Some Cancer Studies Fail to Replicate. That Might Be OK

A years-long effort to validate key cancer biology hit roadblocks and found problems. But maybe this will incentivize scientists to share data.