Cleaned and pre-processed MS/MS datset (build from all positive ionmode spectra in GNPS) - zip file

<p>Large MS/MS dataset build from data that was obtained from GNPS (accessed on 2020-05-11): <a href="https://gnps-external.ucsd.edu/gnpslibrary/ALL_GNPS.json">https://gnps-external.ucsd.edu/gnpslibrary/ALL_GNPS.json</a></p> <p>The data was cleaned and pre-processed using notebooks provided here: https://github.com/iomega/spec2vec_gnps_data_analysis/tree/master/notebooks</p> <ul> <li>112,956 positive ionmode spectra</li> <li>metadata was cleaned and corrected using matchms (https://github.com/matchms/matchms) and lookup routines using PubChem</li> <li>92,954 of the spectra have Smiles and InchiKey (13717 unique InchiKey in first 14 characters)</li> </ul> <p>&nbsp;</p> <p>Was used for the main article on Spec2Vec --&gt; <a href="https://doi.org/10.1371/journal.pcbi.1008724">https://doi.org/10.1371/journal.pcbi.1008724</a></p>