Some batches contain zero or one samples.

[nhg12]

Hi

I am trying to use decontam on my data and I keep getting this error:

Error in isContaminant(physeq2, method = "frequency", conc = "quant_reading") : Some batches contain zero or one sample.

I have checked my data there are no zero quantity values and I have removed the samples with zero OTUs.

What can be going wrong? Please help

[benjjneb]

Do any of your batches contain one sample or less?

You can inspect that by table(my_data_frame$batch-column)

[nhg12]

It gives this error: Error in table(otu_table$batch - column) : object 'column' not found

[nhg12]

This is what I have been doing so far: ##load data in R## I've converted the BIOM ASVs file form QIIME to TSV and then to csv to load in R otu_table <- read.csv(file.choose(), header= TRUE, row.names = 1) taxonomy <- read.csv(file.choose(), header=T, row.names = 1) sample_data <- read.csv(file.choose(), row.names = 1, header =T)

###converted to a phyloseq object ###

count_tab <- otu_table(otu_table, taxa_are_rows=T) sample_info <- sample_data(sample_data) physeq2 <- phyloseq(count_tab, sample_info)

run with decontam

contamdf.freq <- isContaminant(physeq2, method="frequency", conc="quant_reading")

and then it gives me the error

Error in isContaminant(physeq2, method = "frequency", conc = "quant_reading") : Some batches contain zero or one sample.

[benjjneb]

How many samples are in physeq2? You can just physeq2 to get a summary of what's stored in that object.

[nhg12]

phyloseq-class experiment-level object otu_table() OTU Table: [ 8397 taxa and 1 samples ] sample_data() Sample Data: [ 1 samples by 2 sample variables ]

[benjjneb]

That's the problem. Your physeq2 object has just one sample in it. This is probably an issue with how you are exporting data from QIIME and importing it into R.

[nhg12]

HI

So the files that I upload have all the samples. I then do this:

count_tab <- otu_table(otu_table, taxa_are_rows=T) sample_info <- sample_data(sample_data) physeq2 <- phyloseq(count_tab, sample_info)

This is when it considers them as 1 sample. What am I doing wrong?

I have made a biom file and then a tsv file and then a csv file to import in R

[benjjneb]

I'm not sure. Is this being exported from QIIME2? If so I think they have developed documentation about how to export QIIME2 data into R as a phyloseq object that you should take a look at.

In fact someone's made an R pakage for just that purpose: https://github.com/jbisanz/qiime2R