Towards reproducible descriptions of neuronal network models

Area of Science:

• Computational neuroscience
• Scientific communication

Background:

• Effective scientific progress relies on precise idea formulation and exchange.
• Over 50 years of neuronal network simulations have not yielded common practices for describing computational models in neuroscience publications.

Purpose of the Study:

• To address the lack of standardized practices in describing neuronal network models.
• To improve the critical evaluation and re-use of computational models in neuroscience.

Main Methods:

• Analysis of 14 research papers on neuronal network models of varying complexity.
• Observation of diverse approaches in model descriptions, graphical representations, and equation notations.

Main Results:

• Significant variation found in how neuronal network models are described across publications.
• Inconsistent organization, placement of material, graphical representations, and notation were observed.

Conclusions:

• A proposed 'good model description practice' includes guidelines for publications, checklists, table templates, and diagram guidelines.
• This initiative aims to foster debate on human-comprehensible model communication, promoting deeper insights into brain function.