On Fri, Aug 17, 2012 at 9:57 AM, Walter Novak
This residue is in an interesting location as it contacts a mutation, so perhaps we can expect some deviation. I am fairly sure Markus is on the right track (thanks for forwarding this!). I have been experimenting with an alternate conformer (there is some positive difference density, but it isn't as clear cut as many of the other cases).
A simple trick you could try is to reduce the occupancy of the Thr in the single-conformer model (somewhere in the 0.5-0.75 range is good) and re-calculate the maps. Sometimes this makes the difference map clear enough to show where the second conformation goes, since it isn't quite as flattened out by presence of the (full-occupancy) primary conformation. -Nat