Hi All, I am refining several modified nucleotides in ribosomal RNA (N2-methyl-guanosine and similar). While the restraints generation with phenix.reel work quite well, I am unable to get the phosphate oxygen restraints applied properly. Phenix.refine recognizes the links but for some reason the phosphate groups come out quite distorted... For example, angles after refinement: O3 (1) - P (2, modified) - OP2 (2) = 80 deg and O3 (1) - P (2, modified) - OP1 (2) = 127 deg Here is part of the phenix.refine log on reading the coordinates: Chain: "A" Number of atoms: 32526 Number of conformers: 1 Conformer: "" Number of residues, atoms: 1513, 32526 Unusual residues: {'MA6%rna3p': 2, '5MC%rna3p': 3, '5MC%rna2p': 1, '7M G%rna3p': 1, '2MG%rna3p': 1, 'UR3%rna2p': 1, 'M2G%rna3p': 1, '4OC%rna3p': 1} Classifications: {'undetermined': 11, 'RNA': 1502} Modifications used: {'5*END': 1, 'rna3p': 1318, 'rna2p': 195} Link IDs: {'rna3p': 1318, 'rna2p': 194} Chain breaks: 1 I am at a loss about how to fix this as all links and restraints seem to be in place. Many thanks for advice, Gerwald