Hi. I came across with the following post when I was dealing with a similar issue (adding a modified nucleotide to RNA). I could create .cif using elbow/reel as well, but I was wondering where to define the link type between residues in phenix.

You can use the apply_cif_link parameters.

For RNA, I typically see "rna3p', but with my newly created residue linked to normal oligo, the link is formed, but recognized as Link ID 'p'. Also, the newly created molecule is recognized as 'peptide', although I defined _chem_comp.group in .cif file as RNA.

The _chem_comp.group value is ignored by phenix.refine. The classifications are obtained via analysis of the atom names and the bonding in the monomer definitions.

------------------------------------------ Chain: "Q1"       Number of atoms: 74       Number of conformers: 1       Conformer: ""         Number of residues, atoms: 3, 74           Classifications: {'RNA': 2, 'peptide': 1}           Modifications used: {'5*END': 1, 'rna3p': 2}           Link IDs: {'p': 1, 'rna3p': 1} ------------------------------------------

It is strange that it thinks there is a peptide but still applies two rna/dna links. Could you send me (not the list) the part from the pdb file with the Q1 chain? Ralf