HyPhy message board
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl
Theoretical questions >> Sequence Analysis >> Evolutionary Fingerprint
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl?num=1308915894

Message started by beyondblueskies on Jun 24th, 2011 at 4:44am

Title: Evolutionary Fingerprint
Post by beyondblueskies on Jun 24th, 2011 at 4:44am
Dear Sergei,

In Evo-fingerprint analyses, the circles above the line corresponding to dn=ds are under positive selection, while those bellow are negatively selected. But does the intensity of colouration depend on the confidence with which we can conclude that the points belong to that category or the strength of positive selection? Can I interpret a plot where a lot of points are above the line but are diffused throughout (do not form clear-cut circles) as weak positive selection or low-confidence in assigning these sites to a particular class? In short, I wanted to know if I can say one category is under greater strength of positive selection than the other based on the intensity of the colour.

Thank you,
David

Title: Re: Evolutionary Fingerprint
Post by Sergei on Jun 24th, 2011 at 11:17am
Hi David,

The intensity of the color represents site density, i.e. brighter dots on the plot correspond to higher posterior density. In other words, the brighter the dot, the more frequently we expect to find sites with that combination of dS and dN. Confidence in our estimates is represented by the widths of the ellipses. Tight bright circles represent accurate estimates, while diffuse wide ellipsoids -- less confident.

In your case, you can interpret the plot as having evidence of positive selection, but with low confidence in estimating it's strength (probably because there are only few sites contributing to this inference).

Can you attach the plot, so that I can ensure I am not mis-directing you?

Sergei

Title: Re: Evolutionary Fingerprint
Post by beyondblueskies on Jun 24th, 2011 at 12:18pm
Hi Sergei,

So, I can not say 2nd plot has greater selection than the first (file attached). But, can I say that the 2nd plot has more number of sites under positive selection than the first?

I was just wondering if the site, branch and branch-site models implemented in HyPhy and PAML are there just to identify the presence or absence of positive selection or one can compare  the omega estimations as well (to find out which lineage or group has stronger selection pressure than the other).

Thanks for the help :)
David
Untitled_001.png (131 KB | )

Title: Re: Evolutionary Fingerprint
Post by Sergei on Jun 24th, 2011 at 2:34pm
Hi David,

If you are seeking to compare the _strength_ of selection between data sets of lineages, this is probably best done directly, using something like what is described in section 2.7 of Multimedia File Viewing and Clickable Links are available for Registered Members only!!  You need to Login Login

Based on the plots, it does seem like more sites have dN>dS in data set two.

Perhaps you can tell me a bit about what biological question you are trying to answer?

Sergei

Title: Re: Evolutionary Fingerprint
Post by beyondblueskies on Jun 24th, 2011 at 3:56pm
Hi,

Thanks for the help. Just sent you an email with the explanation.

Thank you,
David

HyPhy message board » Powered by YaBB 2.5.2!
YaBB Forum Software © 2000-2024. All Rights Reserved.