Modeller mismatch failure

[jmsstarlight@…]

I still have some problems with the reconstructions of the loop fragments for the structures containing ligands. I've just tested the latest ChimeraX (1.9) and the problem is still there: where I fetch a PDB with the ligand and use it SEQRES to reconstruct the missed loop it produces the error indicating the sequence mismatch. When I remove all ligands, everything is OK.

Example 1 - no problems with loop reconstructions
open 8OU9 ( the chain C contains 5 missed residues)
sequence chain #1/C
modeller refine 1/C:1:all-missing numModels 3 fast true adjacentFlexible 1 protocol standard

Everything works OK - in spite of the presence of the ligand and ZN in the chain C

Example 2 - the similar system with ligand and ZN, which has a problem in loop reconstruction with modeller

open 6BN7 (the chain B contains a gap of 9 residues)

Here is the example I've just tested with a new pdb where the chain B contains the 9 a.a missed fragments

open 6BN7
sequence chain #1/B
modeller refine 1/B:1:230-238 numModels 3 fast true adjacentFlexible 1 protocol standard

this produces the following error

modeller refine 1/B:1:230-238 numModels 1 fast true adjacentFlexible 1 protocol standard:

Dynamically allocated memory at amaxstructure [B,KiB,MiB]: 1514284 1478.793 1.444
read_te_290E> Number of residues in the alignment and pdb files are different: 1305 1178
For alignment entry: 1 6bn7_1
x (mismatch at alignment position 1179)
Alignment DMSPQKFWGLTRSALLPTIP/SMNPPPPETSNPNKPKRQTNQLQYLLRVVLKTLWKHQFA
PDB DMSPQKFWGLTRSALLPTIP/
Match *

I've tried to assosiate the sequence with the structure and it did not find any mismatches:

Disassociated 6bn7 chain B from chain B
Associated 6bn7 chain B to chain B with 0 mismatches

One I removed all ligands- everything works OK:

delete ~protein
modeller refine 1/B:1:230-238 numModels 1 fast true adjacentFlexible 1 protocol standardWebservices job id: 602MBGXI5OKGYO2Z
Modeller job (ID 602MBGXI5OKGYO2Z) finished
Computing secondary structure

[Eric Pettersen]

Fixed. Fix available in next daily build and 1.9 release candidate.

Unlike almost every other PDB entry, this entry has its ions/ligand for chain B after _all_ chains instead of after chain B.

Fix: ​https://github.com/RBVI/ChimeraX/commit/95ea75bd26c666f8dbaa3fbcaa71d219d1b6bc5d