I have a database which had all it's foreign-key relationships dropped at some unknown time in the past (don't ask). I have an old copy of the database which isn't good to restore from, but the schema has the relationships. I'm working from that to create a script to restore the keys.
In updating the tables, I've had to drop statistics from several tables. Do I need to manually recreate those, or can I just run the statistics update procedure when all the tables are updated?
Statistics will be created by the engine when needed, or by an explicit CREATE STATISTICS
Personally, I'd do nothing with the statistics.
Saying that, I'm very surprised you had to drop statistics for any reason. Do you mean indexes? I've never had to drop or create statistics manually, ever.