NMDS (Non-metric multidimensional scaling)

So this works for my data!

I had started with the commands under the section NMDS from
http://userweb.eng.gla.ac.uk/umer.ijaz/bioinformatics/ecological.html

the sample and data samples were prepared according to the above.

and proceed with the following:

library("vegan", lib.loc="~/R/win-library/3.2")
library("lattice", lib.loc="C:/Program Files/R/R-3.2.5/library")
set.seed(2)
abund_table<-read.csv("Sample.csv",row.names=1,check.names=FALSE)
#Transpose the data to have sample names on rows
abund_table<-t(abund_table)

abund_table<-t(abund_table)

example_NMDS=metaMDS(abund_table,
k=2)
plot(example_NMDS)

ordiplot(example_NMDS,type="n")
orditorp(example_NMDS,display="species",col="red",air=0.01)
orditorp(example_NMDS,display="sites",cex=1.25,air=0.01)

And I changed "Treatment 1" and "Treatment 2" to:
treat=c(rep("pH",5),rep("Conductivity",5))
ordiplot(example_NMDS,type="n")
ordihull(example_NMDS,groups=treat,draw="polygon",col="grey90",label=F)
orditorp(example_NMDS,display="species",col="red",air=0.01)
orditorp(example_NMDS,display="sites",col=c(rep("green",5),rep("blue",5)),
air=0.01,cex=1.25)

colors=c(rep("red",5),rep("blue",5))
ordiplot(example_NMDS,type="n")

#Plot convex hulls with colors baesd on treatment
for(i in unique(treat)) {
ordihull(example_NMDS$point[grep(i,treat),],draw="polygon",
groups=treat[treat==i],col=colors[grep(i,treat)],label=F) }
orditorp(example_NMDS,display="species",col="red",air=0.01)
orditorp(example_NMDS,display="sites",col=c(rep("green",5),
rep("blue",5)),air=0.01,cex=1.25)



Thanks to this website!
https://jonlefcheck.net/2012/10/24/nmds-tutorial-in-r/

I have been following the above as Jon is an expert in R for ecological data.

Popular posts from this blog

Correlation

Image
Correlation.R # ============================================================ # Tutorial on drawing a correlation map using ggplot2 # by Umer Zeeshan Ijaz (http://userweb.eng.gla.ac.uk/umer.ijaz) # =============================================================   abund_table<- read.csv ( "SPE_pitlatrine.csv" , row.names = 1 , check.names= FALSE ) #Transpose the data to have sample names on rows abund_table<- t ( abund_table ) meta_table<- read.csv ( "ENV_pitlatrine.csv" , row.names = 1 , check.names= FALSE )   #Filter out samples with fewer counts abund_table<-abund_table [ rowSums ( abund_table ) > 200 , ]   #Extract the corresponding meta_table for the samples in abund_table meta_table<-meta_table [ rownames ( abund_table ) , ]   I changed the parameters #You can use sel_env to specify the variables you want to use and sel_env_label to specify the labes for the pannel sel_env<- c ( "pH" , "Temp" , "TS
Read more

Taxa Plot

Image
So, we have tried. From SILVA NGS, RDP Pipeline, MG-RAST to MEGA, and MEGAN...6.  Two made the cut with my data so far; MEGAN and R.  As a note, I would say the data choose the analysis software. R yet again gave a beautiful taxa plot.  abund_table<- read.csv ( "SPE_pitlatrine.csv" , row.names = 1 , check.names= FALSE )   #Transpose the data to have sample names on rows abund_table<- t ( abund_table ) meta_table<- read.csv ( "ENV_pitlatrine.csv" , row.names = 1 , check.names= FALSE )   #Just a check to ensure that the samples in meta_table are in the same order as in abund_table meta_table<-meta_table [ rownames ( abund_table ) , ]   #Get grouping information grouping_info<- data.frame ( row.names = rownames ( abund_table ) , t ( as.data.frame ( strsplit ( rownames ( abund_table ) , "_" ) ) ) ) # > head(grouping_info) # X1 X2 X3 # T_2_1 T 2 1 # T_2_10 T 2 10 # T_2_12 T 2 12 # T_2_2 T 2 2 # T_2_3
Read more