Dear all,

A quick note to report the problems I have experienced (and a solution) with 
the refinement of a peptide containing both D- and L-amino-acids with refmac5.8 
(Refmac_5.8.0073).
This refinement was done with high resolution data, with hydrogens included in 
the model (important to remember !).

When using the standard dictionary for D-TRYPTOPHAN: $CLIBD_MON/d/DTR.cif, most 
of the geometrical definitions were altered after the inserted D-amino-acid. 
The geometrical Rms deviations are strongly increased already at the first 
cycle (see differences of two runs at the end of my message). 

The first solution was to generate a new cif dictionary from the standard 
TRP.cif file in the following way:

sed -e 's/negativ/positiv/'  -e 's/TRP/DTR/g' -e 's/L-peptide/D-peptide/' 
$CLIBD_MON/t/TRP.cif  > DTRP.cif

Running again refmac5 with this cif file using the LIBIN DTRP.cif instruction 
produced the more sensible results. 

It looked to me that the TRP and DTR cif definitions have diverged at some 
point. Looking more closely at the differences in this two files, pointed me to 
this:
Hydrogen on the indole NE1 is named HE1 in TRP and HNE1 for DRT...
Using a pdb file with HE1 renamed as HNE1 worked with the current DRT file, but 
with slightly worst results.

The DTR.cif file probably needs to be updated (and maybe the other 
D-amino-acids cifs also revisited).

Best regards,
Pierre

PS1. I can send the log and pdb files to developers if they are interested.
PS2. I haven't checked with other D-amino-acids, since I have only this one in 
my structure.

With modified TRP as DTRP.cif:
<                      Initial    Final
<            R factor    0.0865   0.0847
<              R free    0.0913   0.0896
<      Rms BondLength    0.0181   0.0175
<       Rms BondAngle    2.2555   2.2407
<      Rms ChirVolume    0.1131   0.1154

With standard DRT.cif file:
>                      Initial    Final
>            R factor    0.0865   0.1916
>              R free    0.0913   0.2006
>      Rms BondLength    2.4522   1.7415
>       Rms BondAngle   33.3672  32.1460
>      Rms ChirVolume    1.1972   1.0767

With DRT.cif and renamed HE1 as HNE1:
>                      Initial    Final
>           R factor    0.0865   0.0848
>             R free    0.0913   0.0901
>     Rms BondLength    0.0207   0.0200
>      Rms BondAngle    2.4024   2.3884
>     Rms ChirVolume    0.1122   0.1140

Reply via email to