Hi Jake, Yes, that's right. Glad that cleared it up!
Angie ----- "Jacob Biesinger" <[email protected]> wrote: > From: "Jacob Biesinger" <[email protected]> > To: "Angie Hinrichs" <[email protected]> > Cc: [email protected] > Sent: Tuesday, December 15, 2009 10:41:02 AM GMT -08:00 US/Canada Pacific > Subject: Re: [Genome] Drosophila 15 way phylogeny? > > Okay, so if I understand the pipeline, phyloFit is used to make an initial > estimate from a cladogram (no branch lengths). The resulting tree is fed into > phastCons as an initial estimate. phastCons gives estimates for conserved and > non-conserved branch lengths. phastCons is called repeatedly, using the > previous estimate for non-conserved elements as the prior estimate. > > I was under the impression that the tree input into phastCons at the first > iteration was obtained from an external source (i.e., using 4-fold degenerate > sites). Thanks for helping me clear that up. -- > Jake Biesinger > Graduate Student > Xie Lab, UC Irvine > (949) 231-7587 > > > > > On Mon, Dec 14, 2009 at 11:29 AM, Angie Hinrichs < [email protected] > > wrote: > Hi Jake, > > Our models for conserved and nonconserved sites' substitution rates are > estimated from bases that we were able to align across species, so right > there is a big source of bias towards bases that have not diverged beyond > recognition. The "four-fold degenerate positions" subst/site must have been > estimated by a different method from the trees that we feed into phastCons. > > BTW If you would like to see the "makedoc", the file is > kent/src/hg/makeDb/doc/dm2.txt in our source tree ( > http://hgdownload.cse.ucsc.edu/admin/jksrc.zip or see > http://genome.ucsc.edu/admin/cvs.html if you use CVS), and has all the gory > detail of how we generated the track. Search for the "PHASTCONS 15WAY" > section. > > I hope that helps, and please let us know at [email protected] if you have > more questions. That said, Adam Siepel (at Cornell now) created the > estimation methodology behind our conservation scores, so if you really want > to get into the finer points of the statistical basis for the estimates, he's > the best source. > > Angie > > > ----- "Jacob Biesinger" < [email protected] > wrote: > > > From: "Jacob Biesinger" < [email protected] > > > To: [email protected] > > Sent: Friday, December 11, 2009 5:39:52 PM GMT -08:00 US/Canada Pacific > > Subject: [Genome] Drosophila 15 way phylogeny? > > > > > Hi! > > > > I've pulled up an old thread talking about the Drosophila 15-way multiz > > phylogeny at > > https://lists.soe.ucsc.edu/pipermail/genome/2007-January/012530.html > > > > I see the conserved and non-conserved posterior estimates, and the prior > > estimate from the "makedoc:" > > > > >* This is from our makedoc for the 15way Conservation: > > *>* > > *>* #TREE: > > *>* > > (((((((((dm2:0.032478,(droSim1:0.017650,droSec1:0.015740):0.017736):0.026088,(droYak2:0.058116,droEre2:0.055952):0.031922):0.084953,droAna3:0.218988):0.051563,(dp4:0.013624,droPer1:0.015374):0.210705):0.046101,droWil1:0.292357):0.019351,((droVir3:0.109131,droMoj3:0.142623):0.047595,droGri2:0.154583):0.189999):0.113599,anoGam1:0.357817):0.092848,apiMel2:0.386309):0.169225,triCas2:0.169225); > > > > *>* > > * > > > > So this is the tree used as a prior. Although the overall topology is > > similar to what I've seen before (i.e., > > http://web.archive.org/web/20080413030437/http://rana.lbl.gov/~dan/trees.html > > ) > > I'm a bit confused because the total branch length for the tree is only ~ > > 3.14. According to the original 12 drosophila sequencing paper, the branch > > length estimated for four-fold degenerate positions was ~4.5 for the 12 > > drosophilids alone. In this prior tree, the branch length for the > > drosophilids are only ~1.85. Where exactly is this phylogeny coming > > from? > > > > > > Thanks! > > -- > > Jake Biesinger > > Graduate Student > > UC Irvine > > (949) 231-7587 > > _______________________________________________ > > Genome maillist - [email protected] > > https://lists.soe.ucsc.edu/mailman/listinfo/genome > > _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
