A quick way to identify very similar or identical sequences in JalView

I regularly find myself with large edited alignments which contain identical sequences (as the N and C terminal parts have been trimmed). I would like a quick way to find these sequences and remove selected ones (I want to leave very similar ones which are of interest). Does anyone have a good way to do this in JalView. Building a crude tree works for smaller alignments…but for me this takes hours for a larger alignment. Having a distance matrix with identity values (or some better variant), which is linked in the GUI back to the original alignment would be ideal. Just highlighting sequences with >n% identity to at least one other sequence would also do the trick.

Many Thanks
Matthew Hindle

Of course, Jalview should also provide the option to select sequences rather than hide them - which would fix yet another item on the todo list:
http://issues.jalview.org/browse/JAL-250

···

On Tue Jul 17 07:40:39 2012, Matthew Hindle wrote:

Thanks, this seems to work great if I slide the filter to the % level
I want and then delete the highlighted ids I don't want. I now
remember looking at this in the tutorial but didn't notice it
also helpfully highlights them.