gmane.comp.lang.r.phylo

length of attribute (names) when calculatingindependent contrasts

Agus Camacho — 2012-05-24T18:58:05

Hi all,

When calculating independent contrasts of a variable, I get the following
message:

  el atributo 'names' [10] debe tener la misma longitud que el vector [9]
The attribute "names" [10] must have the same length than the vector[9]
(this was translated by me)

After checking the help on the command, i made the following

observations,
 > length(Weight)[1] 10> class(Weight)[1] "numeric"> tree$tip.label
[1] "C.leiolepis"      "C.nicterus"       "C.sinebrachiatus"
 [4] "S.catimbau"       "N.ablephara"      "P.erythrocercus"
 [7] "P.tetradactylus"  "V.rubricauda"     "M.maximiliani"
[10] "P.paeminosus"    > class(tree)[1] "phylo"

       0.9020000        0.7155556        0.5370526        0.5887500
     N.ablephara  P.erythrocercus  P.tetradactylus     V.rubricauda
       0.5460000        0.4127273        0.3970000        0.4725294
   M.maximiliani     P.paeminosus
       0.4833333        0.4677692



However, to me, everything seems to be right about the names and the
vector's length. Any hint?

Thanks to all. I am new to these analyses and appreciate much the help
from the mail list.


Agus

Re: problem with fitcontinuos function

Agus Camacho — 2012-05-23T20:32:14

Thanks to all for your fast answer, this list is amazing!

Graham's answer was perfectly satisfactory, once onnly that eliminated the
warning.

Thanks to all again.
Agus

2012/5/23 Graham Slater <gslater-xwz7R8GQi1g< at >public.gmane.org>

Re: problem with fitcontinuos function

Renata Brandt — 2012-05-23T20:27:24

Hello Agus,

I believe your problem will be solved if you set row names to the object
tp$temperature.

An example follows:
names(temperature)<-row.names(tp)

Please let me know if it worked.
Cheers, Renata

Re: problem with fitcontinuos function

Luke Harmon — 2012-05-23T20:23:32

Hi Agus - 

This might be a problem with node labels - try trimming them off of the tree and see if it works.

tree$node.label<-NULL


On May 23, 2012, at 1:16 PM, Agus Camacho wrote:


Luke Harmon
Assistant Professor
Biological Sciences
University of Idaho
208-885-0346
lukeh-Meo6Lv8EUjg3uPMLIKxrzw< at >public.gmane.org

problem with fitcontinuos function

Agus Camacho — 2012-05-23T20:16:54

Dear all,

fitcontinuous is giving me a warning generally related to non
correspondence between tree tips and data. However name.check and the tree
tips seems ok.

Anybody has a hint on this?

Thanks in advance.
Agus

Fitting  BM model:> tree
Phylogenetic tree with 11 tips and 10 internal nodes.

Tip labels:
C.leiolepis, C.nicterus, C.sinebrachiatus, S.catimbau, N.ablephara,
P.erythrocercus, ...
Node labels:
1, 2, 3, 4, 5, 6, ...

Rooted; includes branch lengths.

General question about the impact of sampling bias incomparative and diversification analyses

Julien Lorion — 2012-05-23T17:14:21

Dear all,

I have a 5-gene phylogenetic tree, dated, quite well resolved, including 50 species and covering 80~90% of the known diversity... I have also LASER diversification analyses and bayesian reconstruction of ancestral states for 5 characters , both taking into account the impact of sampling biases...

I evaluate now the adaptive significance of 5 characters with ape/caper (using Pagel's test, GEE and pGLS) AND their impact of diversification (using the Yule model with covariates). My results makes perfectly sense in terms of biology but I am worrying about their their relevance. Indeed, for 2 of my characters, I have only 50% of data... In another hand, the distribution of data is well balanced among lineages... and I would say that corresponding patterns of evolution inferred from Bayesian analyses are pretty clear...

I now the Beast team is working on an implementation of a Bayesian model for correlation analysis, that takes into account uncertainties (and thus missing data). But it is not ready yet.

So I have a couple of -likely naive- questions:
1. Is there a way to evaluate the impact of this kind of sampling bias on comparative analyses and diversification analyses ?

2. My guess is that the biases will affect much more directly the Yule model with covariates than GEE, pGLS and Pagel's models... as removing taxa for which I don't have data from the tree will induce an additional bias on diversification rates. Am I right ?

So far, I have removed all missing data in analyses including those 2 characters... My next question may sound crazy and/or desperate and I have a guess on the answer already :-), but...
3. How about assuming that the patterns inferred in my reconstruction of ancestral states are correct (free from unsampled reversal or acquisition that may drastically challenge conclusions) and "completing" real data with most probable states sampled during Bayesian approach along tip branches ? and then compare with results obtained without missing data ?

4 Should I definitely forget about the two characters for which I have only 50% data?

Thanks for any answer/comment/suggestion !

Regards

Julien

Julien Lorion
PhD, Post-doctoral fellow of the Japan Society for the Promotion of Science
Japan Agency for Marine-Earth Science and Technology (JAMSTEC)
Marine Ecosystems Research Department
2-15 Natsushima, Yokosuka 237-0061 Japan
Phone: +81-46-867-9570, Fax: +81-46-867-9525

Re: PIC or PGLS for genome-wide SNP screening

Anthony R Ives — 2012-05-23T15:45:04

Ted and Mattia,

For comparisons among models with the same parameters (such as OLS  
vs. PGLS), check out:

Vuong, Q. H. 1989. Likelihood ratio tests for model selection and non- 
nested hypotheses. Econometrica 57:307-333.
Clarke, K. A. 2007. A simple distribution-free test for nonnested  
model selection. Political Analysis 15:347-363.

I haven't tried these tests in this context, but they might be  
appropriate.

Cheers, Tony


On May 23, 2012, at 10:30 AM, Theodore Garland Jr wrote:


Anthony Ragnar Ives
Department of Zoology
UW-Madison
(608) 262-1519


[[alternative HTML version deleted]]

Re: PIC or PGLS for genome-wide SNP screening

Theodore Garland Jr — 2012-05-23T15:30:40

If you have the same set of independent variables, then you just prefer the one (OLS or PGLS) with the higher likelihood.  So far as I am told by Joe Felsenstein, you cannot do a ln maximum likelihood ratio test because the number of parameters is the same, although this paper seems to suggest otherwise:
Mooers, A. O., S. M. Vamosi, and D. Schluter. 1999. Using phylogenies to test macroevolutionary hypotheses of trait evolution in cranes (Gruinae). American Naturalist 154:249-259.

For two models with the same set of independent variables, AIC does not add anything for you.

If you go to something like Regression with an OU process modeled for the residuals, then you do have an additional parameter being estimated and so you can do an ln maximum likelihood ratio test of that model versus OLS and versus PGLS.  For example, see:
Lavin, S. R., W. H. Karasov, A. R. Ives, K. M. Middleton, and T. Garland, Jr. 2008. Morphometrics of the avian small intestine, compared with non-flying mammals: A phylogenetic perspective. Physiological and Biochemical Zoology 81:526-550. [provides Matlab Regressionv2.m, released as part of the PHYSIG package]
Gartner, G. E. A., J. W. Hicks, P. R. Manzani, D. V. Andrade, A. S. Abe, T. Wang, S. M. Secor, and T. Garland, Jr. 2010. Phylogeny, ecology, and heart position in snakes. Physiological and Biochemical Zoology 83:43-54.

Cheers,
Ted

Theodore Garland, Jr.
Professor
Department of Biology
University of California, Riverside
Riverside, CA 92521
Office Phone:  (951) 827-3524
Wet Lab Phone:  (951) 827-5724
Dry Lab Phone:  (951) 827-4026
Home Phone:  (951) 328-0820
Facsimile:  (951) 827-4286 = Dept. office (not confidential)
Email:  tgarland-3vSkeFsW7jA< at >public.gmane.org
http://www.biology.ucr.edu/people/faculty/Garland.html

Experimental Evolution: Concepts, Methods, and Applications of Selection Experiments. 2009.
Edited by Theodore Garland, Jr. and Michael R. Rose
http://www.ucpress.edu/book.php?isbn=9780520261808
(PDFs of chapters are available from me or from the individual authors)

________________________________________
From: r-sig-phylo-bounces-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org [r-sig-phylo-bounces-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org] on behalf of Mattia Prosperi [ahnven-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org]
Sent: Wednesday, May 23, 2012 8:05 AM
To: r-sig-phylo-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org
Subject: [R-sig-phylo] PIC or PGLS for genome-wide SNP screening

Dear all,

I am working on a data set composed of bacterial genomic sequences (a
few genes) associated to phenotypic values (in-vitro resistance to
antibiotics, a numerical value discretised into a binary class). Of
note, the bacterial isolates were sampled non-uniformly at different
times and locations, thus with a possible sampling bias. The data set
is ~1,000 variables and ~1,000 observations.

I have been applying several methods for developing a model to predict
antibiotic resistance from the single nucleotide polymorphisms (SNP)
extracted from a multiple alignment, applying classical statistical
learning and feature selection methods.
Eventually, I found that a logistic regression with main effects,
where the variables were selected first by a univariable chi-square
screening and then by AIC stepwise, was as good as other more complex
and non-linear methods (such as random forests) by comparing different
loss function (AUROC, specificity, sensitivity) distributions  upon
multiple cross-validation runs. Also, the SNP sets selected by
different approaches were highly similar and consistent across several
bootstrap evaluations.

I found that a few relevant (even after Bonferroni correction) SNP
were located in gene regions that are not supposed to be related with
antibiotic resistance. I thought that this might be a consequence of
neutral mutations that became fixed in the population by chance after
a genetic bottleneck (e.g. antibiotic pressure).
I'd like to understand if such SNP that is associated to antibiotic
resistance (and actually not expected to be) is indeed just a random
mutation of an early isolate that was carrying the true resistance SNP
(in another gene region) and that was selected by the antibiotic
pressure, thus transfering both the true resistance SNP and the
"hitchhicking" ones to the offspring. Unfortunately it is not easy to
cross-tabulate SNP in different genes because not all isolates have
been sequenced the same set of genes.

In order to check for fake/true SNP associated to resistance, I
thought I might use a PIC or PGLS approach (after estimating a
phylogenetic tree from the multiple alignment), in the same settings
as the original analysis, i.e. a model selection approach with both
feature and performance evaluation (well, since the coefficients of
PGLS/OLS are the same, it's just a matter of standard errors and
feature set selection), regressing the resistance class as a dependent
variable and using the SNP as covariates.

Is this a reasonable approach? Does it make sense to set up -for
instance- an AIC stepwise selection within a PGLS modeling?
I know that there is a way to check for phylogenetic signal and
therefore decide if the PGLS approach shall be employed. Is this the
way in which one decides for OLS vs. PGLS?

Last but not least, which is the most appropriate covariance matrix
calculation and PGLS implementation for this input-output set (i.e.
categorical variables, binary class)? The "brunch" function within
caper, or compar.gee within ape?

Thanks and apologies if some of the questions are silly.

M. Prosperi.

_______________________________________________
R-sig-phylo mailing list
R-sig-phylo-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo

PIC or PGLS for genome-wide SNP screening

Mattia Prosperi — 2012-05-23T15:05:31

Dear all,

I am working on a data set composed of bacterial genomic sequences (a
few genes) associated to phenotypic values (in-vitro resistance to
antibiotics, a numerical value discretised into a binary class). Of
note, the bacterial isolates were sampled non-uniformly at different
times and locations, thus with a possible sampling bias. The data set
is ~1,000 variables and ~1,000 observations.

I have been applying several methods for developing a model to predict
antibiotic resistance from the single nucleotide polymorphisms (SNP)
extracted from a multiple alignment, applying classical statistical
learning and feature selection methods.
Eventually, I found that a logistic regression with main effects,
where the variables were selected first by a univariable chi-square
screening and then by AIC stepwise, was as good as other more complex
and non-linear methods (such as random forests) by comparing different
loss function (AUROC, specificity, sensitivity) distributions  upon
multiple cross-validation runs. Also, the SNP sets selected by
different approaches were highly similar and consistent across several
bootstrap evaluations.

I found that a few relevant (even after Bonferroni correction) SNP
were located in gene regions that are not supposed to be related with
antibiotic resistance. I thought that this might be a consequence of
neutral mutations that became fixed in the population by chance after
a genetic bottleneck (e.g. antibiotic pressure).
I'd like to understand if such SNP that is associated to antibiotic
resistance (and actually not expected to be) is indeed just a random
mutation of an early isolate that was carrying the true resistance SNP
(in another gene region) and that was selected by the antibiotic
pressure, thus transfering both the true resistance SNP and the
"hitchhicking" ones to the offspring. Unfortunately it is not easy to
cross-tabulate SNP in different genes because not all isolates have
been sequenced the same set of genes.

In order to check for fake/true SNP associated to resistance, I
thought I might use a PIC or PGLS approach (after estimating a
phylogenetic tree from the multiple alignment), in the same settings
as the original analysis, i.e. a model selection approach with both
feature and performance evaluation (well, since the coefficients of
PGLS/OLS are the same, it's just a matter of standard errors and
feature set selection), regressing the resistance class as a dependent
variable and using the SNP as covariates.

Is this a reasonable approach? Does it make sense to set up -for
instance- an AIC stepwise selection within a PGLS modeling?
I know that there is a way to check for phylogenetic signal and
therefore decide if the PGLS approach shall be employed. Is this the
way in which one decides for OLS vs. PGLS?

Last but not least, which is the most appropriate covariance matrix
calculation and PGLS implementation for this input-output set (i.e.
categorical variables, binary class)? The "brunch" function within
caper, or compar.gee within ape?

Thanks and apologies if some of the questions are silly.

M. Prosperi.

Zero value of a significant model parameter duringPGLS

Agus Camacho — 2012-05-23T01:54:25

Dear all,

I just had the following result fitting a Pgls with both, brownian and O-U
correlation structures.
I was wandering what could have made the value and Sd go to zero. Altitude
values were considered continuous but were strongly grouped in two levels.
Might that have to do with the result?

Would appreciate any hint.
There goes the model's summary.
Thanks in advance.


Generalized least squares fit by maximum likelihood
  Model: temperature ~ 1 + altitude
  Data: tp
  AIC BIC logLik
   60  62    -26

Correlation Structure: corMartins
 Formula: ~1
 Parameter estimate(s):
alpha
   21

Coefficients:
            Value Std.Error t-value p-value
(Intercept)    35       2.8    12.7   0.000
altitude        0       0.0    -2.2   0.057

 Correlation:
         (Intr)
altitude -0.93

Standardized residuals:
   Min     Q1    Med     Q3    Max
-1.907 -0.519  0.034  0.734  1.859

Residual standard error: 2.7
Degrees of freedom: 11 total; 9 residual

Re: center of tree

Klaus Schliep — 2012-05-22T10:28:46

Hello Mathias,

I'm not aware of any general  COT in R, your chance to make a contribution.
Midpoint rooting is the COT for a special function F [2], it minimizes
the maximum
distance from the root (COT) to the tips.

Cheers,
Klaus



On 5/18/12, Walter, Mathias <mathias-taBouHiV1h1goHlPtYpdqQ< at >public.gmane.org> wrote:

compute HKY85 distances in R

Juan Antonio Balbuena — 2012-05-22T09:49:55

Hello
I would like to know whether it is possible to compute HKY85 genetic
distances using R. The model is not included in the dist.dna function
of ape and I was wondering whether it is available in another package.
Thank you very much for your attention.

Juan Antonio Balbuena

Re: unrooted tree, spread tips

Emmanuel Paradis — 2012-05-18T11:34:39

Walter, Mathias wrote on 18/05/2012 18:16:

This should not happen. May be a bug. You have an example of this?

E.

Re: unrooted tree, spread tips

Walter, Mathias — 2012-05-18T11:16:34

Hi,

if I use type="r", then quite a lot of lines cross. I tried optimal
tree ordering but as soon as I convert my hclust object to a phylo
object (via as.phylo), the ordering is lost.

Another advantage of the unrooted tree is, that you can often clearly
distinguish main clusters. That is not possible with any other kind of
tree drawing.

--
Kind regards,
Mathias

2012/5/18 Emmanuel Paradis <Emmanuel.Paradis-ysPYPCwjBSM< at >public.gmane.org>:

Re: unrooted tree, spread tips

Emmanuel Paradis — 2012-05-18T10:53:13

Hi,

Walter, Mathias wrote on 18/05/2012 16:42:

No. The only available algorithm for unrooted trees in ape is the 
"equal-angle" one. It seems that for your purpose type="r" might be not 
bad, even though the center of the circle is the root.

Best,

Emmanuel

unrooted tree, spread tips

Walter, Mathias — 2012-05-18T09:42:43

Hi,

I use ape to plot my phylogenetic trees. Often I have to plot unrooted
trees with few clades, but a lot of tips per clade. The tip labels are
hard to read, even if I use lab4ut="axial". Is there any way to spread
the tips? I saw this in some other tree drawing programs.

--
Kind regards,
Mathias

Re: Best way to test correlation between discrete andcontinuous variables ?

Patrik Lindenfors — 2012-05-17T10:04:19

Julien,
Regarding my paper, it was outdated already when published. Stick to the other suggestions given.

Best regards
                              - Patrik Lindenfors -

On 2012-05-17 11:51, Julien Lorion wrote:
 ne: (951) 827-4026 Home Phone: (951) 328-0820 Facsimile: (951) 827-4286 = Dept. office (not confidential) Email: tgarland-3vSkeFsW7jA< at >public.gmane.org http://www.biology.ucr.edu/people/faculty/Garland.html Expe!
 rimental Evolution: Concepts, Methods, and Applications of Selection Experiments. 2009. Edited by Theodore Garland, Jr. and Michael R. Rose http://www.ucpress.edu/book.php?isbn=9780520261808 (PDFs of chapters are available from me or from the individual authors) ________________________________________ From: r-sig-phylo-bounces-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org [r-sig-phylo-bounces-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org] on behalf of Marguerite Butler [mbutler808-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org] Sent: Wednesday, May 16, 2012 12:27 PM To: Julien Lorion Cc: r-sig-phylo-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org Subject: Re: [R-sig-phylo] Best way to test correlation between discrete and continuous variables ? Dear Julien, There is no problem with applying an ANOVA within a phylogenetic framework. This is essentially phylogenetic GLS, which you can implement easily with APE. You can ha
 ve a look at Emmanuel's book (which was just recently came out in the second edition Nov. 2011, by the way). http://www.amazon.com/Analysis-Phylogenetics-Evolution-Emmanuel-Paradis/dp/14614!
 17422/ref=sr_ob_11?s=books&ie=UTF8&qid=1337196146&sr=1-11 In essence, 
you are looking at the phylogeny as a source of "correlated errors" which you are "correcting for" under some assumed model of evolution -- either Brownian motion or Ornstein Ulenbeck. It is viewed as noise which is controlled for in order to see the pattern from ecology, etc. The mechanics of how to incorporate the phylogenetic covariance matrix into the linear model is explained in the appendix of my paper: Butler M.A. Schoener T.W., and Losos J.B. (2000) The relationship between habitat type and sexual size dimorphism in Greater Antillean Anolis lizards. Evolution 54(1):259-272. DOI: http://dx.doi.org/10.1554/0014-3820(2000)054[0259:TRBSSD]2.0.CO;2 Another approach to analyzing the same kind of data is to view the evolution of the quantitative character as being influenced by a number o
 f factors (for example, habitat, symbionts, etc.), which can be thought of as "selective regimes" which influence the evolution of body size. You can then create explicit biological hypothe!
 ses which are translated to mathematical models, and test these hypotheses against each other for the best explanation of the data. This approach has software package developed for it called "OUCH" which is available in R. It is explained and illustrated in this paper: Butler M.A. and King A.A. (2004) Phylogenetic comparative analysis: a modeling approach for adaptive evolution. The American Naturalist 164(6):683-695. DOI: 10.1086/426002 Appologies for the shameless self-promotion:). Marguerite On May 15, 2012, at 9:53 PM, Julien Lorion wrote:
 wo binary variables (habitat and symbiont location) on the body length... In fact, that looks likes ANOVAs from which I wanna remove the phylogenetic bias... The point is that I don't know !
 how to do that in practice. May someone have some advices ? Thanks by advance Best regards Julien _______________________________________________ R-sig-phylo mailing list R-sig-phylo-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org https://stat.ethz.ch/mailman/listinfo/r-sig-phylo

[[alternative HTML version deleted]]

Re: Best way to test correlation between discrete andcontinuous variables ?

Julien Lorion — 2012-05-17T03:51:53

Alexandro, Marguerite and Ted...

Thanks a bunch for your valuable replies... 

Quickly after I sent my sos on the mailing list, I found Ted's paper and implementation of his method in Geiger... I also found the the Brunch algorithm (package Caper), a variation of the Felsenstein's PIC...

I'll add the GLS as Marguerite suggested...and it should be fine... 

I also found a paper by Lindenfors Patrick, introducing another type of phylogenetic ANOVA... But it looks like it has not been implemented yet in any software
http://zootis.zoologi.su.se/research/lindenfors/publications/Lindenfors_2006.pdf

Glad I have found this mailing list !! 

Regards

Julien






On May 17, 2012, at 5:58 AM, Theodore Garland Jr wrote:


Julien Lorion
PhD, Post-doctoral fellow of the Japan Society for the Promotion of Science
Japan Agency for Marine-Earth Science and Technology (JAMSTEC)
Marine Ecosystems Research Department
2-15 Natsushima, Yokosuka 237-0061 Japan
Phone: +81-46-867-9570, Fax: +81-46-867-9525

Re: Best way to test correlation between discreteandcontinuous variables ?

Theodore Garland Jr — 2012-05-16T20:58:25

Dear Julien,

I'll just add two other papers to your reading list (both available at my web page):

Garland, T., Jr., A. W. Dickerman, C. M. Janis, and J. A. Jones. 1993. Phylogenetic analysis of covariance by computer simulation. Systematic Biology 42:265-292.

And the Appendix of this paper:
Lavin, S. R., W. H. Karasov, A. R. Ives, K. M. Middleton, and T. Garland, Jr. 2008. Morphometrics of the avian small intestine, compared with non-flying mammals: A phylogenetic perspective. Physiological and Biochemical Zoology 81:526-550. [provides Matlab Regressionv2.m, released as part of the PHYSIG package]

Cheers,
Ted

Theodore Garland, Jr.
Professor
Department of Biology
University of California, Riverside
Riverside, CA 92521
Office Phone:  (951) 827-3524
Wet Lab Phone:  (951) 827-5724
Dry Lab Phone:  (951) 827-4026
Home Phone:  (951) 328-0820
Facsimile:  (951) 827-4286 = Dept. office (not confidential)
Email:  tgarland-3vSkeFsW7jA< at >public.gmane.org
http://www.biology.ucr.edu/people/faculty/Garland.html

Experimental Evolution: Concepts, Methods, and Applications of Selection Experiments. 2009.
Edited by Theodore Garland, Jr. and Michael R. Rose
http://www.ucpress.edu/book.php?isbn=9780520261808
(PDFs of chapters are available from me or from the individual authors)

________________________________________
From: r-sig-phylo-bounces-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org [r-sig-phylo-bounces-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org] on behalf of Marguerite Butler [mbutler808-Re5JQEeQqe8AvxtiuMwx3w< at >public.gmane.org]
Sent: Wednesday, May 16, 2012 12:27 PM
To: Julien Lorion
Cc: r-sig-phylo-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org
Subject: Re: [R-sig-phylo] Best way to test correlation between discrete and    continuous variables ?

Dear Julien,

There is no problem with applying an ANOVA within a phylogenetic framework. This is essentially phylogenetic GLS, which you can implement easily with APE. You can have a look at Emmanuel's book (which was just recently came out in the second edition Nov. 2011, by the way).

http://www.amazon.com/Analysis-Phylogenetics-Evolution-Emmanuel-Paradis/dp/1461417422/ref=sr_ob_11?s=books&ie=UTF8&qid=1337196146&sr=1-11

In essence, you are looking at the phylogeny as a source of "correlated errors" which you are "correcting for" under some assumed model of evolution -- either Brownian motion or Ornstein Ulenbeck. It is viewed as noise which is controlled for in order to see the pattern from ecology, etc. The mechanics of how to incorporate the phylogenetic covariance matrix into the linear model is explained in the appendix of my paper:

Butler M.A. Schoener T.W., and Losos J.B. (2000)  The relationship between habitat type and sexual size dimorphism in Greater Antillean Anolis lizards.  Evolution 54(1):259-272. DOI: http://dx.doi.org/10.1554/0014-3820(2000)054[0259:TRBSSD]2.0.CO;2

Another approach to analyzing the same kind of data is to view the evolution of the quantitative character as being influenced by a number of factors (for example, habitat, symbionts, etc.), which can be thought of as "selective regimes" which influence the evolution of body size. You can then create explicit biological hypotheses which are translated to mathematical models, and test these hypotheses against each other for the best explanation of the data. This approach has software package developed for it called "OUCH" which is available in R.  It is explained and illustrated in this paper:

Butler M.A. and King A.A. (2004) Phylogenetic comparative analysis: a modeling approach for adaptive evolution. The American Naturalist 164(6):683-695. DOI: 10.1086/426002

Appologies for the shameless self-promotion:).

Marguerite

On May 15, 2012, at 9:53 PM, Julien Lorion wrote:


____________________________________________
Marguerite A. Butler
Associate Professor

Department of Biology
University of Hawaii
2450 Campus Rd., Dean Hall Rm. 2
Honolulu, HI 96822

Office: 808-956-4713
Dept: 808-956-8617
Lab:  808-956-5867
FAX:   808-956-9812
http://www.hawaii.edu/zoology/faculty/butler.html
http://www2.hawaii.edu/~mbutler
http://www.hawaii.edu/zoology/








        [[alternative HTML version deleted]]

_______________________________________________
R-sig-phylo mailing list
R-sig-phylo-0bNBQ1PAWB4BXFe83j6qeQ< at >public.gmane.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo

Re: Best way to test correlation between discrete andcontinuous variables ?

Marguerite Butler — 2012-05-16T19:27:30

Dear Julien,

There is no problem with applying an ANOVA within a phylogenetic framework. This is essentially phylogenetic GLS, which you can implement easily with APE. You can have a look at Emmanuel's book (which was just recently came out in the second edition Nov. 2011, by the way).

http://www.amazon.com/Analysis-Phylogenetics-Evolution-Emmanuel-Paradis/dp/1461417422/ref=sr_ob_11?s=books&ie=UTF8&qid=1337196146&sr=1-11

In essence, you are looking at the phylogeny as a source of "correlated errors" which you are "correcting for" under some assumed model of evolution -- either Brownian motion or Ornstein Ulenbeck. It is viewed as noise which is controlled for in order to see the pattern from ecology, etc. The mechanics of how to incorporate the phylogenetic covariance matrix into the linear model is explained in the appendix of my paper:

Butler M.A. Schoener T.W., and Losos J.B. (2000) The relationship between habitat type and sexual size dimorphism in Greater Antillean Anolis lizards. Evolution 54(1):259-272. DOI: http://dx.doi.org/10.1554/0014-3820(2000)054[0259:TRBSSD]2.0.CO;2

Another approach to analyzing the same kind of data is to view the evolution of the quantitative character as being influenced by a number of factors (for example, habitat, symbionts, etc.), which can be thought of as "selective regimes" which influence the evolution of body size. You can then create explicit biological hypotheses which are translated to mathematical models, and test these hypotheses against each other for the best explanation of the data. This approach has software package developed for it called "OUCH" which is available in R. It is explained and illustrated in this paper:

Butler M.A. and King A.A. (2004) Phylogenetic comparative analysis: a modeling approach for adaptive evolution. The American Naturalist 164(6):683-695. DOI: 10.1086/426002

Appologies for the shameless self-promotion:).

Marguerite

On May 15, 2012, at 9:53 PM, Julien Lorion wrote:

____________________________________________
Marguerite A. Butler
Associate Professor

Department of Biology
University of Hawaii
2450 Campus Rd., Dean Hall Rm. 2
Honolulu, HI 96822

Office: 808-956-4713
Dept: 808-956-8617
Lab: 808-956-5867
FAX: 808-956-9812
http://www.hawaii.edu/zoology/faculty/butler.html
http://www2.hawaii.edu/~mbutler
http://www.hawaii.edu/zoology/

[[alternative HTML version deleted]]

Mentorship program the 2012 iEvoBio conference

Hilmar Lapp — 2012-05-16T18:18:05

The iEvoBio Mentorship Program pairs undergraduate and graduate students with faculty who have extensive experience in evolution, systematics, biodiversity, mathematics and/or software development to enhance their experience at iEvoBio 2012. Mentors will guide participants in collaborative and networking opportunities throughout the two-day conference.

If you would like to gain the wisdom of your colleagues' experiences by participating in this program, please email the iEvoBio 2012 organizing committee at committee-Phy/wAXTOfxAfugRpC6u6w< at >public.gmane.org by June 1, indicating if you have any activities you would like to be included in the mentorship program, or a particular field of interest. You can find more information about the meeting and registration at http://ievobio.org/.

iEvoBio 2012 is sponsored by the US National Evolutionary Synthesis Center (NESCent) and by Biomatters Ltd., in partnership with the Society for the Study of Evolution (SSE) and the Systematic Biologists (SSB).

The iEvoBio 2012 Organizing Committee:
Hilmar Lapp, US National Evolutionary Synthesis Center (chair)
Robert Beiko, Dalhousie University
Nico Cellinese, University of Florida
Robert Guralnick, University of Colorado at Boulder
Rebecca Kao, Denver Botanic Gardens
Ellinor Michel, Natural History Museum, London
Nadia Talent, Royal Ontario Museum
Andrea Thomer, University of Illinois at Urbana-Champaign