Phaser no longer fixes the space group after the first translation
function. Phaser carries the space group with each potential solution. If
the solution is clear at the end, so will the space group be clear.
In practice, we found that the space group was not correct after the first
translation function only when there was pseudocentring present. Any
pseudocentring is accounted for by the new translational NCS correction
factors in Phaser, and with these corrections, the space group is clear (in
our test cases) after the first translation function.
So any issues with space group determination have been addressed in two
ways, and space group determination should be very robust from Phaser-2.5.0
onwards.
Airlie
On Jul 9 2012, David Schuller wrote:
Scaling should be the same in P222 vs. P212121. The only difference is
the exclusion of systematic absences. You may have run into some quirk
of Phaser in the way it handles multiple space groups vs. a single one.
On 07/09/12 14:06, Shya Biswas wrote:
Hi all,
I have a dataset that I scaled in p212121 with cell dimension a=28.9
b=67.1 and c=93.5 however I do not get a right MR solution with this.
So I went back and scaled it in p222 space group and asked phaser to
find the right spacegroup solution for it, this time phaser gave me
the right solution and in p212121 space group. By right solution I
mean the molecule is a dimer and has to be oriented in a particular
way, which I get only when I use p222 scaled data. I am puzzled and
would like to know if anyone can explain this.
thanks,
Shya
