The original question: "I would like to use the lsq commands such as lsq_ex to match together DNA helices. I have rigorously followed the manual for the protein example and that works fine but as soon as I try to use the lsq commands for my structures, I get into all sorts of problems. I can't seem to get the syntax right for specifying the atoms/residues/zone and the example uses CA traces which my DNA obviously does not have. Could some kind person explain what the format is its expecting, since I have tried every conceivable permutation and still get "Error, try again" or some similar, cute but unhelpful message.
One reason I'm using lsq is to get the orientation matrix of one structure from a rotated and translated one (from AMORE). Is this the correct way of doing this or is there a much easier and quicker way which I've overlooked?"
The lsq_* commands are explained in detail in chapter 5 of the tutorial (pub/moron).
However, it's probably much easier & faster to use LSQMAN (bundled with DEJAVU, pub/gerard/dejavu); here you can define exactly *which* types of atoms should be used in the superpositioning (e.g. "nonh" for all non-hydrogens), so it works just as well for dna as it does for proteins Since the two molecules are identical, you don't need. to improve the operator
With LSQMAN, it should work ~ as follows:
read mol1 amore1.pdb read mol2 amore2.pdb atom_types nonh explicit mol1 a1-20 mol2 a1 (*assuming 20 bases numbered 1-20 *) show mol1 mol2 save_operator mol1 mol2 rt_mol1_to_mol2.o .lsq_rt_mol1_to_2 quit
Note that this gives you the operator FROM mol2 TO mol1. You can now read the operator into O and (provided its name begins with ".lsq_rt_") use it in lsq_mol, lsq_obj etc.
The operator file can also be used as is for averaging with RAVE.
With CELLO (pub/gerard/xutil) you can interconvert O and X-PLOR NCS operators (options Import & Export); if you feed the above file into it, you will get an ncs-strict "include" file for X-PLOR.