Error in data.frame and Removed 3 samples with zero total counts (or frequency)

[Paty-Azevedo]

Hi, I need some help with my analysis. I'm trying to plot the frequency of contaminants with Decontam but I have some errors: Error in data.frame(..., check.names = FALSE) : arguments imply differing number of rows: 64, 67 In addition: Warning message: In plot_frequency(physeq1, taxa_names(physeq1)[c(1, 96)], conc = "QUANT_READING") : Removed 3 samples with zero total counts (or frequency). The error in the data frame occurred because in the previous step the program exclude 3 samples with 0 reads and now the data do have not the same quantity of rows. How can I fix that to proceed with the analysis? Thank you so much!

[benjjneb]

Can you clarify the code you are running to get to this error?

Also some description of the underlying data, e.g. number of samples/taxa, and what type of DNA quantitation data you are working with, perhaps also with a summary of that column.

[Paty-Azevedo]

Yes! First I ran this decontam code, and the program excluded 3 samples with 0 total counts: contfreq1 <- isContaminant(physeq1, method="frequency", conc="QUANT_READING") Warning messages: 1: In .is_contaminant(seqtab, conc = conc, neg = neg, method = method, : Removed 3 samples with zero total counts (or frequency). 2: In .is_contaminant(seqtab, conc = conc, neg = neg, method = method, : Removed 3 samples with zero total counts (or frequency).

So I continued the steps and in the step code of the plot the frequency I get this error: Error in data.frame(..., check.names = FALSE) : arguments imply differing number of rows: 64, 67 In addition: Warning message: In plot_frequency(physeq1, taxa_names(physeq1)[c(1, 96)], conc = "QUANT_READING") : Removed 3 samples with zero total counts (or frequency).

We sequenced the 16S genome of samples containing the gut of bees exposed to insecticides. I have 63 samples plus 4 blankets. In qiime demultiplexing, three blankets remain with 0 reads, and I continued with the file I get after qiime processing to work in decontam. Apparently, the program excluded these samples with 0 reads. I attached the images of qiime table-dada2 and of my metadata.

Thanks so much!

[benjjneb]

If you remove the samples yourself before running isContaminant then things should match up before and after.

e.g.

non.zero <- sample_sums(ps) > 0
ps.nonzero <- prune_samples(non.zero, ps)
# Then go forward with ps.nonzero

[Paty-Azevedo]

It worked! Thank you so much! When I continued the analysis I got another error in the presence-absence analysis:

physeq1.pa.neg <- prune_samples(sample_data(physeq1.pa)$CONTROL == "Control Sample", physeq1.pa) Error in h(simpleError(msg, call)) : error in evaluating the argument 'samples' in selecting a method for function 'prune_samples': sam_data slot is empty.

Can you help me with that? Thanks