The choice-wide behavioral association study: data-driven identification of interpretable behavioral components

Area of Science:

• Behavioral neuroscience
• Computational biology
• Machine learning

Background:

• Identifying complex behavioral structures across multiple timescales is challenging.
• Existing methods struggle to pinpoint relevant components for long-term learning and decision-making.

Approach:

• We introduce the choice-wide behavioral association study (CBAS), a data-driven method inspired by genome-wide association studies.
• CBAS employs resampling-based multiple comparisons correction to detect significant action or choice sequences.
• This approach identifies features that differ between groups or correlate with specific covariates.

Key Points:

• CBAS systematically uncovers interpretable behavioral features.
• The method is effective across diverse species, including flies, rats, and humans.
• It provides valuable insights into behavioral tasks at various timescales.

Conclusions:

• CBAS offers a robust framework for dissecting complex behaviors.
• This data-driven approach enhances our understanding of behavioral structure and function.
• The method has broad applicability in analyzing behavioral data.