dear all,

i want to work on two protein-ligand-complexes in the same instance of 
xplor at the same time. in the segment section i called the protein part 
of complex 1 PRO1, the ligand LIG1; and PRO2 and LIG2 accordingly (the 
coordinates of the two proteins partially overlap). after hydrogen 
adding i want to separate the respective complexes from each other using


CONStraints

   INTEraction (segid PRO1) (segid PRO1)
   INTEraction (segid PRO2) (segid PRO2)
   INTEraction (segid LIG1) (segid LIG1)
   INTEraction (segid LIG2) (segid LIG2)

   INTEraction (segid PRO1) (segid LIG1)
   INTEraction (segid LIG1) (segid PRO1)

   INTEraction (segid PRO2) (segid LIG2)
   INTEraction (segid LIG2) (segid PRO2)

 END


then i perform energy minimization and short MD simulation (the random 
seed i set same for each run).
when i compared the coordinate output of this combined runs with two 
separate runs of the single complexes, i noticed quite big rmsds between 
the corresponding structures. this might be no surprise due to the fact 
that i perform energy minimization of the two complexes simultaneously 
and the way of minimizing the energy of the whole system might be 
different from minimizing the single systems. i next tested the same 
setup as described above with the second complex rotated or translated. 
now the system is the same as above, and if the two complexes do not see 
each other (as i want them to) the outcome should be the same, which it 
is not. am i overlooking something really simple? the energy terms i use 
in the minimization step are bond angle impr vdw elec, maybe there is a 
term which is excluded from the interaction constraint? or might it be 
some non-bonded interaction i do not capture with the constraint?

best regards,
benjamin

Reply via email to