Hi Daniel
I was just reading this mail. I have some problem with my reference
structure PDB and trajectory for which RMSD to be calculated.
>> No, as long as the number of atoms and the atom ordering in your target
and reference structures are the same.
I thought the RMS fitting can take care. Actually AMBER does not complain
about atom names and ordering/sequence. So do I need to keep same atom
names and ordering?
e.g. comparing say first few atoms in structure to be compared (in
trajectory) has this sequence
ATOM 1 O5' DG 1 88.750 97.070 107.880 1.00
0.00
ATOM 2 H5T DG 1 89.730 97.100 108.050 1.00
0.00
ATOM 3 C5' DG 1 88.150 98.370 108.180 1.00 0.00
ATOM 4 1H5' DG 1 88.610 99.040 107.590 1.00
0.00
ATOM 5 2H5' DG 1 88.360 98.560 109.130 1.00 0.00
and the reference structure has this:
ATOM 1 HO5' DG5 1 12.843 -51.888 -5.899 1.00
0.00 H
ATOM 2 O5' DG5 1 13.094 -52.038 -5.621 1.00
0.00 O
ATOM 3 C5' DG5 1 13.037 -51.978 -4.813 1.00
0.00 C
ATOM 4 H5' DG5 1 13.096 -51.378 -4.805 1.00
0.00 H
ATOM 5 H5'' DG5 1 13.617 -52.139 -4.669 1.00
0.00 H
Thanks
On Mon, Sep 8, 2014 at 2:54 PM, Daniel Roe <daniel.r.roe.gmail.com> wrote:
> Hi,
>
> On Sun, Sep 7, 2014 at 8:52 PM, Yip Yew Mun <yipy0005.gmail.com> wrote:
> > Hi, I recently ran a simulation of a small alpha helix (~12 residues)
> with REMD and I did a RMSD analysis using just the backbone atoms (N,CA,C).
> Usually for simulations of large proteins, an RMSD of 3 Angstroms is
> usually enough to quantify that there are no major conformational changes.
> However, I realised for small protein systems like mine, an RMSD of 3
> Angstroms isn’t a good criterion to say that I have folded my protein
> successfully. Therefore, I wish to ask:
> >
> > 1) Does the number of atoms affect the RMSD calculation?
>
> For folded conformations not really. Average RMSD is not correlated
> with chain length, although the distribution tends to get narrower for
> longer chains.
>
> > 2) Is there a need for the reference structure to be protonated before
> using it for RMSD analysis since in this case, I’m just using the backbone
> atoms (N,CA,C).
>
> No, as long as the number of atoms and the atom ordering in your
> target and reference structures are the same. Jed Pitera has recently
> done some very interesting work looking at intrinsic characteristics
> of RMSD distributions that I recommend checking out:
> http://pubs.acs.org/doi/abs/10.1021/jp412776d
>
> Hope this helps,
>
> -Dan
>
> >
> > Thanks.
> >
> > Yip Yew Mun (Mr) | PhD Research Scholar | Division of Chemistry &
> Biological Chemistry
> > School of Physical & Mathematical Sciences | Nanyang Technological
> University | Singapore 639798
> > Tel: (+65) 97967803 | Email: yipy0010.e.ntu.edu.sg | GMT+8h
> >
> > _______________________________________________
> > AMBER mailing list
> > AMBER.ambermd.org
> > http://lists.ambermd.org/mailman/listinfo/amber
>
>
>
> --
> -------------------------
> Daniel R. Roe, PhD
> Department of Medicinal Chemistry
> University of Utah
> 30 South 2000 East, Room 307
> Salt Lake City, UT 84112-5820
> http://home.chpc.utah.edu/~cheatham/
> (801) 587-9652
> (801) 585-6208 (Fax)
>
> _______________________________________________
> AMBER mailing list
> AMBER.ambermd.org
> http://lists.ambermd.org/mailman/listinfo/amber
>
_______________________________________________
AMBER mailing list
AMBER.ambermd.org
http://lists.ambermd.org/mailman/listinfo/amber
Received on Wed Sep 17 2014 - 08:00:04 PDT
