CoCoMac (Tracer)
is the main database and contains data from tracing studies on anatomical
connectivity in the macaque cerebral cortex. Currently (February 2001),
it comprises more than 20,000 experimental findings from more than 1,300
tracer injections as well as 5,400 relations between more than 260 brain
maps. The data include:
exact references (pages, tables, figures of respective reports)
experimental conditions (e.g. tracer substance, injection method, animals
used, post-mortem time),
injections (e.g. localisation, volume, concentration, involved laminae),
labelled sites (e.g. localisation, laminar patterns, density),
quantitative data (e.g. exact numbers of labelled cells),
definitions of brain maps,
relations between different parcellation schemes.
The CoCoMac
Online Interface is still in the process of
development and testing. You are invited to register to try this preliminary
test version and we would be happy to receive your comments. Please note
that because of the ongoing development the interface may show unpredictable
results and behaviour.
CoCoMac Online Test Interface at
http://cocomac.org
Alternate development site:
http://www.mon-kunden.de/cocomac
Note that the two sites are completely independent: they may have slightly
different contents and functionality, and they require separate registration.
The interface can be queried automatically using
extended URL strings, whose vocabulary and syntax are documented in a
separate document. Output is either textual in the browser or XML-formatted
according to the CoCoMac XML schema.
If you want to obtain a connectivity matrix from the
interface then a convenient way is to save the connectivity output list in XML
format and open it with the
CoCoMac
IO tool written by Michael Capalbo. This allows you to save the connectivity
data as a list or a matrix in text or MS Excel format.
The CoCoMac Online Interface is complemented by a graphical front-end based on the Catacomb simulator written by Robert Cannon. This demonstrates the power of automated database queries, XML-based data exchange and Java-based visual tools.
Catacomb-CoCoMac interface and downloadable map displays
CoCoMac data can also be displayed using the
CARET
software developed by the Van Essen lab.
CoCoMac Stry
is similar in structure to CoCoMac-Tracer and contains virtually all published
data on functional connectivity from neuronographic studies in the macaque
cerebral cortex. It comprises almost 4000 experimental findings from 245
electrophysiological experiments. The resulting data have been published
in Stephan et al. 2000a (see References).
CoCoDat is
a handy tool for collation of electrophysiological and connectivity data
at the scale of neurons and microcircuitry in any species. Its structure
is fully compatible with the main CoCoMac approach. The database can be download as an Access 2000 mdb file or its contents can be viewed as data catalogs. We also provide an extensive manual and a specific mailing list.
More information on CoCoDat and the CoCoDat mailing list
The Documents page provides extensive manuals both for CoCoMac and CoCoDat
for download.
The References page lists papers documenting the design, contents and
application of these databases.
Higher-order analyses and functional interpretations
on the system level of the brain require systematic collation and integration
of a large number of experimental results. Databases of brain connectivity
have already led to a number of analyses and simulations that provided
valuable insights into the complex nature of cortical organization (e.g.
Felleman & Van Essen 1991; Young 1993; Scannell et al. 1995, 1999;
Burns 1997; Kötter et al. 2001). A major problem, however, is the
divergence of brain maps defined and used by different authors (the "parcellation
problem"). All earlier databases try to cope with this divergence by the
a priori definition of a "reference map" into which the individual findings
of different publications are mapped according to the subjective criteria
of the database collators.
We have tried to overcome the parcellation
problem by developing a new type of connectivity database. Our database CoCoMac
(Collations
of Connectivity data on the Macaque brain) contains the results of hundreds
of tracer studies in the original nomenclature of the respective authors
and is able to convert the data into freely chosen parcellation schemes.
This is achieved by a method called ORT
(Objective Relational Transformation), which operates
on coordinate-independent logical relations between different brain maps
using new classifications for connectivity data and graph-theoretical algorithms.
We use this database to analyse the structural and functional organization
of the cerebral cortex, and to establish structure/function relationships
by implementation of computer models that take into account the real anatomy
of the primate cerebral cortex. These will contribute to our understanding
of disorders such as epileptic spread of activity, functional disconnection
in schizophrenia, and reorganisation after stroke.
CoCoMac databases have been designed to achieve:
High transparency and objectivity
of data representation and data conversion:
Standardised protocols guide data extraction from the literature and data
input into CoCoMac.
Operationalised coding schemes classify the precision of all data types.
Data representation and data interpretation are strictly separated: (1)
only individual findings, no summarised or interpreted data are represented,
(2) data are represented exactly as described by the author (i.e. in her/his
original parcellation scheme).
Data conversion is achieved by Objective Relational Transformation (ORT),
and is thus fully formalised and reproducible.
High flexibility
in usage:
A wide range of data can be queried and analysed for many different purposes,
e.g. exact references, experimental protocols, injection sites, labeled
sites, laminar patterns and density of label, quantitative data (e.g. cell
counts), parcellation schemes, relations between brain maps.
Various constraints can be used for querying the database, e.g. used tracer
substances, left vs. right hemisphere, ipsi- vs. contralateral connections.
Data can be transformed into any user-defined parcellation schemes using
Objective Relational Transformation (ORT).
High adaptability
to
future requirements:
Relational structure of CoCoMac databases allows for flexible extensions.
Further data types (e.g subcortical structures, receptor distributions)
can be easily integrated into the framework of CoCoMac.
|