Check out the text of this recent retraction in Nature Cell Biology:
We regret the lack of integrity of exemplar data in Figures 2–4 in this publication. The irregularities have been investigated by the co-authors and Vanderbilt University in collaboration with the US PHS Office of Research Integrity, and the investigation concluded that exemplar data in this publication were fabricated (http://ori.hhs.gov/content/case-summary-dzhura-igor). We have therefore decided to retract the publication and have been unable to contact the first author, Igor Dzhura, to sign the retraction statement.
The science showing CaMKII activates CaV1.2 Ca2+ channels and facilitates CaV1.2 current is now extensively validated, so our view is that the overall conclusions from this paper are valid. We apologize to the scientific community for any inconveniences or challenges resulting from the publication and retraction of this manuscript.
Emphasis added. Why do journals let fuckers like this routinely claim in retractions that they got lucky and their conclusions based on fraudulent data turned out to be sound, because other subsequent published studies turned out to be consistent? (This kind of thing is pretty common in retractions.)
This means that the editors are allowing a scientific claim to be published in their journal without subjecting that claim to peer review. And the authors whose unvetted scientific claim they are publishing have already proven by virtue of the retraction that their claims going forward should be subject to extraordinary scrutiny!