遇见数据集

TranSpec

收藏
DataCite Commons2025-04-01 更新2024-11-05 收录
官方服务:

资源简介:

## Details of the Training-Data fileIt includes the SMILES of all the molecules in QM9S, as well as the files of infrared and Raman spectra, where the first column is SMILES and the first line is the frequency value of 3000 points from 0-4000cm^-1. The other columns are the spectral intensity values corresponding to SMILES.Includes SMILES and IR spectra files from QM14S, where SMILES is not present in the spectra file, and the lines in the two files correspond to each other.<br>## Details of the Results fileIt includes the correct SMILES of the QM9S datasets, the SMILES candidate results obtained by different models, and the parameters of the corresponding models. Finally, it also provides the SMILES candidate results after model fusion.<br>## Details of Isomer-Distinguish fileIt contains candidate results for the QM9S test sets of isomers, where the first column is the correct SMILES, and columns 2 through 11 are the predicted Top1 to Top10 SMILES, with the isomers in adjacent rows and an empty row inserted in each of the different groups.<br>## Details of the Group-Recognition fileIt contains the correct SMILES graphs for the 12 functional groups tested in the QM9S datasets, as well as the Top1 predicted results.

训练数据文件详情:本文件包含QM9S数据集内所有分子的简化分子线性输入规范(SMILES),以及红外光谱与拉曼光谱文件。其中首列为SMILES,首行为0~4000cm⁻¹范围内3000个采样点的波数值,其余各列为对应SMILES的光谱强度值。本数据集同时包含QM14S的SMILES与红外光谱(IR)文件,其中光谱文件未包含SMILES,且两个文件的各行一一对应。 ## 结果文件详情:本文件包含QM9S数据集的标准SMILES、不同模型生成的SMILES候选结果,以及对应模型的参数信息,最终还提供了模型融合后的SMILES候选结果。 ## 异构体区分(Isomer-Distinguish)文件详情:本文件包含QM9S异构体测试集的候选结果,其中首列为标准SMILES,第2至11列为预测得到的Top1至Top10 SMILES。异构体排列于相邻行中,且每个不同异构体组之间均插入空行进行分隔。 ## 基团识别(Group-Recognition)文件详情:本文件包含QM9S数据集中12种测试官能团的标准SMILES结构,以及对应的Top1预测结果。

提供机构:
figshare
创建时间:
2024-09-30
搜集汇总
数据集介绍
TranSpec 数据集图片
以上内容由遇见数据集搜集并总结生成
二维码
社区交流群
二维码
科研交流群
商业服务