how to control distance between base pairs?

[Andrew Uzilov]

Hi all...

I have a test applet up here:

http://users.soe.ucsc.edu/~auzilov/tmp/plotVarna.U1a.html

The problem occurs when I switch on "Flat exterior loop" (hit Ctrl
+F). Two stems are really close together and overlap now!

If I could increase the distance between G136 and U137 and/or between
U137 and G138, I could move the helices apart and avoid the crowding.
Is there any way to do that?

Thanks!
Andrew

[Yann Ponty]

Hi Andrew,

Unfortunately, this is not possible at the moment.
We could imagine adding a "weight" parameter to each backbone
interval, but this would require a modification of each and every
layout algorithm! I'll give it a thought though...

A cheap fix to this issue, at least for the production of
illustrations, might be to disentangle the structure manually (eg
click and drag the second helix below the outer loop "flat" line to
get it drawn "downward"). If you wish to put this structure on a
website, you could possibly save your VARNA session and use it as an
input file for your applet.

Again, thanks for the feedback,

Yann

[Andrew V. Uzilov]

Ah yes, dragging the helix below the "flat" line looks like it will be
a good fix... thanks!

Perhaps as a heuristic to minimize manual post-processing, VARNA can
detect whether there is clash in "flat" layout and put things
above/below the line to minimize overlap? Yes I know it is a
heuristic, but I think that you will get a LOT of cases like these
with multibranch loop RNAs, and alternating the above/below of helices
facing the multibranch loop automatically would be very nice.

Cheers,
Andrew

[Yann Ponty]

> Perhaps as a heuristic to minimize manual post-processing, VARNA can
> detect whether there is clash in "flat" layout and put things
> above/below the line to minimize overlap?  Yes I know it is a
> heuristic, but I think that you will get a LOT of cases like these
> with multibranch loop RNAs, and alternating the above/below of helices
> facing the multibranch loop automatically would be very nice.

When you say "multibranch loop", you are talking about the external
loop right?
Your suggestion makes perfect sense in this context, especially since
I am fairly certain the problem is NP-complete.

Cannot promise any implementation in the near future, but once again
thanks for the suggestion which goes into the "to do" list (internally
nicknamed the "wish there were more than 24h/day" list).

Regards,

Yann

[Andrew V. Uzilov]

On Mon, May 17, 2010 at 4:10 PM, Yann Ponty <yann....@gmail.com> wrote:
>
>> Perhaps as a heuristic to minimize manual post-processing, VARNA can
>> detect whether there is clash in "flat" layout and put things
>> above/below the line to minimize overlap? Yes I know it is a
>> heuristic, but I think that you will get a LOT of cases like these
>> with multibranch loop RNAs, and alternating the above/below of helices
>> facing the multibranch loop automatically would be very nice.
>
> When you say "multibranch loop", you are talking about the external
> loop right?
> Your suggestion makes perfect sense in this context, especially since
> I am fairly certain the problem is NP-complete.

Oops, that is correct. I indeed meant exterior loop.

> Cannot promise any implementation in the near future, but once again
> thanks for the suggestion which goes into the "to do" list (internally
> nicknamed the "wish there were more than 24h/day" list).

Haha, indeed... yes, we all have lists like that! Flipping the
helices by hand is perfectly fine for now, though.

Thanks,
Andrew