Re: [modeller_usage] strategies for model refinement
To: Thomas Evangelidis <>
Subject: Re: [modeller_usage] strategies for model refinement
From: Mensur Dlakic <>
Date: Mon, 10 Jan 2011 18:21:26 -0700
Cc:
Thomas,
I never really modeled a protein where loop conformation was crucial for
the model itself, so I am usually content to refine the loops so they are
stereochemically acceptable (I hope nobody reading this is cringing in
horror at my relative indifference towards loop modeling). RAPPER for sure
does a good job of improving Ramachandran outliers for loops up to 7-8
residues. The way it does it by "freezing" the protein outside of requested
regions except for one "anchor" residue abutting the region and then
samples the conformational space while making sure that rotamers with
correct stereochemistry are utilized. I usually build 100 loops for each
region and the program picks the one with lowest energy automatically,
which is another reason I use it more than MODELLER for this purpose. For
loops that are 6 or so residues, this takes less than two minutes.
I would not try modeling loops that are 10+ residues, and in fact I usually
omit them from my models if I have long insertions.
Finally, these papers may (or may not, I haven't read them recently) offer
the benchmark you were asking about:
Mensur, are you aware of any benchmark analysis that shows RAPPER's
superiority on loop modeling over MODELLER? I am aware of one that
compares Rosetta, MODELLER and CABS on loops up to 24 aa if I remember
correctly, and shows that in short lengths all 3 programs are equivalent,
but for longer loops the combination of MODELLER with CABS prevails.
I have several proteins with missing regions of varying length (8-47 aa)
which I want to fold. I have tried PyRosetta and MODELLER in the past but
I'm not impressed from the results of loop modeling, especially as the
length increases. I guess these routines were designed to model flexible
animo acid stretches that's why the predicted conformations adopt coiled
coils. To this end I also tried I-TASSER server (which employs Monte Carlo
with Replica Exchange for regions with no templates and is questionably
more accurate in folding long aa stretches) by supplying my initial
structure and excluding all homologues, but the meta-threading program
(LOMETS) it implements always detects traces of "homology" to some
irrelevant structures and uses them as templates. I guess my last resort
is MD with Replica Exchange but I haven't found the time yet to set up my
system and figure out a way to keep the rest of the model rigid, namely to
fold only the missing regions.
This message is slightly off-topic, but I just wanted to share my
experience (and desperation) with other people that might find it helpful
(I hope not the desperation).
Thomas
==========================================================================
| Mensur Dlakic, PhD | Tel: (406) 994-6576 |
| Department of Microbiology | Fax: (406) 994-4926 |
| Montana State University | |
| 109 Lewis Hall, P.O. Box 173520 | http://myprofile.cos.com/mensur |
| Bozeman, MT 59717-3520 | E-mail: |
==========================================================================