Find research datasets worth reusing

QM9-XAS database of 56k QM9 small organic molecules labeled with TDDFT X-ray absorption spectra

<p>Database for training graph neural network (GNN) models in&nbsp;<strong>Integrating Explainability into Graph Neural Network Models for the Prediction of X-ray Absorption Spectra,&nbsp;</strong>by&nbsp;Amir Kotobi,&nbsp;Kanishka Singh,&nbsp;Daniel H&ouml;che,&nbsp;Sadia Bari,&nbsp;Robert H.Mei&szlig;ner,&nbsp;and Annika Bande.</p> <p><strong>Included:</strong></p> <ul> <li>qm9_Cedge_xas_56k.npz: the TDDFT XAS spectra of 56k structures from the QM9 dataset, were employed to label the graph dataset. The dataset contains two pairs of key/value entries: <strong>spec_stk</strong>,&nbsp;which represents a 2D array containing energies and oscillator strengths of XAS spectra, and <strong>id</strong>,&nbsp;which consists of the indices of QM9 structures. This data was used to create the QM9-XAS graph dataset.</li> <li>qm9xas_orca_output.zip: the raw ORCA output of TDDFT calculations for the 56k QM9-XAS dataset consists of excitation energies, densities, molecular orbitals, and other relevant information. This unprocessed output serves as a source to derive ground truth data for explaining the predictions made&nbsp;by&nbsp;GNNs.</li> <li>qm9xas_spec_train_val.pt: processed graph train/validation dataset&nbsp;of 50k QM9 structures. It is used as input to GNN models for training and validation.</li> <li>qm9xas_spec_test.pt: processed graph test dataset&nbsp;of 6k QM9 structures. It is used to test&nbsp;the performance of trained GNN models.</li> </ul> <p><strong>Notes on the datasets:</strong></p> <ul> <li>The QM9-XAS dataset was created using&nbsp;ORCA electronic structure package [Neese, F.,&nbsp;WIREs Computational Molecular Science 2012, 2, 73&ndash;78] to calculate carbon K-edge XAS spectra with the time-dependent density functional theory (TDDFT)&nbsp;method [Petersilka, M.; Gossmann, U. J.; Gross, E. K. U.,&nbsp;Phys. Rev. Lett. 1996, 76, 1212&ndash;1215]</li> <li>The molecular structures of QM9-XAS datasets were sourced from the QM9 database [R. Ramakrishnan, P. O. Dral, M. Rupp, and O. A. Von Lilienfeld,&nbsp;<em>Sci. Data</em>&nbsp;1, 1 (2014)].</li> </ul> <p><strong>Funding:</strong></p> <p>This<strong>&nbsp;</strong>research was funded by&nbsp;HIDA Trainee Network program, HAICU, Helmholtz&nbsp;AI-4-XAS, DASHH and HEIBRiDS graduate schools. For theoretical calculations and model training, computational resources at DESY and JFZ&nbsp;were used.&nbsp;</p>