I want to align some sequences that don't fit clustalw's assumptions. Given a reference sequence (about 10 kbases in length), and a few dozen shorter sequences (about 300 bases in length each), some of which are sense, and some are anti-sense, which don't neccessarily fully overlap the other short sequences, is there an alignment program (ideally available as a web-based program) that can align them all?

Background: In case anyone's wondering, it's not for a final analysis, but to see any common trends in group of suspect sequences.