Thanks a lot! I've learnt two messages: 1. I agree GetBestRMS() is the right tool to compare conformers of the same structure 2. O3A is good for aligning different structures(or structures with different topology). However not only the O3A.Align() rmsd value should be check, the Score should be checked too, in case only a partial match exists with low rmsd value.
Best regards Zhenting Paolo Tosco <paolo.tosco.m...@gmail.com> 于2020年2月24日周一 下午10:25写道： > Hi Zhenting, > > if you compare different conformers of the same structure you won't need > O3A at all, as the atom-atom correspondences will be 1:1, as all conformers > share the same topology; GetBestRMS() is all you need. > > O3A is useful when you need to align molecules which have different > topologies, as it will find non-obvious matches between atoms. > > Regarding open-sorce shape similarity tools you might have a look at > Shape-it: > > > http://silicos-it.be.s3-website-eu-west-1.amazonaws.com/software/shape-it/1.0.1/shape-it.html > > Cheers, > p. > On 24/02/2020 14:16, Gao Zhenting wrote: > > Hi Paolo, > > Thanks for pointing that out. Currently, I am using Open3DAlign to select > diverse conformers of a single structure by measuring the rmsd. According > to your explanation, I will choose O3A.Align() to record the rmsd between > conformers. And my preliminary observation is O3A.Align() gave me a more > expected rmsd value comparing to rdkit.Chem.rdMolAlign.GetBestRMS(probeMol, > refMol). (python code attached) What is your comment on this? > > I do imagine to use Open3DAlign as a shape similarity tool. However with > your advice, I will think it over again. > Do you have any suggestions on an open source shape similarity tool? > > Best regards > Zhenting > > Paolo Tosco <paolo.tosco.m...@gmail.com> 于2020年2月24日周一 下午9:21写道： > >> Hi Zhenting, >> >> O3A.Align() returns the RMSD between the atom pairs that the O3A >> algorithm was able to match across the two molecules. >> >> O3A.Score() returns the score for an alignment. The score is directly >> proportional to: >> >> * the number of atom pairs that could be matched across the two molecules >> * the closeness of the match (i.e., how similar actually the atoms in the >> matched pair are) >> >> It is important to note that Open3DAlign was originally conceived as a >> tool to generate good quality 3D overlays of molecules ahead of a 3D-QSAR >> analysis. It was not designed, nor tested, as a tool to assess 3D >> similarity between molecules and do virtual screening. O3A.Score is used >> internally purely to pick the best overlay between two molecules. >> >> Cheers, >> p >> On 24/02/2020 12:45, Gao Zhenting wrote: >> >> Hi Greg, >> >> >> I am learning Open3D ALIGN >> >> >> example code: >> >> pyO3A = rdMolAlign.GetO3A(mol2py, mol1) >> pyO3A.Align() >> >> >> Questions >> >> - what is the meaning of the value from O3A.Align() >> - is that similar to atom wise RMSD? >> - >> - what is the meaning of the value from O3A.Score() >> - only an online tutorial presented the Score >> - ~ open3dqsar.org > Downloads > Open3DTOOLS tutorial 20130812 >> <http://open3dqsar.org/downloads/Open3DTOOLS_tutorial_20130812.pdf> >> >> >> - >> - Best Regards >> - >> - >> - Zhenting >> >> >> >> _______________________________________________ >> Rdkit-discuss mailing >> listRdkit-discuss@lists.sourceforge.nethttps://lists.sourceforge.net/lists/listinfo/rdkit-discuss >> >>
_______________________________________________ Rdkit-discuss mailing list Rdkit-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rdkit-discuss
