This is getting a little garbled, I seem to have lost track of one of your replies due to post nesting. We’re beyond wild guess territory into “I’m making stuff up based on fragmentary understanding of glances at code I didn’t write”.
It sounds like you should have 26 residues, and your Conformation object has 26 residues, but your FoldTree has 32. The culprit is almost certainly the jump_def line at the top of the residue_pair_whatever file (notice the 32).
Poking around in the (3.1) code for the flag jumps::residue_pair_jump_file, I track the code to protocols/jumping/ResiduePairJumpSetup.cc, function read_file. It’s not likely changed in 3.2.
It looks like it parses the jump_def line as jump start, jump end, cut start, cut end.
Jumps define nonbonded connections between atoms; here it is tethering the metal to the rest of the protein (Rosetta doesn’t consider metal-ligand bonds.) So, your cut start/cut end will be 25.
Your jump end will be the metal (26).
Your jump start will be any of the cysteines; they appeared to use the first one in the sample.
You’ll also need to modify the ZNs to FEs in that file; the V1/V2 virtual atoms are probably okay. You may need to modify the geometry.