POA end gap/semi-global alignment

[ctsa]

Hi, Nice to see recent progress on banded POA! Is there any support for some kind of clipping/end-gap feature for POA to allow semi-global/overlap alignments to the graph?

cc: @lorewar2

[lorewar2]

Hi, The current code does not include clipping, but I am sure that it can be integrated. I'll take a look and try to figure something out. Thanks.

[ctsa]

Thanks! I wasn't sure whether this would need something like a seq to graph version of LCSk++ to setup (I'm only familiar with the rust bio pairwise banding method), would be great if there's a clipping scheme that can be worked out for the POA version.

[lorewar2]

Yes, but LCSk++ was not used for the POA banding; the max scoring cell in the current row was considered as the band center for the next row, and so on. I'll try to implement unbanded POA clipping first (semi-global, local, etc).

[lorewar2]

I see the problem with working with banded POA now,

[ctsa]

I just noticed this looks like it's in the latest release, thanks!

[ctsa]

Will be great to see if the same capability can be worked into the banded POA eventually. This is already a big step ahead though.

[lorewar2]

I am trying to extend LCSk++ to work with POA, but I am not sure if this will work correctly.