I've updated the Target Identification/Validation page in the Drug Discovery Resources. In particular I've added a section on the failure to reproduce literature experiments.

Remember this is an absolutely critical step, almost everything else can be fixed.