I'm not getting the expected result with this string: COC(=O)C(=COC)c1ccccc1COC(N1C)=CC(=N1)C(=C2O)C=C(Cl)C=C2 When used with coot Lidia it gives the expected result (upper picture). But when I run elbow like this: phenix.elbow --smiles="COC(=O)C(=COC)c1ccccc1COC(N1C)=CC(=N1)C(=C2O)C=C(Cl)C=C2" --output="Y7303" --id="Y03" the result is the lower picture. In particular the pyrazole ring (arrow) gets distorted, but other bond lengths are way off. I want to use elbow because it names the atoms as numbered in the order they appear in the smiles string, irrespective of the element. This is good because this is one of a series of compounds in some of which C's are replaced by N's and vice versa so it helps to have the same number for atoms in the same position. What does work is to save the structure as .mol from Lidia and feed that to elbow, which gives the right structure and the elbow atom names. But it would be good to find out why it isn't working directly. my elbow is: electronic Ligand Builder & Optimisation Workbench (eLBOW) 1.12rc1-2801 None