I am receiving an error message with metagenomeSeq that I simply can't figure out:

Error in validObject(.Object) : 

  invalid class “MRexperiment” object: featureNames differ between assayData and featureData

I believe this refers to the features not matching between the count matrix and the taxonomy.  However, I have triple-checked the files, and they do have the same features names, in the same order.  I did some googling, and perhaps came up with a clue, but I am not a savvy R user at all.  The user forum I found indicated that this error message has something to do with using make.names for assayData, but not feature data:

https://stat.ethz.ch/pipermail/bioconductor/2013-November/056100.html

 Here is the code I am using (adapted from the user manual), and I have attached all relevant files:

counts = load_meta("counts.txt", sep = "\t")

pheno = load_phenoData("pheno.txt", sep = "\t", tran=TRUE)

ord = match(colnames(counts$counts), rownames(pheno))

pheno = pheno[ord, ]

head(phono)

taxa = read.delim("taxa.txt", sep = "\t", stringsAsFactors=FALSE)

phenotypeData = AnnotatedDataFrame(pheno)

taxaData = AnnotatedDataFrame(taxa)

AMR = newMRexperiment(counts$counts,phenoData=phenotypeData, featureData=taxaData)

Data from counts.txt:

Data from taxa.txt:

Hopefully I am not making a very obvious mistake and thus wasting anyone's time with this error! Thank you in advance for your time.

Hi Noelle,

Thank you for the informative post! Looking at the line reading in the taxa, your taxa data rownames will be 1,2,3,4,..., etc.

Unfortunately, that's not how the count table row names are encoded. If you change reading your taxa in to:

taxa = read.delim("taxa.txt", sep = "\t", stringsAsFactors=FALSE,row.names=1)

it should work. Also, I should mention we support biom-format.