You are here

Question about homology modelling

4 posts / 0 new
Last post
Question about homology modelling
#1

Hello, everyone!

I have maybe quite a silly question.  What is the best methodic of homology modeling when my template and target have quite different lengths. Like target has 350 and the second has 260. Shall I just get rid of some tail part of longer protein during alignment? And yes, it seems that this is the best homolog for my type of protein.

Post Situation: 
Sun, 2016-08-21 13:44
DmitriiN

Having more residues in the target than template (by this degree) indicates that the target has something else going on the template didn't.  Presumably it either has another domain or two, or it has large flexible tail regions.  Anything not present in the template should be removed from homology modeling, as there's no homology there.  

Does the missing sequence show up as homlogous to another target?  You can solve the domains separately and assemble them.  I'm not sure what the current best domain assembly tool is.

Does the missing sequence show little secondary structure propensity?  It's probably not folded to begin with.  You can use FloppyTail to sample its possible envelope of conformations.

Sun, 2016-08-21 16:23
smlewis

Yes, the two proteins have 7 alpha-helices in common. And the huge tail for the bigger one. 
The problem is that even if I use all the best targets I have now, they still have very similar alpha-helices part, and they all don't have this tail.

So I guess I need to first understand where the tail part starts, then cut it from the sequence and
predict the main part of the target, the alpha-helices. And then use FloppyTail to predict the tail conformation.

But FloppyTail also need some structure of the tail to start modeling. How to get it? Because I am afraid if I separately model the whole protein with tail first and then only main part, I won't be able just to copy paste tail part from PDB to PDB. Maybe I will need vmd to attach the tail..

Mon, 2016-08-22 06:22
DmitriiN

You should do homology modeling on the homologous part and consider the results before deciding on the next step.

As far as re-attaching the tail goes - I'm told the "remodel" application can add arbitrary sequence to add on the tail.  

When I faced this problem for the original FloppyTail case, I wrote a short bit of code "Extender".   I can pass it along to you if you can't get remodel to work.  (Extender  isn't part of the normal release because it doesn't meet code quality standards.  Also A) it's trivial to rewrite it in PyRosetta, and B) I never finished it; it only works for C-terminal tails.  It's pretty likely that Extender is already attached to some message board post on these forums already.

Mon, 2016-08-22 10:28
smlewis