datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
esm-if-cas13-stage0003-assetsinverse-folding-cas13
Inverse-Folding Cas13 reproduction assets
This dataset repository transports the small, versioned data/raw/ and
data/processed/ assets required to reproduce the computational DjCas13d
research workflow in
XuJP264/Inverse-Folding-Cas13.
It contains deposited structural records, their source metadata and SHA256
receipts, a bounded NCBI RefSeq Cas13d snapshot, cited literature artifacts, and
the derived bound evolutionary prior. Per-file source and usage information is
stored… See the full description on the dataset page: https://huggingface.co/datasets/metaXu264/inverse-folding-cas13.crispr-cas-atlas-generator
CRISPR-Cas Atlas – GENERator-ready Dataset
This dataset is a derived, preprocessed version of the CRISPR-Cas Atlas, formatted for causal language model fine-tuning with GENERator.
Source
Original dataset:CRISPR-Cas Atlas v1.0
Ruffolo et al., Design of highly functional genome editors by modeling the universe of CRISPR-Cas sequences, bioRxiv (2024)https://www.biorxiv.org/content/10.1101/2024.04.22.590591v1
Processing
Each CRISPR-Cas operon was converted into a… See the full description on the dataset page: https://huggingface.co/datasets/metaXu264/crispr-cas-atlas-generator.generator-cas-ft-dataset-traingenerator-cas-ft-dataset-mini-traingenerator-cas-ft-dataset-validgenerator-cas-ft-dataset
GENERator Cas Fine-tuning Dataset
This dataset is designed for fine-tuning GENERator on CRISPR-Cas–related DNA sequences.
File
generator_cas_ft_dataset.jsonl
Format
Each line is a JSON object, typically containing:
input : DNA sequence (length multiple of 6 for 6-mer tokenizer)
output : DNA sequence (autoregressive target)
role : cas_protein / crispr / tracrRNA / spacer (if present)
source : CRISPR-Cas Atlas–derived
Usage
from datasets… See the full description on the dataset page: https://huggingface.co/datasets/metaXu264/generator-cas-ft-dataset.generator-cas-ft-trainRISPR-Cas – Train Split
This repository contains the training split of the CRISPR-Cas
DNA generation dataset used for fine-tuning GENERator-style models.
Format
Each line is a JSON object:
{
"sequence": "<cas9>ATG..."
}
crispr-extracted-sequencesgenerator-cas-ft-validgenerator-cas-ft-testgenerator-cas-ft-dataset-test