Automated evaluation of consistency within the PubChem Compound database

Hesam Dashti, Jonathan R. Wedell, William M. Westler, John L. Markley, Hamid R. Eghbalnia
2019 Scientific Data  
Identification of discrepant data in aggregated databases is a key step in data curation and remediation. We have applied the ALATIS approach, which is based on the international chemical shift identifier (InChI) model, to the full PubChem Compound database to generate unique and reproducible compound and atom identifiers for all entries for which three-dimensional structures were available. This exercise also served to identify entries with discrepancies between structures and chemical
more » ... or InChI strings. The use of unique compound identifiers and atom nomenclature should support more rigorous links between small-molecule databases including those containing atom-specific information of the type available from crystallography and spectroscopy. The comprehensive results from this analysis are publicly available through our webserver [http://alatis.nmrfam.wisc.edu/].
doi:10.1038/sdata.2019.23 pmid:30778259 pmcid:PMC6380220 fatcat:3gysuhgzsbbwdlq2jp6s3nlb3y