[phenixbb] problem with alternate conformations

[Nathaniel Echols]

On Thu, Jul 11, 2013 at 5:58 AM, George Phillips <georgep at rice.edu> wrote:
> Thanks, Nat this is helpful.  Would you know the last version that did not
> have this requirement?

Probably 1.7.1, but there are many disadvantages to this - in
particular the geometry isn't going to be nearly as good.

> Are old versions still available?  I suppose I could with a little help, I
> could lobotomize a version of phenix not
> to do this during refinement but check the structures later one by one.
> Its not clear to me why it would work fine for three copies but not four or
> more, so it might be 'fixable'?

I suspect it's simply not a scenario that the authors of Reduce
envisioned when they wrote it.  There may be a straightforward
modification of Reduce that would fix the problem, but I think a more
optimal solution would be to remove the dependency in phenix.refine
entirely.  I receive a lot of other bug reports coming from the same
bottleneck (but I suspect for very different reasons).

-Nat