# Rat Clusterless Spike Recordings With Medial Septal Stimulation for C3PO Latent-State Learning, Frank Lab (DANDI 002044)

## Facts

- Provider: Samuel Bray and the Frank lab
- License: CC-BY-4.0 (https://api.dandiarchive.org/api/dandisets/002044/versions/draft/)
- Stored at: Amazon S3, s3://dandiarchive/dandisets/002044 (us-east-2)
- Access: public, readable without an account
- Category: Neuroscience
- Source: https://github.com/samuelbray32/c3po
- Page: https://scigantic.com/archive/1ddc0ca7-3215-5048-b0a3-4bb5ce697253
- Keywords: Neuroscience, electrophysiology, time series, AWS Open Data

## About

Previously unpublished extracellular recordings from Loren Frank's lab used in Bray et al., "Learning predictive latent structure from unclustered neural spike trains", which introduces C3PO (Contrastive Prediction of Point Process Observations), an unsupervised model that infers latent neural states from spike times and waveform features without spike sorting. The Dandiset holds 256 NWB files (1.16 TB) from two rats (Wallie, 200 files; wilbur, 56 files), each from a single recording day. Files include raw multi-electrode ElectricalSeries, LFP, sorted Units, position and head direction (SpatialSeries, CompassDirection), behavioral events and behavioral time series, plus many per-analysis object files (..._obj- .nwb) exported from the Spyglass pipeline; the main Wallie session is described as medial septal stimulation. The companion repositories samuelbray32/c3po (pip install c3po-neuro, with c3po.utils.train_from_dandi) and samuelbray32/c3po_publication train models directly from these files. Licensed CC BY 4.0; cite Bray et al. and DANDI 002044.

## Load the data

```python
import boto3
from botocore import UNSIGNED
from botocore.client import Config
s3 = boto3.client('s3', region_name='us-east-2', config=Config(signature_version=UNSIGNED))
s3.list_objects_v2(Bucket='dandiarchive', Prefix='dandisets/002044/', Delimiter='/')
```

Lists the files with anonymous access. Runs on any machine, with no cloud account.

## Data structure

Example file: `dandisets/002044/draft/assets.jsonld`

Path pattern: `{prefix}/{YYYYMM}/{seg2}/{filename}`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/1ddc0ca7-3215-5048-b0a3-4bb5ce697253
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/1ddc0ca7-3215-5048-b0a3-4bb5ce697253
- Over MCP: `get_schema_card` with id `1ddc0ca7-3215-5048-b0a3-4bb5ce697253` at https://api.scigantic.com/mcp

## Related topics

- [Neuroscience datasets](https://scigantic.com/topics/neuroscience)
- [Time series datasets](https://scigantic.com/topics/time-series)
- [Electrophysiology datasets](https://scigantic.com/topics/electrophysiology)
