Domain insertion with Non-CAA

Hi all,

I am trying to insert a long fragment (GFPStrand) into another protein (4e1s) using Remodel. I've delete the residues that were originally in 4e1s but left a few at the end as an anchor for the domain insertion. Remodel throws an error about insertion indices being 0 if I don't leave them to anchor it. Remodel accepts the domain insertion and starts on the loop closure. It also appears to close the loop, but then throws this error with trace: 

protocols.forge.remodel.RemodelMover: 
protocols.moves.MonteCarlo: MC: 2  16.1898  13.0953  16.1898  13.0953  0  0  0  accepted score beat last
protocols.moves.TrialCounter:         ccd_move trials=    225;  accepts= 1.0000;  energy_drop/trial=  -0.02888
protocols.moves.TrialCounter:             frag trials=   1660;  accepts= 0.5687;  energy_drop/trial=   0.00549
protocols.forge.remodel.RemodelMover: ** boost_closure_stage
core.chemical.ResidueType: Residue dstor
core.conformation.util: copy_residue_coordinates_and_rebuild_missing_atoms: missing atom ALA:protein_cutpoint_lower OVL1
core.conformation.util: copy_residue_coordinates_and_rebuild_missing_atoms: missing atom ALA:protein_cutpoint_lower OVL2
core.chemical.ResidueType: Residue dstor
core.conformation.util: copy_residue_coordinates_and_rebuild_missing_atoms: missing atom TYR:protein_cutpoint_upper OVU1
loops: adding loop LOOP start: 226  stop: 249  cut: 249  size: 24  skip rate: 0  extended?: True
loops: 
core.chemical.ResidueType: Residue dstor
core.conformation.util: copy_residue_coordinates_and_rebuild_missing_atoms: missing atom ALA:protein_cutpoint_lower OVL1
core.conformation.util: copy_residue_coordinates_and_rebuild_missing_atoms: missing atom ALA:protein_cutpoint_lower OVL2
core.chemical.ResidueType: Residue dstor
core.conformation.util: copy_residue_coordinates_and_rebuild_missing_atoms: missing atom TYR:protein_cutpoint_upper OVU1
protocols.forge.remodel.RemodelMover: at least one loop is beyond tolerance
protocols.forge.remodel.RemodelMover: * 1 / 1   closed / attempts 
core.chemical.ResidueType: Residue dstor
core.chemical.ResidueType: atom name :  HE1 not available in residue CRO
core.chemical.ResidueType: ' CA2' 0x7fca92e844c0
core.chemical.ResidueType: ' CB2' 0x7fca92e84810
core.chemical.ResidueType: ' CG2' 0x7fca92e84b80
core.chemical.ResidueType: ' CD1' 0x7fca92e84f50
core.chemical.ResidueType: ' CE1' 0x7fca92e851f0
core.chemical.ResidueType: ' CZ ' 0x7fca92e85710
core.chemical.ResidueType: ' CE2' 0x7fca92e85970
core.chemical.ResidueType: ' CD2' 0x7fca92e85c20
core.chemical.ResidueType: ' OH ' 0x7fca92e85f00
core.chemical.ResidueType: ' C2 ' 0x7fca92e86570
core.chemical.ResidueType: ' O2 ' 0x7fca92e86800
core.chemical.ResidueType: ' N3 ' 0x7fca92e86ae0
core.chemical.ResidueType: ' CA3' 0x7fca92e86dc0
core.chemical.ResidueType: ' C  ' 0x7fca92e870a0
core.chemical.ResidueType: ' O  ' 0x7fca92e87380
core.chemical.ResidueType: ' C1 ' 0x7fca92e87660
core.chemical.ResidueType: ' CA1' 0x7fca92e87940
core.chemical.ResidueType: ' CB1' 0x7fca92e85500
core.chemical.ResidueType: ' CG1' 0x7fca92e862d0
core.chemical.ResidueType: ' OG1' 0x7fca92e87c60
core.chemical.ResidueType: ' N  ' 0x7fca92e87f40
core.chemical.ResidueType: ' N2 ' 0x7fca92e88220
core.chemical.ResidueType: 
core.kinematics.FoldTree: slide_cutpoint: current= 246 target= 249 FOLD_TREE  EDGE 1 225 -1  JEDGE 225 250 1 N C  END  EDGE 250 251 -1  EDGE 225 246 -1  EDGE 250 247 -1 slide_cutpoint: final tree= FOLD_TREE  EDGE 1 225 -1  JEDGE 225 250 1 N C  END  EDGE 250 251 -1  EDGE 225 249 -1 
ERROR: unknown atom_name: CRO   HE1
ERROR:: Exit from: src/core/chemical/ResidueType.cc line: 3540

I'm relatively certain it's trying to go from centroid back to full atom, but I'm not sure. My fragment pdb is full atom, including hydrogens on the CRO residue. rmoretti had attached a centroid version of CRO somewhere else in the forums that I've supplied as a flag and am reattaching below. The full atom version is in residue_types/extra/CRO.params. 

Any ideas what I'm missing? If I don't need those anchoring residues, that would be great too. Is it just too large a "domain"? My full set of flags are below. Thanks!

Meghan

-s 4e1sShort.pdb
-remodel:blueprint 4e1sShort.blueprint
-extra_res_fa /extra/CRO.params
-extra_res_cen CRO.cen.params
-run:chain A
-remodel:domainFusion:insert_segment_from_pdb GFPStrandRenumbH.pdb
-remodel:RemodelLoopMover:bypass_closure #If I don't have this flag, it makes 50 CCD attempts and seg faults
-remodel:RemodelLoopMover:boost_closure_cycles 0 #Changing this number doesn't affect my error.
-ex1
-ex2
-use_input_sc
-out:file:scorefile remodel.fasc
-in:file:spanfile 4e1s_A_spansShort.txt #This gets read in easily, so the membrane part isn't the problem
-membrane:no_interpolate_Mpair
-membrane:Menv_penalties
-membrane:Mhbond_depth True
-membrane:thickness 12.5
-score:weights membrane_highres_Menv_smooth
-mute core.scoring.MembraneTopology
#-out:levels protocols.loops:trace
-out:levels all:trace