What is Cebra?
CEBRA, or Learnable Latent Embeddings for Joint Behavioural and Neural Analysis, is a novel machine-learning method developed with the aim to map behavioural actions to neural activity - a fundamental goal in neuroscience. The tool has been developed to cater to the increasing interest in modeling neural dynamics during adaptive behaviors as our ability to record large-scale neural and behavioural data grows. The method is unique as it can jointly use behavioural and neural data in both a hypothesis-driven and discovery-driven manner to produce high-performance and consistent latent spaces that reveal the underlying correlates of behaviour. It can be leveraged over single and multi-session datasets for hypothesis testing or be used label-free. CEBRA is adept at handling both calcium and electrophysiology datasets, across tasks, whether sensory or motor, and in simple or complex behaviors across species. Notably, CEBRA can be effectively used for space mapping, uncovering complex kinematic features, and rapid, high-accuracy decoding of natural movies from the visual cortex, thus improving our understanding of neural dynamics and behavior. It also excels at decoding activity from the mouse brain's visual cortex to reconstruct a viewed video, highlighting its promise in neuroscience and behavioral studies.
Pros
- Non-linear techniques
- Creates high-performance latent spaces
- Maps behavioural actions to neural activity
- Reveals behaviour correlates
- Enables hypothesis testing
- Aids discovery-driven analysis
- Validated on calcium datasets
- Validated on electrophysiology datasets
- Useful across sensory tasks
- Useful across motor tasks
- Applicable in simple behaviours
- Applicable in complex behaviours
- Useful for species comparisons
- Operates with single session datasets
- Operates with multi-session datasets
- Label-free usage
- Decodes natural movies from visual cortex
- Efficient in space mapping
- Unveils complex kinematic features
- Code available on GitHub
- Quick and accurate decoding
- Reconstructs visual cortex activity
- Distinguishes meaningful differences
- Makers documentation available
- Open source
- Useful for neuroscience researchers
- Fits timeseries data
- Reveals hidden data structures
- Tests hypotheses on large datasets
- Flexible use with behavioural and neural data
- Ability to decode viewed videos
- Applicable to movie frames decoding
- Produces consistent latent spaces
- Validated in adaptive behaviors contexts
- Applicable to rat hippocampus data
- Applicable to mouse primary visual cortex data
- Works with 2-photon and Neuropixels data
- Handles high-variability data
- Feedforward and self-supervised methods
- Assists in behaviour analysis
- Creates neural dynamics map
- Aggregates behavioural and neural data
- Supports joint behavioural and neural data
Cons
- Limited dataset adaptability
- Requires simultaneous neural-behavioral data
- No live data support
- Potentially complex for non-neuroscientists
- Lacks dataset flexibility
- Requires preexisting hypotheses
- Only supports specific tasks
- Possibly high computational needs
- No adaptability for unsupervised learning
Cebra FAQ
What is the main purpose of Cebra?
Cebra's main purpose is to map behavioural actions to neural activity. This machine learning tool is designed to create consistent and high-performance latent spaces using joint behavioural and neural data. Its primary application is to improve the understanding of neural dynamics during adaptive behaviours.
How does Cebra work with behavioural and neural data?
Cebra works with behavioural and neural data jointly, employing non-linear techniques. The tool can map behavioural actions to neural activity, exposing the underlying neural correlates of behaviour. It also generates neural latent embeddings that are useful for both hypothesis testing and discovery-driven analysis.
Can Cebra be used for hypothesis testing and discovery-driven analysis?
Yes. Cebra can be used for both hypothesis testing and discovery-driven analysis. It can process single or multi-session datasets and can be used label-free, providing flexibility in its application.
What kind of datasets can Cebra handle?
Cebra is capable of handling both calcium and electrophysiology datasets. It is also proficient in working across sensory and motor tasks and is applicable in simple or complex behaviours across a variety of species.
Can Cebra be used on single and multi-session datasets?
Yes. Cebra allows for the use of single and multi-session datasets. This functionality offers flexibility in terms of the quantity and types of data that can be processed.
Does Cebra require any labelling for its use?
No. Cebra doesn't necessarily require any labelling for its use. It can be used label-free, making it highly practical in various neuroscience settings and data analysis.
Is Cebra applicable across different species?
Yes. Cebra can function across different species. It is not restricted by species types in its ability to analyze simple or complex behaviours, expanding its utility significantly in behavioural and neural studies.
What are some key tasks Cebra is adept at handling?
Some key tasks Cebra is adept at handling include space mapping, uncovering complex kinematic features, and rapid, high-accuracy decoding of natural movies from the visual cortex. These advanced functions further improve the understanding of neural dynamics and behaviour.
