Abstract
When the dimensionality of a neural circuit is substantially larger than the dimensionality of the variable it encodes, many different degenerate network states can produce the same output. In this review I will discuss three different neural systems that are linked by this theme. The pyloric network of the lobster, the song control system of the zebra finch, and the odor encoding system of the locust, while different in design, all contain degeneracies between their internal parameters and the outputs they encode. Indeed, although the dynamics of song generation and odor identification are quite different, computationally, odor recognition can be thought of as running the song generation circuitry backwards. In both of these systems, degeneracy plays a vital role in mapping a sparse neural representation devoid of correlations onto external stimuli (odors or song structure) that are strongly correlated. I argue that degeneracy between input and output states is an inherent feature of many neural systems, which can be exploited as a fault-tolerant method of reliably learning, generating, and discriminating closely related patterns.
Similar content being viewed by others
Notes
More generally, if an n-dimensional linear system does not have n distinct eigenvalues, the system is said to be degenerate and contains multiple eigenvectors that are not linearly independent.
A variety of terms have been used to describe these population activity patterns—neural assemblies, neural ensembles, population activity vectors, etc. Within the context of this review, state vectors are a more natural terminology to describe the dimensionality and dynamics of neural circuits.
The development of more rigorous methods of identifying degeneracy is important (see Tononi et al. 1999, for an example), but at present the correlation method has been used as a reasonable first-pass at the problem.
For example, the model networks had a bias in the relative strength of fast and slow synapses in a way that favored the correct pyloric burst order, as is seen in experimental data.
Nonlinear syringeal dynamics can cause abrupt (< 1 ms) sound changes not directly controlled by the bird. However, for longer time scales (> ∼1 ms), there is generally a close relation between vocal muscle activity and acoustic structure.
Assuming that over the course of the 1 s song, all the HVC neurons are on once.
Note that a linear weighting is not strictly necessary for degeneracy; any projection of the high-dimensional RA state space (∼8000 neurons) onto the much lower-dimensional syringeal muscle state space (∼7 muscles) can produce a degenerate mapping between RA states and sound.
Note that each point in the song correlation matrix refers to two sounds in the song, and the corresponding points in the neural state correlation matrix refer to the two neural states that generated those sounds [after a latency correlation for the delay between RA activation and sound production; see Leonardo and Fee (2004) for analytical details].
While there are nonlinearities in the dynamics of the syrinx (Fee et al. 1998) this indicates that similar RA states can produce different sounds. It does not imply the converse (different RA states will produce similar sounds) that is the relevant variable for the present analysis.
RA is not the only source of degeneracy in the song production pathway—it has been shown that very different patterns of vocal muscle activity can generate the same types of sounds (Suthers and Hartley 1996).
The connection to mammalian olfaction, however, is less clear than to other vertebrate systems like the zebrafish.
The mushroom body is thought to be functionally similar to the piriform cortex in mammals.
KC odor selectivity is presumably even lower than 1/17, since only a limited number of odors can be tested in an experiment.
There are roughly 1000 glomeruli in the locust antennal lobe.
In many species, such as flies and moth, each ORN projects to a single glomerulus, the identity of which depends on the specific olfactory receptor it is expressing (Buck 1996). In locust, ORNs project to multiple glomeruli, such that there are more glomeruli than olfactory receptor types.
The LFP is present only during odor-presentation, its frequency is dependent on the PN-LN dynamics and not on the odor identity.
If chemically similar odors produced uncorrelated ORN patterns, PNs would simply inherit the degeneracy in the odor to ORN mapping, rather than generating it.
After the 200 ms of computation leading to the onset of decorrelation.
Individual RA neurons are connected to only a few syringeal muscles; similarly, PNs receive input from only a few types of ORNs.
However, see Barlow (2001) for a critical reinterpretation of this original idea.
Abbreviations
- LFP:
-
local field potential
- LN:
-
local neuron
- PN:
-
projection neuron
- ORN:
-
olfactory receptor neuron
- PVA:
-
population vector average
- RA:
-
robust nucleus of the arcopallium
- STG:
-
stomatogastric ganglion
References
Barlow H (1961) Possible principles underlying the transformation of sensory messages. MIT Press, Cambridge, MA
Barlow H (2001) Redundancy reduction revisited. Network: Computation in neural systems 12:241–253
Bazhenov M, Stopfer M, Rabinovich M, Abarbanel D, Sejnowski T, Laurent G (2001) Model of cellular and network mechanisms for odor-evoked temporal patterning in the locust antennal lobe. Neuron 30:569–581
de Bruyne M, Foster K, Carlson J (2001) Odor coding in the drosophila antenna. Neuron 30:537–552
Bucher A, Prinz A, Marder E (2005) Animal-to-animal variability in motor pattern production in adults and during growth. J Neurosci 25:1611–1619
Buck L (1996) Information coding in the vertebrate olfactory system. Ann Rev Neurosci 19:517–544
Chi Z, Margoliash D (2001) Temporal precision and temporal drift in brain and behavior of zebra finch song. Neuron 32:899–910
Coleman M, Vu E (2005) Recovery of impaired songs following unilateral but not bilateral lesions of nucleus uvaeformis of adult zebra finches. J Neurobiol 63:71–89
Edelman G, Gally J (2001) Degeneracy and complexity in biological systems. Proc Nat Acad Sci 98:13763–13768
Fee M, Shraiman B, Pesaran B, Mitra P (1998) The role of nonlinear dynamics of the syrinx in birdsong production. Nature 395:67–71
Fee M, Kozhevnikov A, Hahnloser R (2004) Neural mechanisms of vocal sequence generation in the songbird. Ann NY Acad Sci 1016:153–170
Fetz EE (1992) Are movement parameters recognizably coded in the activity of single neurons? Behav Brain Sci 15:679–690
Fiete I, Hahnloser R, Fee M, Seung H (2004) Temporal sparseness of the premotor drive is important for rapid learning in a neural network model of birdsong. J Neurophys 92:2274–2282
Fitch H, Kupin J, Kessler I, DeLucia J (2003) Relating articulation and acoustics through a sinusoidal description of vocal tract shape. Speech Commun 39:243–268
Friedrich R, Korsching S (1997) Combinatorial and chemotopic odorant coding in the zebrafish olfactory bulb visualized by optical imaging. Neuron 18:737–752
Friedrich R, Laurent G (2001) Dynamic optimization of odor representations by slow temporal patterning of mitral cell activity. Science 291:889–894
Friedrich R, Habermann C, Laurent G (2004) Multiplexing using synchrony in the zebrafish olfactory bulb. Nat Neurosci 7:862–871
Gardner T, Cecchi G, Magnasco M, Laje R, Mindlin GB (2001) Simple gestures for birdsongs. Phys Rev Lett 87:208101
Georgopoulos A, Schwartz A, Kettner R (1986) Neuronal population coding of movement direction. Science 233:1416–1419
Goldman M, Golowasch J, Marder E, Abbott L (2001) Global structure, robustness, and modulation of neuronal models. J Neurosci 21:5229–5238
Goller F, Cooper B (2004) Peripheral motor dynamics of song production in the zebra finch. Ann NY Acad Sci 1016:130–152
Goller F, Suthers R (1996) Role of syringeal muscles in controlling the phonology of bird song. J Neurophysiol 76:287–300
Golowasch J, Abbott L, Marder E (1999) Activity-dependent regulation of ionic currents in an identified neuron of the stomatogastric ganglion of the crab cancer borealis. J Neurosci 19:RC33
Greenewalt C (1968) Bird song: acoustics and physiology. Smithonian Institution Press, Washington DC
Guckenheimer J, Gueron S, Harris-Warrick R (1993) Mapping the dynamics of a bursting neuron. Philos Trans R Soc Lond B Biol Sci 341:354–359
Gurney ME (1981) Hormonal control of cell form and number in the zebra finch song system. J Neurosci 1:658–673
Hahnloser RHR, Kozhevnikov AA, Fee MS (2002) An ultra-sparse code underlies the generation of neural sequences in a songbird. Nature 419:65–70
Harris-Warrick R, Marder E, Selverston A, Mouline M (1992) Dynamic biological networks: the stomatogastric nervous system. MIT, Cambridge, MA
Heisenberg M, Borst A, Wagner S, Byers D (1985) Drosophila mushroom body mutants are deficient in olfactory learning. J Neuro-genet 2:1–30
Hildebrand J, Shepherd G (1997) Mechanisms of olfactory discrimination: converging evidence for common principles across phyla. Ann Rev Neurosci 20:595–631
Immelmann K (1969) Song development in the zebra finch and other estrilid finches. In: Hinde RA (ed) Bird Vocalizations. Cambridge University Press, New York, pp 61–74
Kanerva P (1988) Sparse distributed memory. MIT Press, Cambridge
Kao M, Doupe A, Brainard M (2005) Contributions of an avian basal ganglia-forebrain circuit to real-time modulation of song. Nature 433:638–643
Konishi M (1965) The role of auditory feedback in the control of vocalization in the white crown sparrow. Z Tierpsychol 22:770–783
Konishi M (1991) Deciphering the brain’s codes. Neural Comput 3:1–18
Laurent G (2002) Olfactory network dynamics and the coding of multidimensional signals. Nat Rev Neurosci 3:884–895
Laurent G, Davidowitz H (1994) Encoding of olfactory information with oscillating neural assemblies. Science 265:1872–1875
Laurent G, Stopfer M, Friedrich R, Rabinovich M, Volkovskii A, Abarbanel H (2001) Odor encoding as an active, dynamical process: experiments, computation and theory. Ann Rev Neurosci 24:263–297
Lee D, Seung H (1999) Learning the parts of objects by non-negative matrix factorization. Nature 401:788–791
Leonardo A (2002) Neural dynamics underlying complex behavior in a songbird. Caltech, PhD thesis, Pasadena
Leonardo A (2004) Experimental test of the birdsong error-correction model. Proc Nat Acad Sci 101:16935–16940
Leonardo A, Fee M (2005) Ensemble coding of vocal control in a songbird. J Neurosci 25:652–661
Leonardo A, Konishi M (1999) Decrystallization of adult birdsong by perturbation of auditory feedback. Nature 399:466–470
Lewis J, Kristan W (1998) A neuronal network for computing population vectors in the leech. Nature 391:76–79
Lohr B, Dooling R (1998) Detection of changes in timbre and harmonicity in complex sounds by zebra finches and budgerigars. J Comp Psych 112:36–47
MacLean J, Zhang Y, Johnson B, Harris-Warrick R (2003) Activity-independent homeostasis in rhythmically active neurons. Neuron 37:109–120
MacLeod K, Laurent G (1996) Distinct mechanisms for synchronizing and temporal patterning of odor-encoding neural assemblies. Science 274:976–979
Malnic B, Hirono J, Sato T, Buck L (1999) Combinatorial receptor codes for odors. Cell 96:713–723
Marr D (1969) A theory of cerebellar cortex. J Physiol (Lond) 202:437–470
Mindlin GB, Gardner TJ, Goller F, Suthers R (2003) Experimental support for a model of birdsong production. Phys Rev E 68
Ngai J, Chess A, Dowling M, Axel R (1993) Expression of odorant receptor genes in the catfish olfactory epithelium. Chem Senses 18:209–216
Nordeen K, Nordeen E (1992) Auditory feedback is necessary for the maintenance of stereotyped song in adult zebra finches. Behav Neural Biol 57:58–66
Nottebohm F, Stokes T, Leonard C (1976) Central control of song in the canary. J Comp Neurology 165:457–486
Olshausen B, Field D (1997) Sparse coding with an overcomplete basis set: a strategy imployed by V1?. Vis Res 37:3311–3325
Olveczky B, Andalman A, Fee M (2005) Vocal experimentation in the juvenile songbird requires a basal ganglia circuit. PLOS (in press)
Perez-Orive J, Mazor O, Turner G, Cassenaer S, Wilson R, GL (2002) Oscillations and sparsening of odor representations in the mushroom body. Science 297:359–365
Perez-Orive J, Bazhenov M, Laurent G (2004) Intrinsic and circuit properties favor coincidence detection for decoding oscillatory input. J Neurosci 24:6037–6047
Petersen R, Panzeri S, Diamond M (2001) Population coding of stimulus location in rat somatosensory cortex. Neuron 32
Prinz A, Bucher A, Marder E (2004) Similar network activity from disparate circuit parameters. Nat Neurosci 7:1345–1352
Reich D, Mechler F, Victor J (2001) Independent and redundant information in nearby cortical neurons. Science 294:2566–2568
Reichmann M, Markham R, Symons R, Rees M (1962) Experimental evidence for degeneracy of nucleotide triplet code. Nature 195:999
Reiner A, Perkel D, Bruce L, Butler A, Csillag A, Kuenzel W, Medina L, Paxinos G, Shimizu T, Streidter G, Wild M, Ball G, Durand S, Guturkun O, Lee D, Mello C, Powers A, White S, Hough G, Kubikova L, Smulders T, Wada K, Dugas-Ford J, Husband S, Yamaoto K, Yu K, Saing C, Jarvis E (2004) Revised nomenclature for avian telencephalon and some related brainstem nuclei. J Comp Neurol 473:377–414
Schneidman E, Bialek W, Berry M (2003) Synergy, redundancy, and independence in population codes. J Neurosci 23:11539–11553
Spiro JE, Dalva MB, Mooney R (1999) Long-range inhibition within the zebra finch song nucleus RA can coordinate the firing of multiple projection neurons. J Neurophysiol 81:3007–3020
Stopfer M, Laurent G (1999) Short-term memory in olfactory dynamics. Nature 402:664–668
Stopfer M, Bhagavan S, Smith B, Laurent G (1997) Impaired odor discrimination on desynchronization of odor-encoding assemblies. Nature 390:70–74
Stopfer M, Jayaraman V, Laurent G (2003) Intensity versus identity coding in an olfactory system. Neuron 39:991–1004
Strotmann J, Wanner I, Helfrich T, Beck A, Meinken C, Kubick S, Breer H (1994) Olfactory neurons expressing distinct odorant receptor subtypes are spatially segregated in the nasal neuroepithelium. Cell Tissue Res 276:429–438
Sturdy CB, Wild JM, Mooney R (2003) Respiratory and telencephalic modulation of vocal motor neurons in the zebra finch. J Neurosci 23:1072–1086
Suthers R, Hartley R (1996) Motor stereotypy and diversity in songs of mimic thrushes. J Neurobiol 30:231–245
Tchernichovski O, Nottebohm F, Ho C, Pesaran B, Mitra P (2000) A procedure for an automated measurement of song similarity. Anim Behav 59:1167–1176
Todorov E (2004) Optimality principles in sensorimotor control. Nat Neurosci 7:907–915
Tononi G, Sporns O, Edelman G (1999) Measures of degeneracy and redundancy in biological networks. Proc Nat Acad Sci 96:3257–3262
Turrigiano G, Abbott L, Marder E (1994) Actiivty-depedent changes in the intrinsic properties of cultured neurons. Science 264:974–977
Vicario DS (1991) Organization of the zebra finch song control system: II. Functional organization of outputs from nucleus Robustus archistriatalis. J Comp Neurol 309:486–494
Vu ET, Mazurek ME, Kuo Y (1994) Identification of a forebrain motor programming network for the learned song of zebra finches. J Neurosci 14:6924–6934
Wehr M, Laurent G (1996) Odor encoding by temporal sequences of firing in oscillating neural assemblies. Nature 384:162–166
Wild JM (2004) Functional neuroanatomy of the sensorimotor control of singing. Ann NY Acad Sci 1016:438–462
Wild J, Goller F, Suthers R (1998) Inspiratory muscle activity during bird song. J Neurobiol 36:441–453
Yu AC, Margoliash D (1996) Temporal hierarchical control of singing in birds. Science 273:1871–1875
Acknowledgements
The ideas in this paper have benefited from discussions with Gilles Laurent and Michale Fee. In addition, I thank Roian Egnor, Mark Konishi and Astrid Prinz for comments on the manuscript. Anthony Leonardo received support from the Helen Hay Whitney Foundation.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Leonardo, A. Degenerate coding in neural systems. J Comp Physiol A 191, 995–1010 (2005). https://doi.org/10.1007/s00359-005-0026-0
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00359-005-0026-0