Dear All, I have a protein-ligand complex cryo-EM data. The ligand is a primary amine, which at the neutral pH used during data collection, should be NH3+ (alkyl-amonium). However, during realspace refine and despite having provided the CIF file (elbow generated alkyl-ammonium form), the molecule keeps "exploding". Instead, the 'NH2 form' (alkyl-amine) remains stable during refinement and most of the pdb entries having alkylammonium molecules have modeled them as the amine form even when they are expected to be positively charged. Thus, it seems to me that other groups may have gone through the same difficulties when trying to refine ammonium groups. We have already tried: * Ligand build in Lidia and tried prodrg with error (*Unrecognised REFMAC atom type 'HCH3' in input*) * Refmac Acedrg (run201 Error in wrapper refmac) Please help me fix this. I could refine using the amine form, but I know that the amine form is not right. Thank you very much, in advance. -- Promod Jaykar, Ph.D.