Data for "DirectMultiStep: Direct Route Generation for Multistep Retrosynthesis"

buyables-stock.txt: 329,384 canonicalized SMILES strings representing buyable building blocks from eMolecules, Sigma-Aldrich, Mcule, ChemBridge Hit2Lead, and WuXi LabNetwork (collected and processed from this repository).chembl_targets.json: 5,000 canonicalized SMILES strings representing 5,000 randomly selected ChEMBL compounds, collected and processed from this Figshare dataset.eMolecules.txt: 23,081,629 canonicalized SMILES strings representing eMolecules screening compounds available as of 2019 (2019-11-01), collected and processed from this repository.uspto_190.txt: 190 canonicalized SMILES strings representing USPTO compounds, collected and processed from this repository.

buyables-stock.txt：包含329,384条规范化简化分子线性输入规范（SMILES）字符串，代表来自eMolecules、Sigma-Aldrich、Mcule、ChemBridge Hit2Lead以及WuXi LabNetwork的可购买合成砌块，从本仓库收集并处理得到。chembl_targets.json：包含5,000条规范化SMILES字符串，代表5000个随机选取的ChEMBL化合物，从该Figshare数据集收集并处理得到。eMolecules.txt：包含23,081,629条规范化SMILES字符串，代表截至2019年11月1日（2019-11-01）可获取的eMolecules筛选用化合物，从本仓库收集并处理得到。uspto_190.txt：包含190条规范化SMILES字符串，代表USPTO化合物，从本仓库收集并处理得到。