how to create a phyloseq object based on a given otu_table and sample_names

[LuffyLuffy]

Hi, there, It is easy to reproduce example of ggrarecurve, however, when I try to create the phyloseq object based on my own data, errors come out. I wonder how to create such a formal phyloseq object for the ggrarecurve. Error in seq.default(0, sdepth, by = step) : 'to' must be a finite number data(test_otu_data) otu1<-as.data.frame(test_otu_data@otu_table) sd1<-as.data.frame(test_otu_data@sam_data) test_otu_data2<-phyloseq(otu_table(otu1,taxa_are_rows = T), sample_names(sd1)) ggrarecurve(test_otu_data2, indexNames=c("Chao1","ACE"), chunks=100) + theme(legend.spacing.y=unit(0.02,"cm"), legend.text=element_text(size=6))

[xiangpin]

You should use sample_data of phyloseq instead of sample_names. In fact, The functions of MicrobiotaProcess were designed to support the phyloseq class and data.frame (nrow sample * ncol feature). So you have two choose. For example in the following.

library(MicrobiotaProcess)
library(ggplot2)
library(phyloseq)

# For phyloseq class
data(test_otu_data)
otu1<-as.data.frame(test_otu_data@otu_table)
sd1<-as.data.frame(test_otu_data@sam_data)

test_otu_data2<-phyloseq(otu_table(otu1,taxa_are_rows = T),
                         sample_data(sd1))
# This is to reproduce
set.seed(1024)
p1 <- ggrarecurve(test_otu_data2,
            indexNames=c("Chao1","ACE"),
            chunks=100) +
      guides(color=guide_legend(keywidth=0.5, keyheight=0.5))+
      theme(legend.spacing.y=unit(0.02,"cm"),
            legend.text=element_text(size=6))

# For data.frame
otu2 <- data.frame(t(test_otu_data@otu_table), check.names=FALSE)
sd2 <- data.frame(test_otu_data@sam_data, check.names=FALSE)
set.seed(1024)
p2 <- ggrarecurve(obj=otu2,
                  sampleda=sd2,
                  indexNames=c("Chao1","ACE"),
                  chunks=100) +
      guides(color=guide_legend(keywidth=0.5, keyheight=0.5))+
      theme(legend.spacing.y=unit(0.02,"cm"),
            legend.text=element_text(size=6))
p <- p1 + p2
p

[LuffyLuffy]

Thanks ! it works! By the way, your rarefaction method is really fast! Previously, I try to use Rarefy in package GUniFrac, and make an iteration for 99 runs in different sampling depth, and the ACE and Chao1 index is not able to be smooth. Your method is great!

[holypeggy]

Hi, xiangpin I am appreciate if you can provide a sample in the ggrarecurve mapping setting, as i want to use the stat_smooth as you did in your recently publised paper. Thanks.

[xiangpin]

If you want to visualize the curve of group, first you should rare phyloseq object.

library(MicrobiotaProcess)
library(phyloseq)
library(ggplot2)

data(test_otu_data)
test_otu_data
set.seed(1024)
ps <- rarefy_even_depth(test_otu_data)
ps

p <- ggrarecurve(
                 ps,
                 factorNames="group",
                 indexNames=c("Observe", "Chao1", "ACE"),
                 chunks=800,
     ) +
     theme_bw()+
     theme(axis.text=element_text(size=8), panel.grid=element_blank(),
           strip.background = element_rect(colour=NA,fill="grey"),
           strip.text.x = element_text(face="bold"))
p

[holypeggy]

It works. Thank you very much!

[Adamyazori]

Hi there, I am getting the error below, meanwhile I have all my packages installed. Any idea on what am missing. Thanks

Error in get_rarecurve(obj = ps, chunks = 400) : could not find function "get_rarecurve"

[xiangpin]

The get_rarecurve is the function in newest(development) version. Please update MicrobiotaProcess following this.

[Adamyazori]

Thank you very much . That works.