natelgrw/ReTiNA

ReTiNA是一个大型的开源数据集，用于训练机器学习模型预测LC-MS工作流程中小分子的保留时间。该数据集不断扩展，包含来自MIT的Coley研究组的新的实验保留时间值。数据集包括641,651个独特化合物的`.smi`列表和6种独特溶剂，用于化学描述符计算。数据集可用于估算新化合物-环境组合的保留时间，辅助LC-MS方法开发中的峰分配，以及特定条件下保留时间预测的机器学习模型训练。

ReTiNA is a large open-source dataset for training machine learning models to predict small molecule retention times in LC-MS workflows. The dataset is actively expanding with new experimental retention time values from the Coley Research Group at MIT. It includes lists of 641,651 unique compounds and 6 unique solvents in the `.smi` format for chemical descriptor calculations. ReTiNA is designed for estimating retention times for new compound-environment combinations, aiding in peak assignment during LC-MS method development, and training ML models for retention time prediction under specific conditions.