Event Abstract

Do gold standards remain gold standards when compiling a large number of published tract-tracing studies into a connectivity database?

  • 1 Radboud University Nijmegen, Netherlands
  • 2 Research Center Jülich, Institute of Neuroscience and Medicine 6, Germany
  • 3 Ludwig-Maximilians-Universität München, Department Biology II, Germany
  • 4 RWTH Aachen University, Faculty of Medicine, Germany
  • 5 RIKEN Brain Science Institute, Japan

Tract-tracing experiments are considered the gold standard when it comes to revealing structural connectivity in the brain. They provide a direct measurement of the axonal tracts of the neuron(s) with axons that start or terminate close to the injection site. Rapid developments in tracing techniques have resulted in a large body of literature, in which each publication describes a few experiments with a given protocol, species and brain atlas. For mouse, several large scale studies are under way in which brain-wide connectivity is measured with an automated tracing and analysis workflow. For ethical reasons however, such initiatives are unlikely to be applied to large mammals. The alternative is to rely on databases which combine findings from individual publications. We focus on a prime example of such a database: the CoCoMac collation of Macaque connectivity. Can the very diverse set of experimental protocols and nomenclatures be combined into a reliable wiring diagram of the brain? To study this, we built a new online platform for CoCoMac (cocomac.g-node.org) [1] with two essential components: (1) a custom wiring diagram builder, in which every detail of the included tracing experiment can be specified; and (2) a graphical nomenclature consistency tool, in which conflicting naming schemes can be resolved interactively. In this talk I will show that small errors in relating brain regions across atlases can have a detrimental effect on the resulting wiring diagram. With the consistent nomenclature, we continue to show different views of the connectivity matrix, such as retrograde vs. anterograde, intralateral vs. contralateral and with and without certain types of tracer. Finally we study the variability of the matrix by leaving out random subsets of tracing data. This variability predicts the effect of adding new data to CoCoMac, and indicates which brain regions require special attention.


Supported by the German INCF Node (BMBF grant 01GQ0801), the Helmholtz Association in the Portfolio theme "Supercomputing and Modeling for the Human Brain", JUGENE Grant JINB33, and EU Grant 269921 (BrainScaleS).


1. Bakker R, Wachtler T and Diesmann M (2012) CoCoMac 2.0 and the future of tract-tracing databases. Front. Neuroinform. 6:30. doi: 10.3389/fninf.2012.00030

Keywords: CoCoMac, macaque, axonal projections, literature database, tract tracing, structural connectivity

Conference: Neuroinformatics 2014, Leiden, Netherlands, 25 Aug - 27 Aug, 2014.

Presentation Type: Poster, not to be considered for oral presentation

Topic: Large-scale modeling

Citation: Bakker R, Wachtler T and Diesmann M (2014). Do gold standards remain gold standards when compiling a large number of published tract-tracing studies into a connectivity database?. Front. Neuroinform. Conference Abstract: Neuroinformatics 2014. doi: 10.3389/conf.fninf.2014.18.00072

Received: 27 Apr 2014; Published Online: 04 Jun 2014.

* Correspondence: Dr. Rembrandt Bakker, Radboud University Nijmegen, Nijmegen, Netherlands, r.bakker@donders.ru.nl

© 2007 - 2017 Frontiers Media S.A. All Rights Reserved