While reviewing a manuscript for the Journal of Organic Chemistry, Caroline Kervarc-Genre and her colleague, Thibault Cantat, researchers at the French Alternative Energies and Atomic Energy Commission, noticed something unusual.

The nuclear magnetic resonance (NMR) spectra buried in the supplementary information had striking irregularities: The baseline was interrupted in some parts, and the noise was the same from one spectrum to the next. “Noise being inherently random, repeating noise is only possible if the spectra are altered [or] fake,” Kervarc-Genre told Retraction Watch.

Starting to suspect something was wrong, she and Cantat, examined other papers by the lead author. They discovered data appeared to have been edited in several of the author’s latest publications. “The fraud was not subtle,” Kervarc-Genre said.