Negative Results: Research papers from authors who tried to test important hypotheses but did not get the outcome they expected. Failed preclinical tests are particularly welcome. These manuscripts must include testing the hypothesis by multiple experimental approaches, rigorously reproducing the experimental models of others that you claim to refute, and meticulous use of both positive and negative controls.
Failure to Replicate: Research papers from authors who could not reproduce someone else’s work of significant importance despite using the same methodology (explaining why the reproduction failed is not mandatory).
(I'm not a big fan of media-specific journal titles like eLife or eNeuro—it's not like we have printJournal of Neuroscience—but I do like some of the changes they are instituting.)
UPDATE: Greg Hickok confirms that Psychonomic Bulletin and Review also encourages negative results (although this isn't spelled out on their website). Is there a central list of "journals that are happy to publish your negative results"?