If it was some kind of fluke and is not easily reproducible but the sample exists and is a room-temperature super-conductor then it becomes a reverse engineering problem.