Back

Social Context Restructures Behavioral Syntax in Mice

Ritter, M.; Shipley, H.; Deiana, S.; Hengerer, B.; Wotjak, C. T.; Brecht, M.; Bogadhi, A. R.

2025-04-19 animal behavior and cognition
10.1101/2025.04.17.648924 bioRxiv
Show abstract

The study of social behavior in mice has grown increasingly relevant for unraveling associated brain circuits and advancing the development of treatments for psychiatric symptoms involving social withdrawal or social anxiety. However, a data-driven understanding of behavior and its modulation in solitary and social contexts is lacking. In this study, we employed motion sequencing ("MoSeq") to decompose mouse behaviors into discrete units ("Syllables") and investigate whether--and how--the behavioral repertoire differs between solitary and dyadic (social) settings. Our results reveal that social context significantly modulates a minority (25%) of syllables, containing predominantly stationary and undirected behaviors. Notably, these changes are associated with spatial proximity to another mouse rather than active social contact. Interestingly, a network analysis of syllable transitions shows that context-sensitive syllables exhibit altered network influence, independent of the number of connected syllables, suggesting a regulatory role. Furthermore, syllable composition changes significantly during social contact events with two distinct sequence families governing approach and withdrawal behaviors. However, no unique syllable sequences mapped to specific social interactions. Overall, our findings suggest that a subset of syllables drives contextual behavioral adaptation in mice, potentially facilitating transitions within the broader behavioral repertoire. This highlights the utility of MoSeq in dissecting nuanced, context-dependent behavioral dynamics.

Published in Frontiers in Behavioral Neuroscience (predicted rank #11) · training set

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.