# SBI for Connectomics: wiring-rule simulations and posteriors for rat barrel cortex connectivity models

## Facts

- Provider: Macke Lab (mackelab)
- License: MIT (https://huggingface.co/datasets/mackelab/sbi-for-connectomics-data)
- Size: 221 files
- Format: TSV (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/mackelab/sbi-for-connectomics-data
- Access: public, readable without an account
- Category: Neuroscience
- Source: https://www.biorxiv.org/content/10.1101/2023.01.31.526269v1
- Page: https://scigantic.com/archive/37c7a88d-abe6-50f8-bc94-bc506ee30b20
- Keywords: Neuroscience, connectomics, TSV, Hugging Face

## About

Companion data for "Simulation-based inference for efficient identification of generative models in connectomics" from the Macke lab. The paper uses simulation-based inference (SBI, neural posterior estimation) to infer the parameters of computational wiring-rule models of the rat barrel cortex, a dense structural model of somatosensory cortex used in connectomics. About 2 to 3 GB across 219 files in three folders that mirror the paths the code expects: data/ (masks and feature tables for the structural barrel cortex model, the subcellular model and the cube model, plus presimulated prior and posterior samples for several wiring-rule parameterizations), results/ (amortized NPE posteriors, prior and posterior predictive samples, simulation-based calibration and cross-validation results, tutorial posteriors) and figures/ (source data and renders for the main and supplementary figures). Pickles are torch/sbi objects loaded with the consbi package and notebooks in github.com/mackelab/sbi-for-connectomics; tables are .txt/.tsv. Download into a clone of that repository with hf download. Released under the MIT License, matching the code repository; cite the bioRxiv preprint (2023.01.31.526269).

## Load the data

```python
import pandas as pd
df = pd.read_csv("hf://datasets/mackelab/sbi-for-connectomics-data/data/cube_model/features/features_set-6_float.tsv", sep='\t')
df.head()
```

## Data structure

File types in sampled files: TSV

Example file: `data/cube_model/features/features_set-6_float.tsv`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/37c7a88d-abe6-50f8-bc94-bc506ee30b20
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/37c7a88d-abe6-50f8-bc94-bc506ee30b20
- Over MCP: `get_schema_card` with id `37c7a88d-abe6-50f8-bc94-bc506ee30b20` at https://api.scigantic.com/mcp

## Related topics

- [Neuroscience datasets](https://scigantic.com/topics/neuroscience)
- [TSV datasets](https://scigantic.com/topics/tsv)
- [Connectomics datasets](https://scigantic.com/topics/connectomics)
