File Sets in DataStage have three key characteristics: they are partitioned across multiple files for parallel processing (A), they are readable by external applications because they use a standard format (D), and they consist of a header file describing the structure plus separate data files (E). Option B is false (file sets are partitioned), and Option C is false (they are multiple files, not a single file).