Antiparallel scan

[ajasja]

If a peptide sequence both begins and ends with an f register, then reversing the sequence and scoring it with the usual scoring function is a rough guide for binding in the antiparallel orientation. Since this are of target states for us we would like to design against them. Hence the following feature request:

Could you add a flag (--antiparallel) to fastscore. When this flag is on then the scoring function should also check the reversed orientations. This basically means that the second string is reversed. For the score matrix the minimum value should be output. In addition to the alignment matrix another matrix should be output with 0 if the maximum is in parallel orientation and 1 if it is in antiparallel.

[dancsi]

1995e2d87f7303111697b025c27cfbe58c35ff69, try if it works. It outputs the antiparallel matrix to a similar binary format as the alignment matrix (8 bit unsigned integers, row after row).

[ajasja]

Thanks! Looks OK, I'll give it a try first thing in the morning.

On 27 October 2016 at 19:50, Daniel Siladji notifications@github.com
wrote:

1995e2d
https://github.com/dancsi/CoiledCoilOrtho/commit/1995e2d87f7303111697b025c27cfbe58c35ff69,
try if it works. It outputs the antiparallel matrix to a similar binary
format as the alignment matrix (8 bit unsigned integers, row after row).

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
https://github.com/dancsi/CoiledCoilOrtho/issues/22#issuecomment-256719674,
or mute the thread
https://github.com/notifications/unsubscribe-auth/AABN1mtqv1Zgn-aBK2OoJ-fETua5EOyyks5q4ORSgaJpZM4KiVQx
.

[ajasja]

Works great, just one more feature request.

For testing purposes (and to perhaps develop an antiparallel only set) I would need to either calculate both antiparallel and parallel orientation, only parallel or only anti-parallel.

I was thinking of either having parallel and antiparallel except boolean values with (example --parallel=1 and antiparallel=1, although setting both to 0 is of course invalid).

Another option, that I prefer a bit more, is having an orientation flag that can accept the values A for antiparallel, P for parallel or B for both. (example --orientation=B

Whatever is easier for you to implement.

Thanks!

[dancsi]

Option B is implemented and pushed.