errors with pdbxyz and analyze

[qinghualiao]

Hello,

I have a system of one Zn2+ atom in a water box (404040 A3) with two Cl- as counterions in pdb format.
I used pdbxyz (Tinker 25.5.3, on Linux) to convert the pdb file to xyz format.

Initially, the atom type for Zn2+ was assigned as 0, 82 for O of water, 11 for H of water, and 60 for Cl-. The parameter file used is amoebabio18.prm.

As I understand, pdbxyz does not deal well with ions and waters, so I manually changed the atom types:
102 for Zn2+, 90 for O of water, 91 for H of water, and 104 for Cl-.

However, when I used analyze to assess the xyz files, it crashed with errors:

Angle 6243-C 6242-H 6244-C 22 9 22
Angle 6246-C 6245-H 6247-C 22 9 22
Angle 6249-C 6248-H 6250-C 22 9 22
Angle 6252-C 6251-H 6253-C 22 9 22
Angle 6255-C 6254-H 6256-C 22 9 22
free(): invalid next size (normal)

Program received signal SIGABRT: Process abort signal.

Backtrace for this error:
#0 0x1527c22e532a
#1 0x1527c22e4503

It seems that O atoms of water were treated as H (atom type 9, Threonine HG2), while H atoms of water were treated as C (atom type 22, Tryptophan CZ3). Obviously, there is no parameter for bond angle 22-9-22 (C-H-C)

I really don't know how this would happen. The attached file is the corrected xyz file.
I really appreciate if you could help me figure it out.

Zn_Amoeba_solv.xyz.zip

All the best,
Qinghua

[jayponder]

Hi, This is easy to fix. In your message, you say you manually used 90 for Water O, 91 for Water H, 102 for Zn+2 and 104 for Cl-. But those are not correct. You have used the "class" numbers, but the Tinker XYZ file should have the "type" numbers. So you want 349 for Water O, 350 for Water H, 361 for Zn+2 and 363 for Cl-. The corresponding atom parameter lines from amoebabio18.prm are shown below. The first number on each line is the "type", and the second number is the "class".

atom 349 90 O "AMOEBA Water O" 8 15.999 2
atom 350 91 H "AMOEBA Water H" 1 1.008 1
atom 361 102 Zn2 "Zinc Ion Zn+2" 30 65.380 0
atom 363 104 Cl- "Chloride Ion Cl-" 17 35.453 0

[qinghualiao]

Dear Prof. Jay Ponder,

Thanks so much! Now it works well.
It is my first setup with AMOEBA force field, I used the atom class as atom type.

All the best,
Qinghua