I'm using ChimeraX with the Modeller plugin to fill in missing residues for my GPCR structure, specifically focusing on helix 8. However, Modeller keeps trying to reconstruct experimental mutations that were added to the N-terminus in the original PDB, even though I manually removed those residues from both the SEQRES and ATOM records in the PDB file. I also tried providing a cleaned FASTA sequence without the modifications, but the issue persists. Is there a way to force ChimeraX/Modeller to completely ignore the experimental N-terminal mutations and only model the missing residues in helix 8?