five

pzweuj/EVOScore-2

收藏
Hugging Face2026-01-28 更新2026-03-29 收录
下载链接:
https://hf-mirror.com/datasets/pzweuj/EVOScore-2
下载链接
链接失效反馈
官方服务:
资源简介:
--- language: - en tags: - biology - ESM - language-model - protein - VEP pretty_name: EVOScore-2 size_categories: - 100M<n<1B license: mit --- # Proteome-wide VESM-3B Variant Effect Scores (MIT Clean Mirror) ![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg) ![Commercial Status](https://img.shields.io/badge/Status-Commercial--Safe-brightgreen) ## ⚠️ Compliance & Mirror Notice This repository is a **strict subset** of the original [ntranoslab/vesm_scores](https://huggingface.co/datasets/ntranoslab/vesm_scores). To ensure commercial compliance and safety, all non-commercial components (VESM3, Sequence-only VESM3, and VESM++) have been **permanently removed** from this mirror. **This repository contains ONLY the `VESM_3B` predictions**, which are based on the ESM2-3B architecture and explicitly released under the MIT License. --- ## Dataset Description This repository provides precomputed **proteome-wide (UniProtKB, hg19, and hg38) variant-effect prediction scores** using the **VESM_3B** model. ### Included Model * **VESM_3B:** A protein language model based on **ESM2 (3B parameter)** architecture. *(Note: The original VESM3 and VESM++ models based on ESM3 have been excluded from this repository).* For more scientific details, please refer to the original GitHub repo: https://github.com/ntranoslab/vesm ## Files Included * `hg38_VESM_3B_scores.parquet.gzip` * `hg19_VESM_3B_scores.parquet.gzip` * `hg38_VESM_3B_scores.vcf.gz` * `hg19_VESM_3B_scores.vcf.gz` * *(And any other Uniprot/metadata files strictly related to VESM_3B)* ## Threshold Recommend | P_threshold | <-11.6 | | --------------- | ----------- | | **B_threshold** | **>-8.5** | ## License <a name="license"></a> The data provided in this repository is distributed under the **MIT License**. > **Attribution:** > This dataset is a mirror of the VESM_3B predictions generated by Ntranos Lab. > While the original repository contained mixed licenses, the content hosted here is exclusively sourced from the MIT-licensed portion.
提供机构:
pzweuj
5,000+
优质数据集
54 个
任务类型
进入经典数据集
二维码
社区交流群

面向社区/商业的数据集话题

二维码
科研交流群

面向高校/科研机构的开源数据集话题

数据驱动未来

携手共赢发展

商业合作