I think that the rebuild tool should allow the a rebuild of a section that can "swing the rest of the protein, without rebuilding it. A good way of thinking about it might be, in a de-novo freezing the sections that you've already built, and then selecting the frozen selection along with the next section down the line, so as to allow free movement of the ends of the rebuilding selection. In this way, the frozen section would not be rebuilt, but would be a part of the rebuild selection, and would enable it to swing freely by the direction of the rebuilding section.
I hope that made sense.
I`m not sure about what you ask, but try to switch in "selection interface" and rebuild only part selected by you.
I hope that you mean.
Here is an example of what I mean:
if in de-novo, you want to rebuild 40-50, that would never form a helix, because 39 and 51 are locked in 3D space. But if you were to rebuild 1-50, the region 40-50 could form a helix, because 39 and 51 aren't holding the segment in a straight line now. The downside of this routine is the fact that the user cannot control the focus of the rebuilding selection, because one cannot rebuild something with any frozen segments, so the program would rebuild all 1-50. I'd want to be able to freeze some areas of a rebuild, so that the program "ignores" them, but allows the frozen area to be moved throughout 3D space.
You need to bend that part (by band and wiggle ie) and when 2 ends of not formed helix will be in distance that allow to form helix rebuild that region. I always rebuilding helix + few loop segments aside to allow some free movement.
Right, that's what I've been doing, but I was thinking in the context of a script. I just thought that it would be simpler to have it automated in that way, as opposed to maybe, banding the beginning and end of a predicted helix together, and then rebuilding from there.
It was just a suggestion, that I think would be helpful best allowing the user to get the folds that he or she would want. The most difficult part about a de-novo (for me) isn't thinking of how the protein might look, it's translating that into the game.
Yes, indeed - that sounds like it would be an incredibly useful tool
