The general way would be opening it with dataLoad, copying it into a larger dataset while filling the gaps, then saving it again.