No, constraints do not improved the Chis, but also don't harm them much.
Which NSDs, from selection or superposition (damsel or damsup)? The
superposition ones:
>>> P1
egp1p_04-1.pdb !! Reference file
egp1p_12-1r.pdb !! NSD = 0.912
egp1p_05-1r.pdb !! NSD = 0.900
egp1p_03-1r.pdb !! NSD = 0.960
egp1p_06-1r.pdb !! NSD = 0.878
egp1p_10-1r.pdb !! NSD = 1.016
egp1p_07-1r.pdb !! NSD = 0.876
egp1p_14-1r.pdb !! NSD = 1.014
egp1p_02-1r.pdb !! NSD = 0.995
egp1p_13-1r.pdb !! NSD = 1.037
egp1p_08-1r.pdb !! NSD = 1.004
egp1p_01-1r.pdb !! NSD = 1.100
egp1p_11-1r.pdb !! NSD = 1.067
egp1p_15-1r.pdb !! NSD = 1.021
egp1p_09-1r.pdb !! NSD = 0.969
>>> P2
egp2p_08-1.pdb !! Reference file
egp2p_09-1r.pdb !! NSD = 0.927
egp2p_12-1r.pdb !! NSD = 1.024
egp2p_10-1r.pdb !! NSD = 1.024
egp2p_13-1r.pdb !! NSD = 1.056
egp2p_07-1r.pdb !! NSD = 0.884
egp2p_06-1r.pdb !! NSD = 0.884
egp2p_01-1r.pdb !! NSD = 0.869
egp2p_03-1r.pdb !! NSD = 1.256
egp2p_14-1r.pdb !! NSD = 0.836
egp2p_15-1r.pdb !! NSD = 0.836
egp2p_02-1r.pdb !! NSD = 1.318
egp2p_05-1r.pdb !! NSD = 1.168
egp2p_04-1r.pdb !! NSD = 1.168
>>> P222
egp22p01-1.pdb !! Reference file
egp22p13-1r.pdb !! NSD = 1.268
egp22p14-1r.pdb !! NSD = 1.584
egp22p15-1r.pdb !! NSD = 1.584
egp22p02-1r.pdb !! NSD = 1.030
egp22p03-1r.pdb !! NSD = 1.030
egp22p07-1r.pdb !! NSD = 1.442
egp22p06-1r.pdb !! NSD = 1.442
egp22p04-1r.pdb !! NSD = 0.851
egp22p05-1r.pdb !! NSD = 0.851
egp22p08-1r.pdb !! NSD = 0.905
egp22p09-1r.pdb !! NSD = 0.905
egp22p12-1r.pdb !! NSD = 1.392
egp22p10-1r.pdb !! NSD = 1.392
>>> P4
egp4p_09-1.pdb !! Reference file
egp4p_07-1r.pdb !! NSD = 0.923
egp4p_08-1r.pdb !! NSD = 0.923
egp4p_05-1r.pdb !! NSD = 0.886
egp4p_02-1r.pdb !! NSD = 0.861
egp4p_01-1r.pdb !! NSD = 0.861
egp4p_03-1r.pdb !! NSD = 0.998
egp4p_04-1r.pdb !! NSD = 0.998
egp4p_06-1r.pdb !! NSD = 1.176
egp4p_13-1r.pdb !! NSD = 1.079
egp4p_12-1r.pdb !! NSD = 1.079
egp4p_14-1r.pdb !! NSD = 0.868
egp4p_15-1r.pdb !! NSD = 0.868
egp4p_10-1r.pdb !! NSD = 1.309
egp4p_11-1r.pdb !! NSD = 1.309
>>>P5
egp5S206-1.pdb !! Reference file
egp5S202-1r.pdb !! NSD = 0.675
egp5S205-1r.pdb !! NSD = 0.789
egp5S210-1r.pdb !! NSD = 1.044
egp5S212-1r.pdb !! NSD = 0.998
egp5S215-1r.pdb !! NSD = 0.929
egp5S204-1r.pdb !! NSD = 0.740
egp5S201-1r.pdb !! NSD = 1.077
egp5S211-1r.pdb !! NSD = 1.126
egp5S213-1r.pdb !! NSD = 0.773
egp5S209-1r.pdb !! NSD = 1.192
egp5S207-1r.pdb !! NSD = 1.213
egp5S214-1r.pdb !! NSD = 1.270
egp5S208-1r.pdb !! NSD = 0.710
Em 29-07-2010 13:11, Kushol Gupta escreveu:
Fred,
Two cents - I think the P1 SAXS solution should strongly guide your choice
of symmetry constraint above all else in this case: do any of the
symmetry-restrained shape reconstructions *improve* the statistics (chi) and
stability of the shape (NSD) when compared to the P1 result? Also, it sounds
like you have other data - do the theoretical Rs, f/fo, etc of the shapes
generated agree well with your other measurements?
Cheers,
Kushol
Kushol Gupta, Ph.D.
Research Associate
Van Duyne Laboratory - HHMI/Univ. of Pennsylvania School of Medicine
kgu...@mail.med.upenn.edu
215-573-7260 / 267-259-0082
Of course, 222 has not a 4 axis, otherwise it would be a 4-fold axis.
But that's the output of the program. P4 exp. model has a 4-fold axis
along the longest axis, while the P222 MODEL has a 4-fold axis along the
smallest, which doesn't make any sense. Can you imagine something build
up with 4 identical subunits and 222 symmtry, but without a 4-fold axis
at the molecular level (I mean at the envelop resolution level)?
Em 29-07-2010 12:32, Vellieux Frederic escreveu:
Hi,
To quote you: "even my P222 experimental envelop does have a 4-fold
axis" - this is not suprising, a particle with 222 symmetry does not
have 4-fold symmetry. There are 3 mutually perpendicular 2-fold axes
that intersect at the origin (of the "particle", of the molecule) [and
for the nomenclature, these axes are named the P Q and R axes].
Fred.
Fred wrote:
Thanks all of you who promptly replied my question.
I should have been more precise. I was referring to the symmetry of
the tetrameric particle (point symmetry) at the molecular level not
at the atomic level. This question has arisen because I have
collected some SAXS data of my protein in solution and I don't have a
molecular model to superpose to the experimental envelop. Others
experimental data, gel filtration and NAT-PAGE, suggest a tetrameric
particle. On the other side, P1, P2, P222 and P4 experimental
envelops are quite different. So, I am not sure which symmetry to
take. Considering the native state (no ligands at all), 4 identical
subunits and that the interface of oligomarization have to be
conserved, I would take P222 or P4. However, I can be able to imagine
such spacial arrangement without a 4-fold axis at the molecular
level. Indeed, even my P222 experimental envelop does have a 4-fold
axis.
I appreciate if you could add some more comments on this.
Thanks in advance,
Fred
