[gmx-users] wrong charge assignment(?) on ATP in rtp file

[chiba.s.ac]

Dear gmxusers,

I am going to carry out an MD simulation using G53a6 FF, whose system contains some ATP molecules.  In order to know whether the calculation is possible or not, I looked for a topology of ATP in ffG53a6.rtp.  Although the topology exists, each total charge of charge group 9, 10 and 11 is not a whole number (-0.925,　-0.925 and -1.15 respectively).  It seems to be wrong.

My questions are as follows:
1) How can I fix them, otherwise can I get the appropriate topology.
2) There are not only ATP but many residues which are not listed in the original paper of G53a5/6 “23 (13) 1656-1676, Journal of Computational Chemistry”.  How were these parameters derived?　Is it consistent with the method described in the paper, i.e. are charges fitted to reproduce the corresponding experimental value?

I really appreciate all of replies.

Best regards,
Shuntaro.