A context-free grammar for Caenorhabditis elegans behavior
Gupta, S.; Gomez-Marin, A.
Show abstract
Hierarchy is a candidate organizing principle of ethology, where actions grouped into higher order chunks combine in specific ways to generate adaptive behavior. However, demonstrations of hierarchical organization in behavior have been scarce. Moreover, it remains unclear how such underlying organization allows for behavioral flexibility. Here we uncover the hierarchical and flexible nature of Caenorhabditis elegans behavior. By describing worm locomotion as a sequence of discrete postural templates, we identified chunks containing mutually substitutable postures along the dynamics. We then elucidated the rules governing their interactions. We found that stereotypical roaming can be described by a specific sequence of postural chunks, which exhibit flexibility at the lowest postural level. The same chunks get combined differently to produce dwelling, capturing non-stereotypical actions across timescales. We show that worm foraging is organized hierarchically --a feature not explainable via Markovian dynamics--, and derive a context-free grammar governing its behavior --which is different than a regular grammar, or a hidden Markov chain. In sum, in making the analogy with human language concrete (but not literal) our work demonstrates, in line with the foundational insights of classical ethologists, that spontaneous behavior is orderly flexible. Once more, investigating the humble nematode suggests that everything human has its roots in lower animal behavior.\n\n\n\nO_FIG O_LINKSMALLFIG WIDTH=180 HEIGHT=200 SRC=\"FIGDIR/small/708891v1_ufig1.gif\" ALT=\"Figure 1\">\nView larger version (25K):\norg.highwire.dtl.DTLVardef@168e648org.highwire.dtl.DTLVardef@1e6b2eeorg.highwire.dtl.DTLVardef@ed3ecdorg.highwire.dtl.DTLVardef@1b60741_HPS_FORMAT_FIGEXP M_FIG Graphical abstract\n\nC_FIG
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Neuro-Cognitive Multilevel Causal Modeling: A Framework that Bridges the Explanatory Gap between Neuronal Activity and Cognition 96%
- The hourglass organization of the Caenorhabditis elegans connectome 95%
- Learning compositional sequences with multiple time scales through a hierarchical network of spiking neurons 95%
Similar papers in this journal
"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.