1

I'm trying to get the coefficients of the taxa that generate the most differences between the groups in my permutest of types of supplies for the animals. I paste the results of the permutest:

permutest(betadispersionAlimentación, pairwise = TRUE) enter image description here

When I try to get the coefficients, Studio answer me with this:

coefficients(betadispersionAlimentación)["C",] NULL

The object to get the coefficients is the type of: Permanova? Betadisper? Permutest?

i don't know how can I get the taxa (genus, Family, Specie...) that contribute to the differences.

can anybody help me, please?

Thanks a lot!

Srvcl
  • 11
  • 2

1 Answers1

0

For anybody who have the same problem. From the day that I posted the question I have worked about that. To get de genus or the specie that more contribute to the dissimilarity I used the vegan library.

Here I used the Vegan Distance to create de distance object

DIST_ABUNDANCES <- vegdist(t(OTU_ABUNDANCES))

Here I took the anova function to create all the values.

anova(betadisper(DIST_ABUNDANCES, META_ABUNDANDES$Sexo))

Here I made the permutest to compare all the values of the variable Sexo

permutest(betadisper(DIST_ABUNDANCES, META_ABUNDANDES$Sexo), pairwise = TRUE)

Here I took the correlation coefficients of the anova to seek the objetive

COEF_ABUNDANCES <- coefficients(PERMANOVA_ABUNDANCES)["Sexo1",]

I created the top coeficients object

TOP.COEF_ABUNDANCES <- COEF_ABUNDANCES[rev(order(abs(COEF_ABUNDANCES)))[1:20]]

Here I designed the plot of the coefficients

par(mar = c(3, 14, 2, 1))
barplot(sort(TOP.COEF_ABUNDANCES), horiz = T, las = 1, main = "Top taxa")

I followed this tutorial: https://mibwurrepo.github.io/Microbial-bioinformatics-introductory-course-Material-2018/multivariate-comparisons-of-microbial-community-composition.html

Srvcl
  • 11
  • 2
  • Also, to establish the parameters of the plot you can access to this information. In my case, I had to set a bigger size to the left side parameter of the plot. http://rfunction.com/archives/1302 – Srvcl Jan 07 '23 at 23:15