All the programs use the same format for storing sequences. This includes the database, the query and any sequences extracted by scanps or sortsco. The format is as follows:
>IDENTIFIER TITLE LINE one letter code in capitals terminated by * >IDENTIFIER2 Title line one letter code..... * etc
This is the format of the NBRF-PIR database distributed for VAX. I use this format for historical reasons. If anyone can suggest which format is the most commonly used for database scanning, then I will support this. I guess that FASTA format as used by BLAST would be a good one to include...