By working with journals and crystallographers, databases aim to catch fraud earlier. Yeah
In 2022, a study revealed that about 800 papers published in crystallography and exotic-chemistry journals originated from a paper mill.
This publication is to be congratulated for taking action to ensure the integrity of the scientific record. The criteria they used to identify dodgy and compromised papers could easily serve as a checklist for the review of a bibliography.
In response to Bimler’s study, the Cambridge Structural Database (CSD), one of the most popular crystallography databases, flagged 992 crystal structures mentioned in the implicated papers. As of April 2023, the CSD had formally retracted 209 of the flagged structures after journals retracted some of the implicated papers. But many of the flagged structures remain in the database because they come from articles in predatory journals, which often have no interest in retracting papers they know to be fraudulent.
Databases have established systems for spotting fraud both when scientists submit a structure and when users later browse the database. There seems to be a consensus that databases such as the CSD are doing a good job of weeding out fraudulent crystal structures. Meanwhile, databases are recruiting experts to check the validity of structures and making it easy for users to review one another’s submissions.
Chawla, D.S. (2023) Crystallography Databases Hunt for Fraudulent Structures. ACS Central Science. 2023 Oct 9;9(10):1853-1855. doi: 10.1021/acscentsci.3c01209. PMID: 37901172; PMCID: PMC10604003.
Publisher (Open Access): https://pubs.acs.org/doi/10.1021/acscentsci.3c01209
