{\overrightarrow{w}}\): A publication about the directed weighted cyclic coefficient is unknown. The directed weighted cyclic coefficient is implemented here as follows:
with \({\rm dist}^w_i(j,k)\) is the weighted version of \({\rm dist}_i(j,k)\) with the weighted path length.
Cyclic network coefficient \({\rm CyclC}^\rightarrow \): The cyclic coefficient of the network is the average cyclic coefficient of its nodes:
Hubness and authoritativeness: A hub is a node that points to many authorities and an authority is a node that has numerous input connections from many hubs (Kleinberg 1999; Sporns et al. 2007). The hubness \({\rm Hub}(i)\) of a node i is:
with the authoritativeness \({\rm Auth}(i)\)
An iterative algorithm is used to calculate a fixed point of these equations.
Vulnerability V: The vulnerability V is the maximum relative decrease of the global efficiency removing a single node.
where \({\rm GE}(i)\) is the global efficiency of the graph (\(N \backslash \{i\}, \{(j,k) \in E | j \ne i \ne k \}\)) that originates by removal of node i and all edges adjacent to i. The weighted version is analog using the weighted global efficiencies.
Random models
The following random graph models are compared to the real network of the intrinsic amygdala connectivity. By comparing the average path length and the cluster coefficient of the models with the real network it is feasible to determine a model that is most similar to the real network.
Erdös Rényi graph:
where n is the number of vertices and p is the probability that an edge \((i, j)\) exists, for all \(i, j\). The degree distribution of the Erdös Rényi random graph is binomial in terms of
Watts–Strogatz graph: The small-world model of Watts–Strogatz is a random graph generation model that provides graphs with small-world properties. The network (initially it has a non-random lattice structure) is build by linking each node to its \(\langle k \rangle \) closest neighbors using a rewiring probability p. Hence, an edge has the probability p that it will be rewired as a random edge. The number of rewired links can be estimated by:
Barabasi–Albert graph: The Barabasi–Albert graph is used to generate preferential attachments between nodes. The probability \(p_i\) that the new node is connected to node i is
The degree distribution of a Barabasi–Albert network is scale free following the power law distribution of the form:
Eipert graph: The modified Eipert model (EN: Eipert network) is based on the Barabasi–Albert graph. However, the algorithm starts at a fixed number of nodes and edges are added iteratively.
Ozik–Hunt–Ott graph: The Ozik–Hunt–Ott model (OHO) (Ozik et al. 2004) is a small-world randomization approach that was modified for directed networks and a fixed number of edges. The OHO-model uses a growing mechanism in which all connections are made locally to topographical nearby regions.
Rewiring graph: The rewiring-models connects each target of an edge of a network to another target node.
Power law:
\(\Delta \) is the deviation (error) of an empirical distribution of degrees from the power law function. A small \(\Delta \) value means that the empirical distribution is similar with the power law function.
Rights and permissions
About this article
Cite this article
Schmitt, O., Eipert, P., Kettlitz, R. et al. The connectome of the basal ganglia. Brain Struct Funct 221, 753–814 (2016). https://doi.org/10.1007/s00429-014-0936-0
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00429-014-0936-0