Hi Chen, I used SUPPOSE previously. It worked very well. The compositions of the structures have to be the same! It works on Linux. http://structbio.vanderbilt.edu/~jsmith/suppose/
If the compositions of your structures are different, you may first load them to PyMOL, then align them use its commandline. http://pldserver1.biochem.queensu.ca/~rlc/work/teaching/BCHM823/pymol/alignment/ If the compositions/sequences of your structures are very very different, but their overall structure are similar, you can do the alignment using a PyMOL plugin called 'cealign'. It works on both Mac and Linux. http://pymolwiki.org/index.php/Cealign Here is my note on how to run SUPPOSE. #### remove the outputs from the previous runs rm -rf *.rot.pdb mean.pdb all.pdb # if the input file names are XXX.pdb, those of the output files are XXX.rot.pdb and a mean.pdb #### fit the structrue models based on the CA atoms, and then calculate the RMSD of the backbone atoms suppose -mat -rot -mean -fit ":11-52,85-97:CA" -calc ":11-52,85-97:C,CA,N" restrt_*.pdb # 0.639 #### fit the structrue models based on the CA atoms, and then calculte the RMSD of the heavy atoms suppose -mat -rot -mean -fit ":11-52,85-97:CA" -calc ":11-52,85-97:C*,N*,O*,S*" restrt_*.pdb # 1.23 # Do not visulize the rotated pdb files with molmol at this step, because molmol will shift the individual files. # That is, the output files have already been perfectly supperimposed, # but molmol gives an illusion that they have not been rotated correctly. #### join the pdb with the AWK script ./joinpdb -o all.pdb *.rot.pdb # input files: *.rot.pdb # output file: all.pdb #### display the rotated pdb with molmol now. molmol all.pdb # molmol can only display the combined pdb (i.e., 20 models in one file) correctly!!! HTH, Steven On Wed, Oct 15, 2014 at 4:05 PM, Chen Zhao <[email protected]> wrote: > Dear all, > > I am just wondering whether there is a simple command line tool for RMSD > calculation between two aligned structures? I just cannot find a good > answer online, and ccp4bb can always impress me. > > Thank you so much in advance, > Chen > -- Steven Chou
