MoSeq

By analyzing spontaneous behavior, without pre-defining which actions should be measured, MoSeq provides a deeper understanding of behavioral phenotypes while maintaining scientific interpretability through visualization and interrogation of the individual behaviors driving observed effects.

Detail of offering

Understanding behavior through motion sequencing

MoSeq analyzes video recordings of freely moving rodents and automatically identifies recurring behavioral motifs, referred to as “syllables”. Rather than focusing on a limited number of predefined measures such as rearing or locomotor activity, the platform evaluates the full behavioral repertoire expressed during a recording session and organizes behavior into distinct, quantifiable modules.

This approach enables researchers to:

• Characterize behavioral phenotypes in disease models
• Detect treatment-related changes in spontaneous behavior
• Compare behavioral signatures across pharmacological classes
• Profile novel compounds against established reference datasets
• Explore behavioral outcomes alongside other translational endpoints in future integrated study designs

Interpretable machine learning

A key advantage of MoSeq is that behavioral outputs can be visualized and interrogated. Rather than generating an opaque classification, specific behaviors contributing to differences between groups are presented in video clips. These visualizations overlay multiple examples of a behavioral syllable, enabling rapid interpretation of patterns such as grooming, rearing, locomotor activity, wall investigation, or other spontaneous behaviors.

There is only ever one animal in the arena at a time.

High-throughput behavioral phenotyping

Transpharmation operates a high-throughput recording configuration capable of assessing eight animals simultaneously, allowing efficient collection of behavioral datasets across treatment groups. Recording sessions are analyzed centrally, with behavioral outputs generated from video datasets and summarized through a range of behavioral metrics.

In addition to syllable-based outputs, the platform captures conventional behavioral measures including:

• Velocity
• Movement paths
• Dwell time within arena regions

Comparing compounds and disease phenotypes

By integrating information across multiple behavioral syllables, MoSeq can be used to compare pharmacological profiles and examine similarities or differences between treatments, disease states, and control populations. Advanced visualization approaches help researchers understand whether a test article produces behavioral changes that resemble known pharmacological classes or modifies disease-associated behavioral phenotypes.

Applications currently include the investigation of neurological disease models, such as epilepsy-related phenotypes, alongside the ongoing development of an internal reference database spanning therapeutic areas including pain, epilepsy, depression, cognition, and neuropsychiatric pharmacology.

Decision support for discovery programs

MoSeq adds a rich behavioral dimension to preclinical studies, providing objective, data-driven measures that can complement traditional efficacy endpoints. The platform is particularly valuable where behavioral changes are complex, multifactorial, or difficult to capture using conventional observational methods alone.