scbirlab/cyc-pep-6-12mer-650M-2024

数据集名为Cyclic peptides (650 million)，包含约6.5亿个几乎唯一的环状和线性肽，每个肽由6到12个氨基酸组成，并包含一些计算得出的分子属性。数据集由Eachan Johnson策划，由The Francis Crick Institute资助，使用MIT许可证。数据集的主要用途是开发围绕环状肽的化学模型。数据集的结构包括线性肽和环状肽的标识符、氨基酸序列、SMILES字符串、分子量等属性。数据集的创建是为了开发一个大型的SMILES字符串肽数据集，数据生成使用了schemist工具。

The dataset contains approximately 650 million (almost) unique cyclic and linear peptides, each comprising 6-12 amino acids, with some calculated molecular properties. The dataset was curated by [@eachanjohnson](https://huggingface.co/eachanjohnson) and funded by The Francis Crick Institute. The purpose of creating the dataset is to develop chemistry models around cyclic peptides, and the data was generated using the chemist tool. The dataset structure includes various identifiers and molecular properties for both linear and cyclic peptides.