Amgen is releasing data on three of its failed efforts to confirm findings from high-profile journal articles, Nature News reports. The company has posted its data at the newly launched Preclinical Reproducibility and Robustness channel at F1000 Research, part of the Faculty of 1000.
By making this data public, Amgen says it hopes it will encourage others to do so and help tackle the issue of irreproducibility in research, particularly preclinical research. It's also, Sasha Kamb, senior vice president for research at Amgen, tells ScienceInsider, "to help improve the self-correcting nature of science to benefit society as a whole, including those of us trying to create new medicines."
In 2012, Amgen researchers made a splash when they announced that they were unable to reproduce the findings of 47 out of 53 landmark cancer papers.
The data they've posted, though, deal with other findings. As Nature News reports, these three postings counter findings of a Science paper that indicated that a cancer drug might also work as an Alzheimer's disease treatment, of two papers — one in Biochemical and Biophysical Research Communications and the other in the Journal of Clinical Investigation — that linked a certain gene to insulin sensitivity, and of a Nature paper that said inhibiting one protein could led to increased degradation of other proteins linked to neurodegenerative diseases.
However, Stanford University's John Ioannidis notes at Nature News that other efforts to encourage the reporting of replication studies and negative results have had limited success.