I think dave is best suited to answer these questions. But he is busy,
so I take a shot ...
In sander (and other programs), the "atom" indices in the bonding arrays
are not the atom numbers. They are the pointers to the 3-D coordinate
arrays. In other words, it points to the X-coordinate of that atom
within the array. So, atom one would have an index of 0, atom two would
have an index of 3, and so on. This dated back to the first version of
AMBER when 64K bytes of memory was considered "huge" (well, relatively).
In the new format, because of the backward compatibility concern (to the
existing code) and because not that many people want to re-write this
part of the program, the simple solution is to keep it that way. But
"I6" can not accommodate large systems. A nice thing about the new
format is that it allows future expansion (because the format
information is provided). However, on our way, we probably forgot to
update the manual.
yong
> -----Original Message-----
> From: owner-amber.scripps.edu
> [mailto:owner-amber.scripps.edu] On Behalf Of Herbert Georg
> Sent: Friday, October 24, 2003 10:38 AM
> To: amber.scripps.edu
> Subject: AMBER: prmtop file format
>
>
> I was taking a look at the prmtop file format description in
> the amber7 manual
> and in the amber site but it seems that my prmtop file
> doesn't match the
> description. It seems that some new flags were added and the
> format of the flags
> has changed. Could someone help me out with this new format?
> For example, in the flag BONDS CONTAINING HYDROGEN the manual
> and the site have
> the following description:
>
>
> FORMAT(12I6) (IBH(i),JBH(i),ICBH(i), i=1,NBONH)
> IBH : atom involved in bond "i", bond contains hydrogen
> JBH : atom involved in bond "i", bond contains hydrogen
> ICBH : index into parameter arrays RK and REQ
>
>
> and my prmtop file has the following for that flag:
>
>
> %FLAG BONDS_INC_HYDROGEN
> %FORMAT(10I8)
> 0 3 1 6 9 1 18
> 21 5 18
> 24 5 18 27 5 30 33
> 5 30 36
> 5 30 39 5 42 45 1
> 48 51 1
> 57 60 7 63 66 7 69
> 72 10
>
>
> According to the description above, there is an atom numbered 'zero'.
> Furthermore, there are atoms numbered above 26 which is the
> total number of atoms.
>
>
> --------------------------------------------------------------
> ---------
> The AMBER Mail Reflector
> To post, send mail to amber.scripps.edu
> To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
>
-----------------------------------------------------------------------
The AMBER Mail Reflector
To post, send mail to amber.scripps.edu
To unsubscribe, send "unsubscribe amber" to majordomo.scripps.edu
Received on Fri Oct 24 2003 - 16:53:01 PDT
