I would be careful with Rfactors at the early stages of refinement,
especially in the presence of twinning (apparent or real). You can see
weird behaviour of Rfactors in this presentation:
www.ysbl.york.ac.uk/refmac/Presentations/Refmac_Erice_workshop.ppt,
slides 18-20.
In short if you have perfect twinning (you seem to have perfect
twinning along one axis, the second axis seems to be ghost) and you
have modelled twinning then random R-factor could be as low as 40%.
I.e. even if your Rfactor after refinement is around 40 you cannot be
sure that you have solved the structure. Best strategy would be to
analyse electron density very carefully. Can you see some new features
that are not in the model? What happens if you remove several residues
and refine? Do they come back.
If you want you can send your data and I can try to analyse here to
see what is going on
regards
Garib
On 19 Jul 2010, at 08:41, 孙庆祥 wrote:
Dear Collins and all,
Thanks for the reply. Please find the following pdb header. For your
information, this is after refmac5 with twin-refinement option
selected. This is the best R values that I'm getting. Cutting
resolution does not help much in getting the Rs down...
HEADER SWISS-MODEL SERVER ( pa-JAN-g) XXXX
COMPND SS-MODEL
REMARK 3
REMARK 3 REFINEMENT.
REMARK 3 PROGRAM : REFMAC 5.5.0109
REMARK 3 AUTHORS : MURSHUDOV,VAGIN,DODSON
REMARK 3
REMARK 3 REFINEMENT TARGET : MAXIMUM LIKELIHOOD
REMARK 3
REMARK 3 DATA USED IN REFINEMENT.
REMARK 3 RESOLUTION RANGE HIGH (ANGSTROMS) : 1.52
REMARK 3 RESOLUTION RANGE LOW (ANGSTROMS) : 74.96
REMARK 3 DATA CUTOFF (SIGMA(F)) : NONE
REMARK 3 COMPLETENESS FOR RANGE (%) : 91.56
REMARK 3 NUMBER OF REFLECTIONS : 27851
REMARK 3
REMARK 3 FIT TO DATA USED IN REFINEMENT.
REMARK 3 CROSS-VALIDATION METHOD : THROUGHOUT
REMARK 3 FREE R VALUE TEST SET SELECTION : RANDOM
REMARK 3 R VALUE (WORKING + TEST SET) : 0.42077
REMARK 3 R VALUE (WORKING SET) : 0.42170
REMARK 3 FREE R VALUE : 0.40609
REMARK 3 FREE R VALUE TEST SET SIZE (%) : 5.4
REMARK 3 FREE R VALUE TEST SET COUNT : 1584
REMARK 3
REMARK 3 FIT IN THE HIGHEST RESOLUTION BIN.
REMARK 3 TOTAL NUMBER OF BINS USED : 20
REMARK 3 BIN RESOLUTION RANGE HIGH : 1.522
REMARK 3 BIN RESOLUTION RANGE LOW : 1.561
REMARK 3 REFLECTION IN BIN (WORKING SET) : 1804
REMARK 3 BIN COMPLETENESS (WORKING+TEST) (%) : 82.44
REMARK 3 BIN R VALUE (WORKING SET) : 0.529
REMARK 3 BIN FREE R VALUE SET COUNT : 121
REMARK 3 BIN FREE R VALUE : 0.622
REMARK 3
REMARK 3 NUMBER OF NON-HYDROGEN ATOMS USED IN REFINEMENT.
REMARK 3 ALL ATOMS : 984
REMARK 3
REMARK 3 B VALUES.
REMARK 3 FROM WILSON PLOT (A**2) : NULL
REMARK 3 MEAN B VALUE (OVERALL, A**2) : 18.096
REMARK 3 OVERALL ANISOTROPIC B VALUE.
REMARK 3 B11 (A**2) : 8.45
REMARK 3 B22 (A**2) : -12.86
REMARK 3 B33 (A**2) : 4.41
REMARK 3 B12 (A**2) : 0.51
REMARK 3 B13 (A**2) : -1.32
REMARK 3 B23 (A**2) : -8.65
REMARK 3
REMARK 3 ESTIMATED OVERALL COORDINATE ERROR.
REMARK 3 ESU BASED ON R VALUE (A):
0.032
REMARK 3 ESU BASED ON FREE R VALUE (A):
0.026
REMARK 3 ESU BASED ON MAXIMUM LIKELIHOOD (A):
0.082
REMARK 3 ESU FOR B VALUES BASED ON MAXIMUM LIKELIHOOD (A**2):
5.306
REMARK 3
REMARK 3 CORRELATION COEFFICIENTS.
REMARK 3 CORRELATION COEFFICIENT FO-FC : 0.812
REMARK 3 CORRELATION COEFFICIENT FO-FC FREE : 0.853
REMARK 3
REMARK 3 RMS DEVIATIONS FROM IDEAL VALUES COUNT RMS
WEIGHT
REMARK 3 BOND LENGTHS REFINED ATOMS (A): 1016 ; 0.008 ;
0.022
REMARK 3 BOND ANGLES REFINED ATOMS (DEGREES): 1382 ; 1.482 ;
1.974
REMARK 3 TORSION ANGLES, PERIOD 1 (DEGREES): 122 ; 9.603 ;
5.000
REMARK 3 TORSION ANGLES, PERIOD 2 (DEGREES): 46 ;33.363 ;
26.087
REMARK 3 TORSION ANGLES, PERIOD 3 (DEGREES): 176 ;21.472 ;
15.000
REMARK 3 TORSION ANGLES, PERIOD 4 (DEGREES): 2 ;13.956 ;
15.000
REMARK 3 CHIRAL-CENTER RESTRAINTS (A**3): 144 ; 0.102 ;
0.200
REMARK 3 GENERAL PLANES REFINED ATOMS (A): 774 ; 0.007 ;
0.022
REMARK 3
REMARK 3 ISOTROPIC THERMAL FACTOR RESTRAINTS. COUNT RMS
WEIGHT
REMARK 3 MAIN-CHAIN BOND REFINED ATOMS (A**2): 624 ; 0.917 ;
1.500
REMARK 3 MAIN-CHAIN ANGLE REFINED ATOMS (A**2): 1020 ; 1.686 ;
2.000
REMARK 3 SIDE-CHAIN BOND REFINED ATOMS (A**2): 392 ; 1.691 ;
3.000
REMARK 3 SIDE-CHAIN ANGLE REFINED ATOMS (A**2): 362 ; 2.734 ;
4.500
REMARK 3
REMARK 3 ANISOTROPIC THERMAL FACTOR RESTRAINTS. COUNT RMS
WEIGHT
REMARK 3 RIGID-BOND RESTRAINTS (A**2): 1016 ; 0.952 ;
3.000
REMARK 3
REMARK 3 NCS RESTRAINTS STATISTICS
REMARK 3 NUMBER OF NCS GROUPS : NULL
REMARK 3
REMARK 3 TWIN DETAILS
REMARK 3 NUMBER OF TWIN DOMAINS : 4
REMARK 3 TWIN DOMAIN : 1
REMARK 3 TWIN OPERATOR : H, K, L
REMARK 3 TWIN FRACTION : 0.307
REMARK 3 TWIN DOMAIN : 2
REMARK 3 TWIN OPERATOR : -H, K, -K-L
REMARK 3 TWIN FRACTION : 0.191
REMARK 3 TWIN DOMAIN : 3
REMARK 3 TWIN OPERATOR : -H, -K, K+L
REMARK 3 TWIN FRACTION : 0.198
REMARK 3 TWIN DOMAIN : 4
REMARK 3 TWIN OPERATOR : H, -K, -L
REMARK 3 TWIN FRACTION : 0.304
REMARK 3
REMARK 3
REMARK 3 TLS DETAILS
REMARK 3 NUMBER OF TLS GROUPS : NULL
REMARK 3
REMARK 3
REMARK 3 BULK SOLVENT MODELLING.
REMARK 3 METHOD USED : MASK
REMARK 3 PARAMETERS FOR MASK CALCULATION
REMARK 3 VDW PROBE RADIUS : 1.40
REMARK 3 ION PROBE RADIUS : 0.80
REMARK 3 SHRINKAGE RADIUS : 0.80
REMARK 3
REMARK 3 OTHER REFINEMENT REMARKS:
REMARK 3 HYDROGENS HAVE BEEN ADDED IN THE RIDING POSITIONS
REMARK 3 U VALUES : REFINED INDIVIDUALLY
REMARK 3
SSBOND 1 CYS A 3 CYS A 24
SSBOND 2 CYS A 17 CYS A 41
SSBOND 3 CYS A 43 CYS A 54
SSBOND 4 CYS A 55 CYS A 60
SSBOND 5 CYS B 3 CYS B 24
SSBOND 6 CYS B 17 CYS B 41
SSBOND 7 CYS B 43 CYS B 54
SSBOND 8 CYS B 55 CYS B 60
CRYST1 18.232 39.633 77.526 104.82 90.08 90.04 P 1
SCALE1 0.054849 0.000038 0.000087 0.00000
SCALE2 0.000000 0.025232 0.006677 0.00000
SCALE3 0.000000 0.000000 0.013343 0.00000
>when you say that the Rfree is 0.4, what statistic are you quoting?
For all of your data or the highest shell It is clear from the data
reduction statistics that your data do not go to 1.57 angstroms.
There are no data at that resolution. The <I> is 1/2 of the error;
<I> 24 and error about 40. The Rmerge for the shell is 59% which is
close enough to random for me to have little confidence in it.
>It would help us if you just cut and paste the statistics into the
question.
>You are right, the density looks great. There is something funny
about the statistics if this is a normal 2fo-fc map.
>cheers,
>-ed collins
>
>On Jul 18, 2010, at 10:39 PM, 孙庆祥 wrote:
>
>> <Density over model.png>
>
您想拥有和网易免费邮箱一样强大的软件吗？
