Hi Luke, Thank you so much for your reply! I never thought about the possibility to model the trend as a fixed effect. After you mentioned it, I think it makes sense to me.
Indeed, after inspecting my data merely by eyes, I think I do find a trend that species with longer branches to have longer DNA deletions (the trait that I mentioned that always increases). It's just amazing how I didn't find this pattern in the dataset that I've stared at for months. Anyway, thank you for your suggestion, and I'll incorporate branch length into the model. Best, Yanzhu > -----原始邮件----- > 发件人: "Luke Matthews" <lmatt...@rand.org> > 发送时间: 2022-10-18 22:18:22 (星期二) > 收件人: "r-sig-phylo@r-project.org" <r-sig-phylo@r-project.org> > 抄送: > 主题: [R-sig-phylo] reply to alternative error structures in PGLS > > Hi Yanzhu, > > You raise and excellent question. I think the Brownian model would not be statistically correct if implemented naively to a dataset that shows a strong trend over time (increase but never decrease). There are a couple ways the overall model, however, might be adjusted to accommodate this. If change happens at speciation points, then entering a fixed effect variable for the number of nodes root-to-tip would, in theory, account for the trend as a fixed effect. Then the phylogenetic error term is OK that it is modeling Brownian motion, because it is modeling the residual after the trend was already netted out by the fixed effect. > > Depending on your biological system, the molecular branch length, or the time branch length (with fossils) might be appropriate fixed effects to model a trend like this. > > There are also a variety of methods out there to model variation in the rates of a trait change down various branches of a tree. Those might be relevant to your question, but others on this list serve know much more about them than I do. > > My best > Luke > > -----Original Message----- > From: R-sig-phylo <r-sig-phylo-boun...@r-project.org> On Behalf Of r-sig-phylo-requ...@r-project.org > Sent: Tuesday, October 18, 2022 6:00 AM > To: r-sig-phylo@r-project.org > Subject: [EXT] R-sig-phylo Digest, Vol 177, Issue 2 > > Send R-sig-phylo mailing list submissions to > r-sig-phylo@r-project.org > > To subscribe or unsubscribe via the World Wide Web, visit > https://stat.ethz.ch/mailman/listinfo/r-sig-phylo > or, via email, send a message with subject or body 'help' to > r-sig-phylo-requ...@r-project.org > > You can reach the person managing the list at > r-sig-phylo-ow...@r-project.org > > When replying, please edit your Subject line so it is more specific than "Re: Contents of R-sig-phylo digest..." > > > Today's Topics: > > 1. Alternative error structures in PGLS (jiyan...@ioz.ac.cn) > > ---------------------------------------------------------------------- > > Message: 1 > Date: Tue, 18 Oct 2022 09:49:19 +0800 (GMT+08:00) > From: jiyan...@ioz.ac.cn > To: r-sig-phylo@r-project.org > Subject: [R-sig-phylo] Alternative error structures in PGLS > Message-ID: <3bdf3a4c.4b8c.183e8c7eb1d.coremail.jiyan...@ioz.ac.cn> > Content-Type: text/plain; charset="utf-8" > > Dear all, > > > > I have a question related to the residual structure of PGLS models. If I understand correctly, PGLS models implemented in R assume a Brownian motion (or lambda model) of the correlation structure of the residuals when using corBrownian or corPagel functions. However, I am just wondering if anyone has any idea if these models could still be applied for PGLS tests where the dependent variable does not follow Brownian motion but instead, for example, can only increase instead of decrease through time? If the residuals of OLS exhibit phylogenetic signals, would it be sufficient to justify the use of Brownian motion to correct for phylogeny? > > > > Many thanks! > > > > Yanzhu > > > > > Yanzhu Ji > > Postdoctoral researcher > > Key Laboratory of Zoological Systematics and Evolution | Institute of Zoology | Chinese Academy of Sciences > [[alternative HTML version deleted]] > > > > ------------------------------ > > Subject: Digest Footer > > _______________________________________________ > R-sig-phylo mailing list > R-sig-phylo@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-phylo > > > ------------------------------ > > End of R-sig-phylo Digest, Vol 177, Issue 2 > ******************************************* > __________________________________________________________________________ > > This email message is for the sole use of the intended r...{{dropped:6}} > > _______________________________________________ > R-sig-phylo mailing list - R-sig-phylo@r-project.org > https://stat.ethz.ch/mailman/listinfo/r-sig-phylo > Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/ </r-sig-phylo-boun...@r-project.org></r-sig-phylo@r-project.org></lmatt...@rand.org> _______________________________________________ R-sig-phylo mailing list - R-sig-phylo@r-project.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo Searchable archive at http://www.mail-archive.com/r-sig-phylo@r-project.org/
