[phenixbb] Group ADP not working

[Pavel Afonine]

Hi Alejandro,

1) it does group isotropic B refinement but it's hidden as inside step 
of TLS refinement, so you don't see it as a separate step.

2) You get different B-factors because you do TLS refinement. Look at 
TLS formula and you will see that each individual atoms is supposed to 
get different anisotropic B. Nothing wrong about it.

Please let me know if you have any other questions or problems or my 
explanations were not clear!

BTW, you can use parameter files instead of typing such a long command line.

Pavel.
 

On 6/3/2008 7:16 AM, alebus at pasteur.edu.uy wrote:
> Dear all,
>
> I am refining a structure (2.9angs res; P21) with phenix (Version:  
> 1.3b; Release tag: rc6; linux platform)
>
> I want to carry on group_ADP refinement, according to the ('lengthy')  
> command line :
>
> phenix.refine eden_unique1.hkl hlcoeff_blur.hkl abril28_phenix.pdb  
> input.experimental_phases.labels="PA_BLUR PB_BLUR PC_BLUR PD_BLUR"  
> ordered_solvent=false simulated_annealing=false  
> simulated_annealing.mode=first  
> strategy=rigid_body+individual_sites+group_adp+tls main.ncs=tru
> e main.number_of_macro_cycles=10 xray_data.low_resolution=15.0  
> xray_data.high_resolution=2.7 tls_group_selections.params  
> rigid_body_selections.params group_anomalous_1.params  
> elbow.acp_A_pdb.001.cif restraints_edits.params  
> convert_to_isotropic=true
>
> everything works just fine (including correct handling of an 'unusual'  
> ligand; selected NCS restraints; TLS; etc)...except for the group ADP  
> : phenix keeps doing individual atom B-factor refinement
>
> (e.g. these are some pertinent remarks in the output pdb
>
> REMARK ****************** REFINEMENT STATISTICS STEP BY STEP  
> ******************
> REMARK leading digit, like 1_, means number of macro-cycle
> REMARK 0    : statistics at the very beginning when nothing is done yet
> REMARK 1_bss: bulk solvent correction and/or (anisotropic) scaling
> REMARK 1_xyz: refinement of coordinates
> REMARK 1_rbr: rigid body refinement
> REMARK 1_gbr: group B-factor refinement
> REMARK  
> ------------------------------------------------------------------------
> REMARK  R-factors, x-ray target values and norm of gradient of x-ray target
> REMARK  stage     r-work r-free  xray_target_w  xray_target_t
> REMARK    0    :  0.3951 0.4346  -6.345353e+00  -6.295977e+00
> REMARK    1_bss:  0.3189 0.3519  -6.182904e+00  -6.132416e+00
> REMARK    1_rbr:  0.3199 0.3553  -6.164872e+00  -6.111524e+00
> REMARK    1_bss:  0.3197 0.3548  -6.158236e+00  -6.106543e+00
> REMARK    1_xyz:  0.3145 0.3612  -6.243414e+00  -6.163029e+00
> REMARK    1_adp:  0.2890 0.3327  -6.111496e+00  -6.011406e+00
> REMARK    2_bss:  0.2862 0.3290  -6.107396e+00  -6.004629e+00
> REMARK    2_xyz:  0.2801 0.3278  -6.104559e+00  -5.995221e+00
> REMARK    2_adp:  0.2783 0.3213  -6.062947e+00  -5.954846e+00
> REMARK    3_bss:  0.2783 0.3214  -6.063301e+00  -5.955103e+00
> REMARK    3_xyz:  0.2755 0.3218  -6.065099e+00  -5.950739e+00
> REMARK    3_adp:  0.2751 0.3218  -6.062715e+00  -5.947567e+00
> ...
> and so on, always doing 'X_adp' (and not 'X_gbr') on the different steps...
>
> starting from a single value on the B-factor column in the input pdb,  
> of course I end up having individually varied figures in the end,  
> which I don't want
>
> I will greatly appreciate any help on what am I doing wrong.
>
> Thanks,
>
> Alejandro
>
>
> _______________________________________________
> phenixbb mailing list
> phenixbb at phenix-online.org
> http://www.phenix-online.org/mailman/listinfo/phenixbb
>