# Scorpion Toxin Binder Design: SPR-Validated De Novo Protein Binders

## Facts

- Provider: Anthropic (Claude Science; Horwitz, van Geelen, Kajderowicz, Shanehsazzadeh)
- License: CC-BY-4.0 (https://huggingface.co/datasets/Anthropic/scorpion-toxin-binder-design/raw/main/README.md)
- Size: 5,108 files
- Format: CSV, MP4, JPEG (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/Anthropic/scorpion-toxin-binder-design
- Access: public, readable without an account
- Category: Structural Biology
- Source: https://www-cdn.anthropic.com/31ac578806cffd8b3d92d8affe6edc4c04d502e1.pdf
- Page: https://scigantic.com/archive/9fdc2833-af91-577b-9e14-46c9c44d9c4f
- Keywords: Structural Biology, CSV, MP4, JPEG, Hugging Face

## About

Data release for de novo protein binder design against five scorpion toxins (Cn2, Ts1, AaH II, AaH I and LqhIII), published by Anthropic with the report De novo design of protein binders against scorpion toxins with Claude (October 2026). Claude designed the proteins; two contract research organizations, Adaptyv Bio and Twist Bioscience, measured binding by surface plasmon resonance (SPR). design.csv holds all 211 designs with sequence, each lab's result and KD against each toxin, and a binder_against label; seventeen designs are binders (seven against Cn2, four against AaH II, four against AaH I, two against LqhIII, none against Ts1). Further tables give in silico scores, a Cn2 competition assay against the neutralizing antibodies LR, RU1 and 9004G, a baculovirus particle (BVP) ELISA for polyreactivity, the toxin sequences and control antibodies. sensorgrams/ has per-design SPR traces and kinetic fits, structures/ has design models and predicted structures, and supporting_tables/ documents how each design was made and expressed; COLUMNS.md explains every column. Measured 2026-11-14: CSV tables at the root plus the assets/, sensorgrams/, structures/ and supporting_tables/ folders. Open the tables with pandas (read uuid as text) or datasets.load_dataset. Useful for benchmarking binder design pipelines, structure prediction confidence versus wet-lab outcome, and antitoxin research. Licensed CC BY 4.0; cite the Anthropic report.

## Load the data

```python
import pandas as pd
df = pd.read_csv("hf://datasets/Anthropic/scorpion-toxin-binder-design/BVP_ELISA.csv")
df.head()
```

## Data structure

File types in sampled files: CSV, MP4, JPEG

Example file: `BVP_ELISA.csv`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/9fdc2833-af91-577b-9e14-46c9c44d9c4f
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/9fdc2833-af91-577b-9e14-46c9c44d9c4f
- Over MCP: `get_schema_card` with id `9fdc2833-af91-577b-9e14-46c9c44d9c4f` at https://api.scigantic.com/mcp

## Related topics

- [CSV datasets](https://scigantic.com/topics/csv)
- [JPEG datasets](https://scigantic.com/topics/jpeg)
- [Structural Biology datasets](https://scigantic.com/topics/structural-biology)
- [MP4 datasets](https://scigantic.com/topics/mp4)
