Converting mothur's groups and names files into Unifrac sample ID mapping file

[Roey Angel]

Hi,
Does anyone here have a script (perl/ python) for converting mothur's groups and names files into Unifrac sample ID mapping file?

Thanks in advance,
Roey

[Greg Caporaso]

Hi Roey,
We don't have anything like that in QIIME, but would be interested in
incorporating it if someone from the user community wants to
contribute one. On a related note, mothur now supports the biom
format, so it should be easy to move otu tables back and forth between
the tools.

Greg

[Roey Angel]

Hi Greg,
Thanks for your reply.

For the R users, here's a quick function that would convert mothur's .count file (the output of count.seqs) to UniFrac sample ID mapping file.

in mothur, run something like
count.seqs(name=final.names, group=final.groups)

then in R

Counts2ID <- function(data, filename = "sample", convert = FALSE) {
  require(reshape2)
  sample.ID <- subset(data, select = -c(total)) # remove column 'total'
  sample.ID <- melt(sample.ID, id = "Representative_Sequence") # melt data frame
  names(sample.ID) <- c("Sequence.ID", "Sample.ID", "Sequence.abundance") # rename variables
  sample.ID <- sample.ID[sample.ID$Sequence.abundance > 0, ] # remove 0 values
  sample.ID <- sample.ID[order(sample.ID$Sequence.ID), ]
  write.table(sample.ID, file = paste(filename,".ID",sep = ""), quote = FALSE, sep = "\t", col.names = FALSE, row.names = FALSE)
  return()
}

filename <- "final.seq.count"
data <- read.table(filename, header=T) # read file
Counts2ID(data, filename)

I would've loved to contribute but haven't had the chance to properly learn python till now.

Roey

[Greg Caporaso]

Thanks Roey!

Greg