Has anyone done work with BioJava on very large sequences (i.e. contigs)? The types of issues we're thinking about are keeping a sub-set of the sequence in memory, but ensuring that the indices of the bases are accurate. Has anyone dealt with this? Greg