[gmx-users] Problems with pdb2gmx simulating a iron-sulfur cluster

[martinez]

Hi everybody

Thank you very much Tsjerk. I'm sorry for asking all these questions 
but some points are still unclear to me, and I didn't find much 
information on the mailing list archives or in the manual that could 
help me.

Grompp is complaining about the angles because in the .top file there 
are no entries for the angle type, because pdb2gmx just creates the 
bond, but doesn't specify the type like you told me.
I've edited the ff...bon.itp file as follows but it seems useless.
; bond-, angle- and dihedraltypes for specbonds:
[ bondtypes ]
S      S       2    gb_33
NR     FE      2    gb_32
; cystine - heme link (is CR1-S, use CH2-S):
S      CR1     2    gb_30
; cystine - fs link:
FE     S       2    gb_48

[ angletypes ]
CH1    CH2    S     2   ga_15
CH2    S      S     2   ga_5
CR1    NR    FE     2	ga_33
NR     FE    NR     2   ga_16
; cystine - heme link (is CH2-S-CR1, use CH2-S-CH2):
CH2    S     CR1    2   ga_3
; cystine - heme link (is S-CR1-C/CH2, use CHn-CH2-S):
S      CR1   C      2   ga_15
S      CR1   CH2    2   ga_15
; cystine - fs link
S      FE    S      2   ga_47
FE      S    FE     2   ga_48
etc...
where I specified ga_47, ga_48, gb_48...
Is there a way that grompp could take these types or do I have to edit 
the topology file manually to specify the type?

It seems strange to me because there are also no entries for the bond 
types, and grompp doesn't give me any warnings for that, why?

Thank you again

Nicolas