Bota, Mihail
Coauthors(s): Michael A. Arbib University Park Campus, Hedco Neuroscience Bldg, Room 7, Los Angeles, CA, 90089 email: arbib@pollux.usc.edu
University of Southern California, Human Brain Project
Neurobiology
University Park Campus, Hedco Neuroscience Bldg, Room 10A, Los Angeles, CA, 90089
rana.usc.edu:8376/~mbota/index2.html

The NeuroHomology Database

If one wants to define and evaluate the degree of homology between two brain structures from two species, then one has to correlate with what makes a given brain structure distinguishable from other structures. To define a neural structure, neuroscientists use numerous attributes including gross morphology, relative location, cytoarchitecture, types of cell responses to different ways of stimulation, and function. We identified eight criteria that can make a brain area distinguishable and can be used to evaluate the degree of homology between two brain structures. These eight criteria are the morphology of cells within a brain structure, the relative position, the cytoarchitecture, chemoarchitecture (neurotransmitters and enzymes that are found within a brain structure), myeloarchitecture, afferent and efferent connections, and function. Accordingly, we take into account all these attributes when evaluating the degree of homology of a pair of brain structures from two species. The NeuroHomology database is a summary database which contains related data on the brains of humans, monkeys and rodents, and provides a tool for discovery and evaluation of homologies between pairs of brain structures from different species. The database contains three interconnected entities: Brain Structures, Connections, and Homologies. The Brain Structures part of the Neurohomology database has as objects brain structures, as found in literature. The Connections part of the Neurohomology Database has as objects brain structures related by neuroanatomical connections. The Homologies part of the Neurohomology database can handle data about brain structures from any species. The NeuroHomology database can be searched independently by brain structures, connections and homologies. The search of existent data in the database and insertion of new data or annotations to any entry can be performed online. The brain structures are entered in the database as found in the investigated literature. For humans and monkeys we have used the Neuroanatomica glossary (Bowden & Martin, 1997). The hierarchy level and hierarchy path for each brain structure is entered as interpreted from the literature. The Connections part of the database evaluates the connection confidence level as well as the technique confidence level for each entry in the database. The quantification of the connection confidence level as reflected in an investigated paper is analogous with that used in the Neuroscholar database (Burns, 1997). Taking into account the relative advantages and limitations of each commonly used tract-tracing technique, we provide a method of quantification of the technique that was used to reveal a given connection. For each recorded connection in the database, we calculate a combined confidence level that takes into account the connections confidence level and the technique confidence level. If a given connection appears in a number of papers, then an overall confidence level is computed. The process of definition of homology at the level of brain structures implies a process of inference from distinct clusters of attributes. Thus, the concept of degree of homology is more appropriate to use when discussing homologies of brain structures across species. The degree of homology can take a minimal value of zero (no homology) and a maximal one of 1 (all criteria are fulfilled). For a given pair of brain structures from two different species, the degree of homology increases only if the number of fulfilled criteria is increased in all searched papers and remains constant if the same criteria are found in any number of citations. The importance of the NeuroHomology Database is that it provides an objective evaluation of the degree of homology between two brain structures. This can be especially useful for comparative neuroanatomists and computational neuroscientists. The database can be seen by comparative neuroanatomists as a tool for systematization of the existent data and by computational neuroscientists as a tool for evaluation of the degree of reliability to model brain structures from one species by using experimental data or characteristics of brain structures from other species. Moreover, the objective evaluations of connections and homologies can lead to predictions for future experiments for further subdivisions of brain nuclei and new homologies between brain structures. See http://bsl9.usc.edu/scripts/webmerger.exe?/database/homologies-main.html for further information. References: Bota, M., and Arbib, M.A. (2000). Neurohomology Database. To appear in Arbib M.A and Grethe J. (eds) Computing the Brain: A Guide to Neuroinformatics , Academic Press. Bowden, D.M, Martin R.F. (1997). A digital Rosetta stone for primate brain terminology. In: Bloom, F.E, Bjorklund, A., Hokfelt, T. (eds), Handbook of Chemical Neuroanatomy vol. 13: The Primate Nervous System, part I, pp:1-37 Burns, G.A.P.C (1997) Neural connectivity of the rat: Theory, methods and applications Oxford University D.Phil. Thesis. Nieuwenhuys, R, ten Donkelaar, H.C, Nicholson, C (1998). The central nervous system of vertebrates. Springer, pp:273-326