itsmisterbrown/microfiltr

accessory functions for processing microbial community data

microfiltR: marker gene processing, filtering and exporting functions

accessory functions for processing microbial community data

for a sample workflow, see README.ipynb

The associated manuscript is currently in the pipeline, so feel free to cite however you'd like until then :)

microfiltr's Issues

Issues sourcing the function

Hi,

I am getting an error message when I try to source the function using this line of code:

source("/Users/bpb/Documents/GitHub/microfiltR/microfiltR_source_code.R")

Error message:
Error in file(filename, "r", encoding = encoding) :
cannot open the connection
In addition: Warning message:
In file(filename, "r", encoding = encoding) :
cannot open file '/Users/bpb/Documents/GitHub/microfiltR/microfiltR_source_code.R': No such file or directory

make.dataset.biom metadata invalid format

Hello

Firstly, thanks for making a very easy use package to move phyloseq objects into the biom format!
Unfortunately, I think I found some bugs when metadata is used

The problem:
I was able to successfully use make.dataset.biom(ps=ps, paths ect, writeFASTA=TRUE, rename=TRUE) to generate the biom v1 table but when i tried to import this file into Qiime2 it couldn't parse the metadata. Looking at the file text itself I found my data had been formatted in the following way:
"columns":[{"id":"S1P1PM","metadata":["Plantation","Mineral","Replicate","Site 1"]}....
where plantation, mineral, replicate and site are variables within specific columns of my physeq metadata

comparing this to the example "rich dense OTU table" within the biom docs, some problems become obvious.
e.g.:
"columns":[
{"id":"Sample1", "metadata":{
"BarcodeSequence":"CGCTTATCGAGA",
"LinkerPrimerSequence":"CATGCTGCCTCCCGTAGGAGT",
"BODY_SITE":"gut",
"Description":"human gut"}},
Firstly the list after metadata should be created with curly braces instead of square braces
Secondly each variable should be predicated by the variable type so that as an example with my data it would look like
"columns":[{"id":"S1P1PM","metadata":{"Location":"Plantation","Layer":"Mineral","Treatment":"Replicate","Site":"Site 1"}}....

Thanks in advance!

taxa_are_rows bug in write.dataset.biom function

the current version (data = as.matrix(t(phyloseq::otu_table(ps)))) can only deal with ps object whose taxa are in column.

should be:

 #generate biom file
 suppressWarnings(ps.b <- biomformat::make_biom(
      if (taxa_are_rows(ps)) {
          data = as.matrix(phyloseq::otu_table(ps))
  } else {
      data = as.matrix(t(phyloseq::otu_table(ps)))
  },

sample_metadata = as.data.frame(phyloseq::sample_data(ps)),
observation_metadata = as.data.frame(phyloseq::tax_table(ps)), 
matrix_element_type = "int"
  )
  )