afg1/rnacentral_subset

--- license: cc0-1.0 size_categories: - 1M<n<10M --- This is a parquet-ified subset of the RNAcentral active fasta file available here: https://ftp.ebi.ac.uk/pub/databases/RNAcentral/releases/24.0/sequences/rnacentral_active.fasta.gz I have preprocessed it a bit, requiring only sequences less than 8192 nt long, and having no ambiguous nucleotides (i.e. no Ns or other non standard things) This dataset is about 10% of the overall, and comprises 3,252,483 (3.2 million) sequences, or 2,642,703,990 (2.6 billion) bases. The train/val/test split is 60/20/20

原始信息汇总

数据集概述

数据集基本信息

• 许可证: CC0-1.0
• 数据集大小: 1M<n<10M

数据来源与处理

• 原始数据: 来自RNAcentral的active fasta文件，地址为https://ftp.ebi.ac.uk/pub/databases/RNAcentral/releases/24.0/sequences/rnacentral_active.fasta.gz
• 预处理: 仅包含长度小于8192 nt且无模糊核苷酸（无N或其他非标准字符）的序列

数据集规模

• 序列数量: 3,252,483 (3.2 million)
• 总碱基数: 2,642,703,990 (2.6 billion)

数据集分割

• 训练/验证/测试比例: 60/20/20