If that's truly what I want to do - you would have to join your tables or concatenate them. It all depends on what columns you have in each table - If you want to bring new columns, you will have to use join (or keep, with an additional prefix left/right/inner/outer depending what's the data you want to have at the end of this transformation process). If you only want to bring a small number of columns - create mapping tables and ApplyMap() in the destination table.
However - concatenate will bring across new rows. If the structure between two (or more tables) is exactly the same - QV will concatenate them naturally, during the load script. But you mentioned that you already have 4 facts tables, which suggest they aren't the same and the columns contain different information (or possibly - they are called differently as they may be coming from various sources). You can of course force QV to concatenate two tables by adding Concatenate(destination_table_name).
It's very hard to advise what exactly you must do at this point - without knowledge how your tables look like or knowing which one should be use as the final facts table (considering the size, information that holds, etc...) it's a difficult task.
As few people before me mentioned already - is there a special reason you want to perform this operation? If the performance isn't an issue for you, I'm not sure what exactly you are trying to achieve.
