The normalisation in getpcoa() function does not work

[fernando-aq]

Hi, I tried to use

PCoA_res <-
get_pcoa(obj = ps_object, distmethod = "bray")

It worked well. But when I do

PCoA_res <-
get_pcoa(obj = ps_object, distmethod = "bray", method = "log")

It says:
Error in h(simpleError(msg, call)) :
error in evaluating the argument 'object' in selecting a method for function 'otu_table': (subscript) logical subscript too long

[xiangpin]

I suggest you use the newest MicrobiotaProcess. Because MicrobiotaProcess has a significant change to work with tidyverse ecosystem better. The newest tutorial is here. You can first use mp_decostand to normalize the Abundance, then perform the pcoa analysis. The following is an example.

> library(MicrobiotaProcess)
> data(mouse.time.mpse)
> mouse.time.mpse
# A MPSE-tibble (MPSE object) abstraction: 4,142 × 11
# OTU=218 | Samples=19 | Assays=Abundance | Taxonomy=Kingdom, Phylum, Class, Order, Family, Genus, Species
   OTU    Sample Abundance time  Kingdom Phylum Class Order Family Genus Species
   <chr>  <chr>      <int> <chr> <chr>   <chr>  <chr> <chr> <chr>  <chr> <chr>
 1 OTU_1  F3D0         579 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
 2 OTU_2  F3D0         345 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
 3 OTU_3  F3D0         449 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
 4 OTU_4  F3D0         430 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
 5 OTU_5  F3D0         154 Early k__Bac… p__Ba… c__B… o__B… f__Ba… g__B… s__un_…
 6 OTU_6  F3D0         470 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
 7 OTU_7  F3D0         282 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
 8 OTU_8  F3D0         184 Early k__Bac… p__Ba… c__B… o__B… f__Ri… g__A… s__un_…
 9 OTU_9  F3D0          45 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
10 OTU_10 F3D0         158 Early k__Bac… p__Ba… c__B… o__B… f__Mu… g__u… s__un_…
# ℹ 4,132 more rows
# ℹ Use `print(n = ...)` to see more rows

This is original Abundance stored in a MPSE object. If you has a phyloseq, you can use as.MPSE to convert it to MPSE.
Then you can use mp_decostand to normalize it. Here, I use the clr method, log, hellinger, total etc can also be
supported. Then the result was added to the MPSE by default.

> mouse.time.mpse %>% mp_decostand(.abundance=Abundance, method='clr', pseudocount=1)
# A MPSE-tibble (MPSE object) abstraction: 4,142 × 12
# OTU=218 | Samples=19 | Assays=Abundance, clr | Taxonomy=Kingdom, Phylum, Class, Order, Family, Genus, Species
   OTU    Sample Abundance   clr time  Kingdom   Phylum Class Order Family Genus
   <chr>  <chr>      <int> <dbl> <chr> <chr>     <chr>  <chr> <chr> <chr>  <chr>
 1 OTU_1  F3D0         579  4.72 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
 2 OTU_2  F3D0         345  4.20 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
 3 OTU_3  F3D0         449  4.47 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
 4 OTU_4  F3D0         430  4.42 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
 5 OTU_5  F3D0         154  3.40 Early k__Bacte… p__Ba… c__B… o__B… f__Ba… g__B…
 6 OTU_6  F3D0         470  4.51 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
 7 OTU_7  F3D0         282  4.00 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
 8 OTU_8  F3D0         184  3.58 Early k__Bacte… p__Ba… c__B… o__B… f__Ri… g__A…
 9 OTU_9  F3D0          45  2.19 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
10 OTU_10 F3D0         158  3.43 Early k__Bacte… p__Ba… c__B… o__B… f__Mu… g__u…
# ℹ 4,132 more rows
# ℹ 1 more variable: Species <chr>
# ℹ Use `print(n = ...)` to see more rows

Next, let's use the normalized to perform the pcoa analysis

mouse.time.mpse %>% 
   mp_decostand(.abundance=Abundance, method='clr', pseudocount=1) %>% # normalization
   mp_cal_pcoa(.abundance=clr, distmethod='euclidean') %>%  # perform the pcoa
   mp_adonis(.abundance=clr, .formula=~time, distmethod='euclidean', action='add') %>% # perform adonis
  {.  ->> mpse.clr.pcoa} %>% # Save to a new mpse
   mp_plot_ord(.group=time, show.sample=T, show.adonis=T) # visualization

[fernando-aq]

Thank you so very much for your reply. I hope you are keep maintaining this.