Welcome, Guest. Please Login
YaBB - Yet another Bulletin Board
 
  HomeHelpSearchLogin  
 
Spidermonkey questions (Read 1925 times)
Alex_Wong
YaBB Newbies
*
Offline


Curious HyPhy user

Posts: 1
Spidermonkey questions
Aug 24th, 2012 at 3:01pm
 
Hello,

I am potentially interested in using Spidermonkey to look for co-evolution between proteins. Before I dive in too deeply, I have a few questions -
(1) Are there any in-principle concerns with using Spidermonkey on multiple proteins instead of just one? Presumably I could input a concatenated alignment to do so?
(2) Is there a minimum number of taxa that are recommended to see a signal if there is one?
(3) Again on the topic of power - given the choice between closely related taxa or more distantly related taxa, is there a preference?

Any guidance would be much appreciated! Thanks,
Alex Wong
Back to top
 
 
IP Logged
 
Sergei
YaBB Administrator
*****
Offline


Datamonkeys are forever...

Posts: 1658
UCSD
Gender: male
Re: Spidermonkey questions
Reply #1 - Aug 28th, 2012 at 9:48am
 
Hi Alex,

In response to your queries:

1). No, so long as the proteins have the same gene trees. Spidermonkey assumes a single phylogeny for all sites.
2). Spidermonkey derives power from co-occuring substitutions, which are a function of the number of taxa and sequence divergence. I would say that you need at least 5 substitutions at a site to gain good power.
3). Basically you want to maximize the number of substitutions without losing homology. I would include a mix of distantly and closely related taxa -- the former give you power, the latter help map substitutions more accurately.

Sergei
Back to top
 

Associate Professor
Division of Infectious Diseases
Division of Biomedical Informatics
School of Medicine
University of California San Diego
WWW WWW  
IP Logged