0
$\begingroup$

Based on presence absence data, I'm attempting to find indicator species for different sample sites. I have 25 different groups or "clusters" as I've labelled them. Running the R code below, only results for 13 of the 25 groups was returned.

Is the output truncated, i.e. is there a max output parameter in R I need to fix? Or does this imply that there simply aren't any statistically relevant indicator species in the groups not included in the output?

Code used:

library(indicspecies)

data = read.csv("ISAI.txt", header= TRUE, sep= "\t") #Load up data from file

abund = data[,3:ncol(data)] #Collect only numerical data, starts in 3rd column
cluster = data$Cluster #Groups are defined in the Cluster column of the imported data

inv = multipatt(abund, cluster, func = "r.g", control = how(nperm=1000),duleg = TRUE) #Run analysis.

summary(inv) #print out results

Notes:

For the func parameter, "r" was used due to data being in presence/absence format. The "g" corrects for groups with different numbers of members. The duleg = True prevents group site combinations from being assessed as I'm not interested in those.

$\endgroup$

1 Answer 1

1
$\begingroup$

Got it! :)

I switched the summary(inv) to summary(inv, alpha=1) to get all possible results and all my groups showed up.

So, the answer to my own question is that the groups are missing from the output because they don't have any statistically significant OTUs correlated with those groups.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.