{"success":true,"data":{"author":"Marks Lab, Harvard Medical School","bucketAccessType":"public","bucketName":"Marks-lab/PromoterZoo","bucketPrefix":"","category":"Genomics & Bioinformatics","cloudProvider":"huggingface","createdAt":"2026-11-21T20:45:58.846Z","description":"Training data for LOL-EVE (Language of Life, Evolutionary Variant Effects), a genomic language model from the Marks Lab at Harvard Medical School that scores the effect of indels and other variants in gene promoters. The dataset holds 13,611,373 promoter DNA sequences of 1,000 base pairs taken around transcription start sites (TSS) across many species grouped into evolutionary clades, giving the model evolutionary context for each gene's regulatory region. Each record has the sequence, gene symbol, species, clade, chromosome, strand, genomic start and stop, sequence length and direction (towards or away from the TSS), and both strands are included. Stored as one 6.4 GB data.parquet plus Arrow shards under dataset/, it loads with the datasets library or pandas and can be filtered by gene or species. Suited to pretraining DNA language models, comparative promoter analysis and cis-regulatory evolution studies. Released by the Marks Lab under the MIT License.","displayTitle":"PromoterZoo: 13.6 Million Multi-Species Promoter Sequences for Training the LOL-EVE Variant Effect Model (Harvard Marks Lab)","huggingFaceUrl":"https://huggingface.co/datasets/Marks-lab/PromoterZoo","id":"74b31416-c2c6-5fc7-ab54-4ae1c287ddcb","license":"MIT","licenseEvidenceUrl":"https://huggingface.co/datasets/Marks-lab/PromoterZoo","objectCountEstimate":13,"probeStatus":"ok","referenceUrl":"https://huggingface.co/datasets/Marks-lab/PromoterZoo","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"Marks-lab/PromoterZoo\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"Marks-lab/PromoterZoo\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=Marks-lab, license:mit, size_categories:1M<n<10M, format:parquet, downloads=62"],"format":"mixed","generatedAt":"2026-11-21T20:48:48.294Z","layout":{"example":"data.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"504152311504159499f7011594d7724c15c61f1500120000cacc7b78e8030000434341434747434143414743414754545454475443544343414747050510544754474109100c41434143012c205443434754544741470133052900470101010a053c0054050b0454470520004701140441470123105447475454051e185447414354544701220841434101090054016f01281043414741430106044343016901740054012301700953014b084741410137004101a60041018c051b0116059701810043015b01620112059c0447540179104141434141012a01140454540507054e005409f101990054018200410112005405b20148058710545441545401990145044343051d01d7213b055d04434101ae0522214c015e0118059b01710047012b044147215701ea2d090843434101f501770101098a04414105830047211001410164013b01980108084741542500004321b00054218c0047099201432115011a21190910054f011304434701cb004305662169005421d72de200470538019b00432596004301d00131053401612161410525d305e00160052d054a454000410111210621ff01fc0510214304434109aa0047019001f7210b254e216a012725940199014004544721af097d0041011201d001f001882116051121d101b4098201c30194017c21ad291a01da2566458241f6017b01710104015401a801e8058f","encoding":"hex","key":"data.parquet","truncatedAt":512}],"sampleObjects":["data.parquet","dataset/data-00000-of-00007.arrow","dataset/data-00001-of-00007.arrow","dataset/data-00002-of-00007.arrow","dataset/data-00003-of-00007.arrow"],"sidecars":[{"bytes":"---\nlanguage:\n- en\nlicense: mit\ntask_categories:\n- text-generation\n- genomic-sequence-modeling\nsize_categories:\n- 10M<n<100M\nconfigs:\n- config_name: default\n  data_files: \"*.parquet\"\n  default: true\n---\n\n# PromoterZoo\n\n## Dataset Description\n\nThis dataset contains promoter sequence training data for the LOL-EVE (Language of Life - Evolutionary Variant Effects) model. The dataset includes 13.6 million promoter sequences from various species, providing a comprehensive foundation for training genomic sequence models.\n\n## Dataset Structure\n\n- **Total Sequences**: 13,611,373\n- **Species**: Multiple species across different clades\n- **Sequence Length**: 1,000 base pairs (promoter regions)\n- **Direction**: Both towards and away from TSS\n\n## Features\n\n- `sequence`: DNA sequence (1000bp promoter region)\n- `gene`: Gene symbol\n- `species`: Species name\n- `clade`: Evolutionary clade\n- `chrom`: Chromosome\n- `strand`: Strand orientation (+/-)\n- `start`: Genomic start position\n- `stop`: Genomic stop position\n- `sequence_len`: Length of sequence (1000)\n- `direction`: Direction relative to TSS (towards/away)\n\n## Usage\n\n```python\nfrom datasets import load_dataset\n\n# Load the dataset\ndataset = load_dataset(\"cshearer/PromoterZoo\")\n\n# Access the data\nprint(f\"Dataset size: {len(dataset['train'])}\")\nprint(f\"Features: {dataset['train'].features}\")\n\n# Example: Get sequences for a specific gene\ngene_sequences = dataset['train'].filter(lambda x: x['gene'] == 'a1bg')\nprint(f\"Sequences for a1bg: {len(gene_sequences)}\")\n\n# Example: Get sequences from a specific species\nspecies_sequences = dataset['train'].filter(lambda x: 'homo' in x['species'].lower())\nprint(f\"Human sequences: {len(species_sequences)}\")\n```\n\n## Data Processing\n\nThe sequences are processed as follows:\n- Promoter regions are extracted 1000bp upstream of transcription start sites\n- Both forward and reverse strand sequences are included\n- Sequences are labeled with gene, species, and genomic coordinates\n\n## Citation\n\nIf you use this dataset in your research, please cite:\n\n```bibtex\n@article{loleve2024,\n  title={LOL-EVE: Language of Life - Evolutionary Variant Effects},\n  author={Your Name and Collaborators},\n  journal={Nature},\n  year={2024}\n}\n```\n\n## License\n\nThis dataset is released under the MIT License.\n","encoding":"utf8","key":"README.md","name":"README.md","size":2282,"truncated":false},{"bytes":"{\n  \"citation\": \"\",\n  \"description\": \"\\n    Training data for LOL-EVE model containing promoter sequences from multiple species.\\n    This dataset provides the foundation for training genomic sequence models to predict\\n    variant effects in promoter regions.\\n    \",\n  \"features\": {\n    \"sequence\": {\n      \"dtype\": \"string\",\n      \"_type\": \"Value\"\n    },\n    \"gene\": {\n      \"dtype\": \"string\",\n      \"_type\": \"Value\"\n    },\n    \"species\": {\n      \"dtype\": \"string\",\n      \"_type\": \"Value\"\n    },\n    \"clade\": {\n      \"dtype\": \"string\",\n      \"_type\": \"Value\"\n    },\n    \"chrom\": {\n      \"dtype\": \"string\",\n      \"_type\": \"Value\"\n    },\n    \"strand\": {\n      \"dtype\": \"string\",\n      \"_type\": \"Value\"\n    },\n    \"start\": {\n      \"dtype\": \"int64\",\n      \"_type\": \"Value\"\n    },\n    \"stop\": {\n      \"dtype\": \"int64\",\n      \"_type\": \"Value\"\n    },\n    \"sequence_len\": {\n      \"dtype\": \"int64\",\n      \"_type\": \"Value\"\n    },\n    \"direction\": {\n      \"dtype\": \"string\",\n      \"_type\": \"Value\"\n    }\n  },\n  \"homepage\": \"https://huggingface.co/cshearer/PromoterZoo\",\n  \"license\": \"mit\"\n}","encoding":"utf8","key":"dataset/dataset_info.json","name":"dataset_info.json","size":1082,"truncated":false},{"bytes":"{\n  \"dataset_name\": \"PromoterZoo\",\n  \"description\": \"Complete training data for LOL-EVE model containing 13.6M promoter sequences\",\n  \"homepage\": \"https://huggingface.co/datasets/cshearer/PromoterZoo\",\n  \"license\": \"mit\",\n  \"size\": 13611373,\n  \"features\": [\n    \"sequence\",\n    \"gene\",\n    \"species\",\n    \"clade\",\n    \"chrom\",\n    \"strand\",\n    \"start\",\n    \"stop\",\n    \"sequence_len\",\n    \"direction\"\n  ]\n}","encoding":"utf8","key":"dataset_info.json","name":"dataset_info.json","size":407,"truncated":false}],"sourceObjectKey":"data.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"Marks-lab/PromoterZoo\")\nprint(ds)","starterSamplePath":"data.parquet","totalObjects":13},"summary":"Training data for LOL-EVE (Language of Life, Evolutionary Variant Effects), a genomic language model from the Marks Lab at Harvard Medical School that scores the effect of indels…","title":"PromoterZoo: 13.6 Million Multi-Species Promoter Sequences for Training the LOL-EVE Variant Effect Model (Harvard Marks Lab)","updatedAt":"2026-11-22T01:00:05.576Z","visibility":"public","websiteUrl":"https://github.com/debbiemarkslab"}}