• Paper Digest

HBP scientists propose guidelines for describing network connectivity

15 September 2022

Researchers who work with neuronal network models – simplified representations of brains – need to "speak the same language" so that their results can be understood and reproduced. Scientists at the Human Brain Project now propose guidelines for the unambiguous description of network connectivity by formalizing concepts already in use in the computational neuroscience community. To provide an intuitive understanding of network properties, they also propose a graphical notation for network diagrams unifying existing diagram styles. The researchers found that published descriptions of network connectivity are often incomplete and imprecise. This may in turn lead to wrong predictions of network activity when attempting to reproduce model results, as even small differences between rules may lead to different network dynamics.

Authors Prof. Sacha van Albada and Dr. Johanna Senk. Photo left: © Mareen Fischinger Photo right: © Forschungszentrum Jülich GmbH / SBC Lehmann

“Precise definitions of connectivity rules are crucial for the correct and efficient algorithmic implementation of high-level connectivity routines in simulators,” said neuroscientist Johanna Senk, the lead author of the study. Senk added that the researchers hope that the guidelines they propose will increase reproducibility, because they will ensure that when two 
neuroscientists use the same term or symbol they mean the same thing.

Most neuronal network models that have been developed until now still have limited complexy, noted Prof. Sacha van Albada, the senior author of the study. She pointed out that the new EBRAINS digital research infrastructure built by the HBP will make complex models manageable for individual scientists and in this way help to overcome the current  complexity barrier.

Original Publication:

Senk J, Kriener B, Djurfeldt M, Voges N, Jiang H-J, Schüttler L, et al. (2022) Connectivity concepts in neuronal network modeling. PLoS Comput Biol 18(9): e1010086. https://doi.org/10.1371/journal.pcbi.1010086