Re: [modeller_usage] DOPE profile of target structure
To: Hanudatta Atreya <>
Subject: Re: [modeller_usage] DOPE profile of target structure
From: Modeller Caretaker <>
Date: Wed, 19 Mar 2008 09:13:05 -0700
Cc:
Hanudatta Atreya wrote:
Should the residue-wise DOPE profile of the target modeled structure
necessarily match that of the template? This is because in the tutorial
online, the DOPE profile shown indeed follows very closely that of the
template.
No - since the profiles are derived from the DOPE energy, the absolute
values cannot be compared between different structures. But you can
qualitatively compare the rough shape of the profile between template
and target to give a visual indication of your alignment quality (if the
shapes are similar but shifted, your alignment probably is also shifted).
In my case, the modeled structure has 1.0 Angstrom RMSD with the
template. However, the residue wise DOPE profile do not match with that
of template.
It makes no sense to compare your model with the template, because it
will always be fairly similar by construction. And the DOPE profile, as
explained above, is not conclusive.
Also, other structure quality scores as I checked with
PROCHECK, MOLPROBITY etc. are poor. So does this indicate non-optimal
modeling? Especially the Clash-scores are high. Should I carry out one
more mimimization step myself to loosen the distortions?
If assessment methods indicate a poor model, by all means try to
optimize it some more. But you may get better results by simply building
more models and ranking them.