To 
Ed Pozharski, 
Thanks for suggestion. I have checked the electron density of all the residues 
of the refined pdb ( till now, with R(work)/ R(free) = 0.24/0.30) and also the 
ones showing as "Ramachandran outliers" in different loop region. it seems the 
residues fit well to the electron density. But away from the loops, in a alpha 
helical region some electron density is still there close to two residues, even 
though the two residues are fitted well and placed within the allowed region in 
Ramachandran plot. Can that be the cause of four residues (in in four different 
loop region) to be observed as outliers in Ramachandran plot ?? 
To 
Monica Mittal 
Hii, thanks for your advice. i have tried that once, but i end up with the same 
problem again. If i look at the residues of the loop region including the 
"Ramachandran outliers", it is fitting well to the electron density. As i have 
explained to Ed Pozharski, can residues outside loop region responsible for 
that?? 



----- Original Message -----

From: "Monica Mittal" <[email protected]> 
To: "Ajit Roy" <[email protected]> 
Sent: Monday, April 20, 2015 1:32:46 PM 
Subject: Re: [ccp4bb] loop refinement 



Hi 
Just an advice why dont you delete the loop regions and refine your structure 
and try tracing the difference fourier density to guide you the correct path !! 
And den start filling your loop regions one by one and refining iteratively 
with each addition. 
HTH 
On Apr 18, 2015 3:55 PM, "Ajit Roy" < [email protected] > wrote: 


Dear all, 

While refining the structure of a protein I’m facing problems with some 
residues present in the loop region of the protein. There are four residues in 
four different loops in the protein which is always showing as outliers in 
Ramachandran plot even after multiple steps of refinement (i.e. even 93% of 
residues of the refined protein are in preferred region). 
The way i am refining the loops is by using the module; “fit loop” in Phenix, 
through which I am fitting residues in single loop followed by refinement, and 
then reiterating the same process for three more other loops containing the 
Ramachandran outliers. 

The problem I’m facing in the following process is that: after refining a 
single loop, the corresponding residue within the loop (i.e. Ramachandran 
outlier) is showing under the preferred region but, after using the refined pdb 
(i.e. obtained after the first loop refinement ) for second loop refinement, 
the residue (present in the 1st loop) is again showing as outliers after 2nd 
refinement, while the residue (i.e. Ramachandran outlier) in 2nd loop is 
showing within the preferred region. In this way I am only able to fix a single 
residue (i.e. Ramachandran outlier) in a single loop, or to a maximum of two 
with a different combinations of refinement (for example: refinement of loop 1 
followed by loop 3 instead of loop 2 and so on …) is there any other approach 
to address the problem ?? 

Is there any way in which multiple loops can be refined in a single time ?? 

Any suggestions to address this problem is deeply appreciated. 

P.S. For identifying the loops in protein i used secondary structure alignment 
with sequences of similar proteins, whose structures are known and matching 
with the hitherto refined structure of the protein and the ramachandran 
outliers are in the middle of the loop. 

As a novice my sincere apology, if the description of the problem is not clear 
enough for you to understand. 




Reply via email to