Pointless (P41) vs. Phaser (P43)

From: Young-Jin Cho
Date: 11 October 2011 20:02
Hi all,
I recently got diffraction data of 214 AA protein.  When I processed
the data, pointless suggested me a space group as P41. However, when I
ran Phaser with 'all choices of alternate space group', it gave me a
pdb file with P43. Additionally, phenix.xtriage suggested me P422 with
twin laws (h,-k,-l).
Anyway, it seems like P43 is right space group (I tried to process and
refine all possible space groups, and  based on R values P43 is most
reasonable).
At this point, I'd like to know more first why different programs
suggested different space group, and which is the most reliable way to
determine the space group. Also, what is the biggest difference
between these space groups. In practice, which way is the best method
to get a right space group other than trying all possible space groups
as I just did.

Thanks in advance,

Young-Jin

----------
From: Frederic VELLIEUX
P4(1) and P4(3) are enantiomorphic space groups. The only difference is the helix (one way, or another). No difference in the diffraction pattern. Hence a program (or a
crystallographer) cannot distinguish the 2 based on the diffraction pattern.

Once you start phasing, e.g. by molecular replacement, there is one solution for one enantiomer and far worse statistics for the other enantiomer. This solves it.

And I'd like to point out that it is always a good idea to keep in mind the alternative space groups even if you think you are solving the structure but haven't solved it
completely yet.

Fred.