does not work for "enterotype" and my data
2533245542 opened this issue · 4 comments
Hi, I tried using "enterotype" which comes from phyloseq and it got stucked at fix_duplicate_tax() and here is the related information:
For my data, it got stuck at parsePhyloseq,
I am sorry that I could not share my data but I am happy to follow any instructions.
It turns out that enterotype data does not have the full taxonomy information:
> head(tax_table(enterotype))
Taxonomy Table: [6 taxa by 1 taxonomic ranks]:
Genus
-1 NA
Bacteria NA
Prosthecochloris "Prosthecochloris"
Chloroflexus "Chloroflexus"
Dehalococcoides "Dehalococcoides"
Thermus "Thermus" As compared to GlobalPatterns:
> head(tax_table(GlobalPatterns))
Taxonomy Table: [6 taxa by 7 taxonomic ranks]:
Kingdom Phylum Class Order Family Genus
549322 "Archaea" "Crenarchaeota" "Thermoprotei" NA NA NA
522457 "Archaea" "Crenarchaeota" "Thermoprotei" NA NA NA
951 "Archaea" "Crenarchaeota" "Thermoprotei" "Sulfolobales" "Sulfolobaceae" "Sulfolobus"
244423 "Archaea" "Crenarchaeota" "Sd-NA" NA NA NA
586076 "Archaea" "Crenarchaeota" "Sd-NA" NA NA NA
246140 "Archaea" "Crenarchaeota" "Sd-NA" NA NA NA
Species
549322 NA
522457 NA
951 "Sulfolobusacidocaldarius"
244423 NA
586076 NA
246140 NA We currently need the taxonomic tree in the data. I will also point out a limitation that the parsers need the full dataset. If you remove Bacteroides genus, all species belonging to it should not be present to make the tree inferred complete.
Sure, thank you. I have not figured out the exact problem of my dataset. When I am trying to approach it from the other angles, I find there might be a big in annotation. In the standard example, I tried to annotate Crenarchaeota with red, and an error occurred. Here is what happenned for reproducible purpose:
library(phyloseq)
data("GlobalPatterns")
GP = GlobalPatterns
GP = transform_sample_counts(GlobalPatterns, function(otu) otu/sum(otu))
GP = filter_taxa(GP, function(x) max(x)>=0.01,TRUE)
GP = fix_duplicate_tax(GP)
tr = parsePhyloseq(GP)
p = tree.backbone(tr, size=1)
anno.data <- data.frame(node=c("Crenarchaeota"),
color='red', stringsAsFactors = FALSE)> clade.anno(p, anno.data)
Error in filter_impl(.data, quo) : Result must have length 271, not 0The parsePhyloseq function actually prepends the phylogenic level to the taxonomy name. So simply changing "Crenarchaeota" to "p__Crenarchaeota" solves the problem. I apologize for seeing this in seven months. Hopefully you figured something out...
Fixed by pull request #15