How do you validate that your experimental assays are yielding robust, reproducible data before moving to the next phase of discovery at GSK?