SteinBlog

A study recently published in Acta Cryst. (2007). D63, 941-950 by E. N. Brown and S. Ramaswamy on the quality of protein crystal structures revealed a positive correlation between a journal’s impact factor and the error rate in its published protein structures.  The graph shown in my previous posting assumes the theoretical situation of a similar distribution of errors in crystal structures in all journals.  If that was true, 50% of all papers would have a worse-than-average and 50% a better-than-average quality. As the graph shows, the reality is different and high-impact journals publish more papers with low quality crystal structures. This can both be attributed to high-impact journals publishing cutting-edge and therefore less well defined structures as well as to authors rushing to publish high-impact work, according to a report in CE&N (Aug 20, 2007, 11).