C-Terminal proline in AMOEBA
Posted: Sat Apr 01, 2017 6:51 pm
Hi,
A system I'm working with has a C-terminal proline on one of the chains. I noticed in an old github post (https://github.com/pandegroup/pdbfixer/issues/112), that the type for HA needed to be changed from 55 to 54. Even after that correction, however, I notice the following issue:
The multipole reference frame definition for type 51 (CA) is
multipole 51 50 52 54
This works fine for regular PRO, where
CA is type 51, N is type 50, C is type 52, and HA is type 54.
However, for CPRO,
CA is type 51, N is type 50, C is type 233 and HA is type 54. Hence the corresponding multipole definitions are missing. I checked the original amoebapro13.prm from TINKER, and the same issue is present there as well. Would it be appropriate to make a copy of the multipole parameters of the CA from PRO and use them for the CA from CPRO, or would the electrostatic environment be different enough between a mid-sequence PRO and a CPRO that this would not be a good option? In that case, would new parameters need to be derived for that CA for AMOEBA itself?
Thanks.
A system I'm working with has a C-terminal proline on one of the chains. I noticed in an old github post (https://github.com/pandegroup/pdbfixer/issues/112), that the type for HA needed to be changed from 55 to 54. Even after that correction, however, I notice the following issue:
The multipole reference frame definition for type 51 (CA) is
multipole 51 50 52 54
This works fine for regular PRO, where
CA is type 51, N is type 50, C is type 52, and HA is type 54.
However, for CPRO,
CA is type 51, N is type 50, C is type 233 and HA is type 54. Hence the corresponding multipole definitions are missing. I checked the original amoebapro13.prm from TINKER, and the same issue is present there as well. Would it be appropriate to make a copy of the multipole parameters of the CA from PRO and use them for the CA from CPRO, or would the electrostatic environment be different enough between a mid-sequence PRO and a CPRO that this would not be a good option? In that case, would new parameters need to be derived for that CA for AMOEBA itself?
Thanks.