[R-sig-eco] RDA

[Richards, Christina]

Dear Professor Oksanen and R-sig-eco community,

We corresponded several years ago about our problem of using 3 matrices in capscale because the capscale formula x~y + condition (z) does not allow a matrix of values for z. I was wandering if there have been any developments about having 3 or 4 matrices in the same formula, since now we would like also to combine an additional set of data (environmental)

We would like to identify the the effect of y1 and y2 on x with z partitioned out where:
x=phenotype
y1= dataframe of dna methylation with individuals listed in column 1 and 0/1 data across ~80 columns
y2= environmental data for each individual
z= dataframe of genetic loci with individuals listed in column 1 and 0/1 data across ~80 columns

Or really just identify how y1,y2, and z contribute to x.

Thanks for any ideas!
Christina

Christina Richards, Ph.D.
MOPGA-GRI scholar
University of T�bingen
Plant Evolutionary Ecology group
Auf der Morgenstelle 5
D-72076 T�bingen

Associate professor
University of South Florida
Department of Integrative Biology
4202 East Fowler Avenue SCA 127
NES 107 (shipping)
Tampa, FL 33620
(813)974-5090
(813)974-3263 FAX
http://www.ecologicalepigenetics.com
Twitter: @EcolEpig
Facebook: Ecological Epigenetics

Under Florida law, all email to and from this address is considered public record.
________________________________
From: Jari Oksanen <jari.oksanen using oulu.fi>
Sent: Thursday, November 26, 2015 7:24 PM
To: Richards, Christina <clr using usf.edu>
Cc: Marcelino de la Cruz <marcelino.delacruz using upm.es>; r-sig-ecology using r-project.org <r-sig-ecology using r-project.org>; Robertson, Marta <mhr using mail.usf.edu>; Verhoeven, Koen <K.Verhoeven using nioo.knaw.nl>; Alvarez, Mariano <malvare2 using mail.usf.edu>; Foust, Christy <christyfoust using mail.usf.edu>
Subject: Re: [R-sig-eco] RDA


> On 26 Nov 2015, at 17:27 pm, Richards, Christina <clr using usf.edu> wrote:
>
> Hello!
>
> That is very helpful and seems to work! Thank you!!
>
> I did not realize we could use raw data in capscale, is this true only because it is conditioning variables? At any rate it recapitulates our analysis using partial mantel in 2 species with 2 different results (one significant, the other not), so I'm inclined to believe its doing something similar.
>
If you supply raw community matrix (observations times species), capscale() will internally turn it into dissimilarities using specified distance= method (defaults �euclidean�). You should verify that the selected distance= method is the one you need. The variables on the right-hand-side of the formula must be in �raw� (observations times variables) format. They cannot be distances.

Cheers, Jari Oksanen

> ___
> From: Marcelino de la Cruz <marcelino.delacruz using upm.es>
> Sent: Wednesday, November 25, 2015 3:28 AM
> To: Richards, Christina; r-sig-ecology using r-project.org
> Subject: Re: [R-sig-eco] RDA
>
> Hi Christina,
>
> I think this could work:
>
> You should combine the *raw* y and z (i.e.not the Euclidean matrices) in
> the same data.frame (e.g. "yz"), and call capscale like this:
>
> capscale (x ~y1+y2+...+yn + Condition(z1+z2+...+z80), data=xy)
>
>
> Cheers!
>
> Marcelino
>
>
>
> --
> Marcelino de la Cruz Rot
> Depto. de Biolog�a Y Geolog�a
> Universidad Rey Juan Carlos
> M�stoles Espa�a
>
>
> El 24/11/2015 a las 22:11, Richards, Christina escribi�:
>>
>>
>> Hello!
>>
>>
>> We are trying to do a partial RDA with 3 matrices but our x matrix has a lot of missing data. We could use instead distance matrices which imputed missing data, but when we try to use capscale and the euclidean matrices, it seems we have to use the formula x~y + condition (z) and we cannot use a matrix of values for z.  We would like to identify the the effect of y on x with z partitioned out where:
>>
>>
>> x= dataframe of dna methylation with individuals listed in column 1 and 0/1 data across ~80 columns
>>
>> y= habitat for each individual
>>
>> z= dataframe of genetic loci with individuals listed in column 1 and 0/1 data across ~80 columns
>>
>>
>> Christina Richards, Ph.D.
>> University of South Florida
>> Department of Integrative Biology
>> 4202 East Fowler Avenue SCA 127
>> NES 107 (shipping)
>> Tampa, FL 33620
>> (813)974-5090
>> (813)974-3263 FAX
>> http://www.ecologicalepigenetics.com
>> Twitter: @EcolEpig
>> Facebook: Ecological Epigenetics
>>
>>      [[alternative HTML version deleted]]
>>
>> _______________________________________________
>> R-sig-ecology mailing list
>> R-sig-ecology using r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology
>>
>
> _______________________________________________
> R-sig-ecology mailing list
> R-sig-ecology using r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-ecology


	[[alternative HTML version deleted]]