HyPhy message board
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl
HYPHY Package >> HyPhy feedback >> IFEL LTR
http://www.hyphy.org/cgi-bin/hyphy_forums/YaBB.pl?num=1150888730

Message started by YER on Jun 21st, 2006 at 4:18am

Title: IFEL LTR
Post by YER on Jun 21st, 2006 at 4:18am
Dear Sergei,

I am trying to dive into the population-based analyses of dN/dS ratios based on your paper in PLoS Computational Biology. One thing that is not clear to me is how to implement the test to determine the significance of IFEL test. I have scrolled through the various options that might comprise this test, but was unable to find it. The subtreeselectioncomparison.bf can be used to carry out the IFEL test for individual sites in a sample, but the LRT to assess the significance of any temporal heterogeneity of dN/dS across sites in a population sample does not seem to be in the standard options of HYPHY. Is this correct? If so, could you point out how to set up this LRT and especially how to assign tips, interiors, and the two site classes?

Thanks again, Yer

Title: Re: IFEL LTR
Post by Sergei on Jun 21st, 2006 at 8:13am
Dear Yer,

There are several tests described in the PLoS paper (Multimedia File Viewing and Clickable Links are available for Registered Members only!!  You need to Login Login).

1). Test for selection along internal branches at every site of a given sample. This can be achieved by Standard Analyses->Positive Selection->QuickSelectionDetection->...->Two Rate FEL -> ... -> Internal Branches Only (when prompted for branch class).

2). Differential IFEL - compare strengths of selection along internal branches at each site of two samples. Standard Analyses->PositiveSelection->CompareSelectivePressureIVL.bf

3). Given a sample, and a partitioning of codons into two non-overlapping (but not necessarily covering the entire sequence) sets, test whether the mean strength of selection along internal branches differs between the two sets (Population Level Adaptation Test). This is not included in standard analyses, but you can do it like this.

Run an IFEL analysis from (1). This will generate a nucleotide fit file among other things. When you look at the results of IFEL, partition all codons into those which (i) Have only tip dN > 0 (ii) Have internal branch dN>0. You need to generate a file that looks like


Code (]
partSpec={};partSpec["Tips"):

="18-20,36-38,48-50,51-53,63-65,69-71,84-86,126-128,180-182,183-185,186-188,189-191,213-215,2
16-218,225-227,240-242,243-245,291-293,294-296,297-299,303-305,312-314,318-320,372-374,375-377,384-386,390-392,396-398,40
5-407,408-410,411-413,414-416,423-425,435-437,438-440,450-452,453-455,459-461,462-464,468-470,471-473,489-491,504-506,549
-551,561-563,573-575,576-578,579-581,582-584,585-587,588-590,591-593,594-596,597-599";
partSpec["Internals"]="6-8,21-23,24-26,57-59,123-125,171-173,246-248,282-284,351-353,363-365,399-401,465-467,477-479,483-
485,498-500,510-512";

I'll write a little script which takes the output of IFEL and generates it for you and post it later today (hopefully).

With the nucleotide fit and the partition file, now download Multimedia File Viewing and Clickable Links are available for Registered Members only!!  You need to Login Login and feed it the two files for the IFEL LRT test.

Hopefully this helps,
Sergei

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