Editorial: preclinical data reproducibility for R&D - the challenge for neuroscience

The inability to reproduce published findings has been identified as a major issue in science. Reports of only a low percentage of landmark studies being reproduced at pharmaceutical companies like Bayer (Prinz et al. 2011) gained much interest in the scientific community and raised high levels of concerns. A more recent analysis from Amgen (Begley and Ellis 2012) suggested that those non-reproducible studies may have an even stronger impact on the field than those that can be reproduced, possibly because the more remarkable and exciting findings are reported in higher impact journals. Evidently, this is not just a problem of pharmaceutical industry. About half of respondents from faculty and trainees at the academic MD Anderson Cancer Center, Houston, Texas, had experienced at least one episode of inability to reproduce published data in a survey by Mobley et al. (2013) and comparable figures may be expected in neuroscience.