Hi Noel,
thank you for your message. I finally got around to prepare a self contained example which you can run directly on colab https://colab.research.google.com/drive/1MJ8UBs7YgmIkfduQrf9axXQZy3o3gcMd?usp=sharing If you uncomment the mol.endModify() line at the end all the residues in the mol OBMol become unlinked. If the line is commented the PDB file is written out correctly and can be reread by OpenBabel. We figured out a workaround in the meantime, where we save the file to disk and reload it because afterwards we are making some modifications (like changing torsion angles). If we do not call endModify and eg try to set a new torsion directly on "mol" we get a segmentation fault. Best and thanks for your help, Simon Simon Dürr Doctoral Assistant Laboratory of Computational Biochemistry and Chemistry EPFL SB ISIC LCBC BCH 4105 (Bât. BCH) CH-1015 Lausanne phone: +41 78 607 2524 i...@simonduerr.eu<mailto:i...@simonduerr.eu> https://simonduerr.eu<https://simonduerr.eu/> ________________________________ From: Noel O'Boyle <baoille...@gmail.com> Sent: Saturday, October 17, 2020 4:13:31 PM To: Dürr Simon Leonard Cc: openbabel-discuss@lists.sourceforge.net Subject: Re: [Open Babel] mol.EndModify() and upgrade to ob3.0 Hi Simon, I'm finding it a bit confusing to work out what's going on here. If you could either provide the simplest possible self-contained example, or else describe what you want to do. At the back of my mind is the fact that molA += molB is how you add one molecule to another in Open Babel - it is not expected that users copy over each atom one-by-one. Regards, Noel On Thu, 15 Oct 2020 at 13:41, Dürr Simon Leonard <simon.du...@epfl.ch<mailto:simon.du...@epfl.ch>> wrote: Hi, I am currently upgrading our code from openbabel 2.4.2 to openbabel 3.0 (using the python bindings). I am having trouble with the EndModify command. What the code is doing: 1. Load a pdb file and a mol2 file of an amino acid (aa_mol) into seperate obMols 2. Set both ObMol to BeginModify() 3. Calculate a rotation vector and remove overlapping atoms via DeleteAtom() from aa_mol and pdb obmol 4. Then create a new obatom in the pdb file and copy all information and remake the bonds new_atom.SetId(id) new_atom.SetVector(frag_atom.GetVector()) new_atom.SetAtomicNum(frag_atom.GetAtomicNum()) new_atom.SetType(frag_atom.GetType()) new_atom.SetSpinMultiplicity(frag_atom.GetSpinMultiplicity()) new_atom.SetFormalCharge(frag_atom.GetFormalCharge()) #new_atom.SetImplicitValence(frag_atom.GetImplicitValence()) #old 2.4.2 replaced with line below new_atom.SetImplicitHCount(frag_atom.GetImplicitHCount()) new_atom.SetPartialCharge(frag_atom.GetPartialCharge()) for obbond in openbabel.OBMolBondIter(fragment): begin_atom = obbond.GetBeginAtom() end_atom = obbond.GetEndAtom() mol.AddBond(obbond.GetBeginAtomIdx() + prevAtoms, obbond.GetEndAtomIdx() + prevAtoms, obbond.GetBondOrder(), obbond.GetFlags()) 5. Renumber the atoms so that the newly inserted atoms are at the desired location 6. Make a bond between the fragment and the CA atom in the pdb file 7. Set name of residue. 8. Set Dihedral angle. Now this code was working perfectly fine in ob2.4.2, but now in 3.0 it does not anymore. I get a Segmentation fault when setting the Dihedral angle. If I move the mol.EndModify() before the SetTorsion command the segmentation fault disappears but all the atoms become unlinked (Residue UNL). If I write out a PDB file just before the EndModify command everything looks alright (correct CONECT statements, positions, order etc.). Any ideas where the problem is? Thank you for your help, Simon Simon Dürr i...@simonduerr.eu<mailto:i...@simonduerr.eu> https://simonduerr.eu<https://simonduerr.eu/> _______________________________________________ OpenBabel-discuss mailing list OpenBabel-discuss@lists.sourceforge.net<mailto:OpenBabel-discuss@lists.sourceforge.net> https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
_______________________________________________ OpenBabel-discuss mailing list OpenBabel-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-discuss
