Hello,
I hope someone can help me. I am fairly new to R and I just finished my first coding.
Phyloseq gave me one phylum [Thermi] because of the square brackets it is not possible for me to subsample it (or do anything else with it).
Is there any method to change [Thermi] into Thermi?
Thank you!
This will seem like magic at first. Applying it to your object will depend on whether it's a freestanding character string, as shown below, or part of a data frame.
library(stringr)
target <- "[Thermi]"
target
#> [1] "[Thermi]"
pattern <- "[:alpha:]+"
pattern
#> [1] "[:alpha:]+"
str_extract(target,pattern)
#> [1] "Thermi"
Created on 2020-11-02 by the reprex package (v0.3.0.9001)
Thank you!
The magic worked with your example, but unfortunately not with my data. Usually that happens, when I did something wrong or somewhere is a typo, but I can't figure it out:
pattern <- "[:alpha:]+"
control <- phyloseq::subset_samples(gps_phylum, treatment ="Control")
control_otu <- data.frame(phyloseq::otu_table(control))
str_extract(control_otu, pattern)
#> [1] "c" "c" "c" "c" "c" "c" "c" "c" "c" "c" "c"
Warning:
In stri_extract_first_regex(string, pattern, opts_regex = opts(pattern)) :
argument is not an atomic vector; coercing
head(control_otu)
#>[Thermi] 0 0 0 0
#>Chloroflexi 0 0 0 0
#>Actinobacteria 0 24 3 0
I thought that might be a problem. What does
str(control_otu)
display?
str(control_otu)
#>'data.frame': 12 obs. of 11 variables:
$ Hauck.001.1A16: num 0 0 0 6 155 ...
$ Hauck.002.1B16: num 0 0 24 0 44 0 0 0 2 654 ...
$ Hauck.003.1C16: num 0 0 3 0 133 ...
$ Hauck.004.2A16: num 0 0 0 8 88 ...
$ Hauck.005.2B16: num 0 5 0 203 101 ...
$ Hauck.006.2C16: num 0 0 9 129 105 ...
$ Hauck.007.3A16: num 0 0 0 0 69 ...
$ Hauck.008.3B16: num 0 0 25 3 60 ...
$ Hauck.009.3C16: num 0 0 0 4 37 ...
$ Hauck.011.4B16: num 0 0 2 374 109 ...
$ Hauck.012.4C16: num 2 0 7 0 67 0 0 0 6 878 ...
``I had hoped this would show the name of the column with the phyla. Assuming there is one by that name, this works (reverses the selection logic now to exclude punctuation)
suppressPackageStartupMessages({
library(dplyr)
library(stringr)
})
target <- "[Thermi]"
pattern <- "[:punct:]"
my_data <- data.frame(phyla = c("[Thermi]", "Bryozoa"))
my_data %>% mutate(phyla = str_remove_all(phyla,pattern))
#> phyla
#> 1 Thermi
#> 2 Bryozoa
Created on 2020-11-02 by the reprex package (v0.3.0.9001)
I am so sorry! It didn't work.
library(dplyr)
library(stringr)
})
pattern <- "[:punct:]"
control_otu <- data.frame(phyla = c("[Thermi]", "Actinobacteria", "Bacteroides", "Chlamydiae",
"Chloroflexi", "Cyanobacteria", "Firmicutes", "Planctomyces", "Proteobacteria",
"Tenericutes", "TM6", "Verrucomicrobia"))
control_otu %>% mutate(phyla = str_remove_all(phyla,pattern))
#> phyla
1 Thermi
2 Actinobacteria
3 Bacteroides
4 Chlamydiae
5 Chloroflexi
6 Cyanobacteria
7 Firmicutes
head(control_otu)
#> head(control_otu)
phyla
1 [Thermi]
2 Actinobacteria
3 Bacteroides
4 Chlamydiae
5 Chloroflexi
6 Cyanobacteria
control_otu before:
Chloroflexi 0 0 0 0
Actinobacteria 0 24 3 0
Cyanobacteria 6 0 0 8
Proteobacteria 155 44 133 88
and after the [Thermi]-to-Thermi procedure:
phyla
1 Thermi
2 Actinobacteria
3 Bacteroides
4 Chlamydiae
5 Chloroflexi
6 Cyanobacteria
I noticed, that the [Thermi] show up also in the step before:
gps_phylum <- phyloseq::tax_glom(gps, "Phylum")
phyloseq::taxa_names(gps_phylum) <- phyloseq::tax_table(gps_phylum)[, "Phylum"]
phyloseq::otu_table(gps_phylum)[1:5, 1:5]
#Melt and plot
phyloseq::psmelt(gps_phylum) %>%
ggplot(data = ., aes(group=group, x = group, y = Abundance)) +
geom_boxplot(outlier.shape = NA) +
geom_jitter(aes(color = OTU), height = 0, width = .2) +
labs(x = "", y = "Abundance\n") +
facet_wrap(~ OTU, scales = "free")
This code gives me already otu [Thermi] in the plot.
It did. It just needs to be saved:
control_otu %>% mutate(phyla = str_remove_all(phyla,pattern)) -> control_otu
Worked!
But now I'm stuck at the next step:
control_otu <- control_otu %>%
t(.) %>%
as.data.frame(.) %>%
mutate(Other = Thermi + Chlamydiae + Chloroflexi + Cyanobacteria + Verrucomicrobia) %>%
dplyr::select(-Thermi, -Chlamydiae, -Chloroflexi, -Cyanobacteria, -Verrucomicrobia)
#Error: Problem with `mutate()` input `Other`.
#x Object 'Thermi' not found
#ℹ Input `Other` is `Thermi + Chlamydiae + Chloroflexi + Cyanobacteria + #Verrucomicrobia`.
#Run `rlang::last_error()` to see where the error occurred.
control_otu before:
[Thermi] 0 0 0 0
Chloroflexi 0 0 0 0
Actinobacteria 0 24 3 0
Cyanobacteria 6 0 0 8
Proteobacteria 155 44 133 88
and after the [Thermi]-to-Thermi procedure:
1 Thermi
2 Actinobacteria
3 Bacteroides
4 Chlamydiae
5 Chloroflexi
6 Cyanobacteria
I'm not sure how you intend to handle these. One way is to simply filter them out
suppressPackageStartupMessages({
library(dplyr)
library(stringr)
})
pattern <- "[:punct:]"
control_otu <- data.frame(phyla = c("[Thermi]", "Actinobacteria", "Bacteroides", "Chlamydiae", "Chloroflexi", "Cyanobacteria", "Firmicutes", "Planctomyces", "Proteobacteria", "Tenericutes", "TM6", "Verrucomicrobia"))
control_otu %>% mutate(phyla = str_remove_all(phyla, pattern)) -> control_otu
the_others <- c("Thermi", "Chlamydiae", "Chloroflexi", "Cyanobacteria", "Verrucomicrobia")
control_otu %>% filter(!(phyla %in% the_others))
#> phyla
#> 1 Actinobacteria
#> 2 Bacteroides
#> 3 Firmicutes
#> 4 Planctomyces
#> 5 Proteobacteria
#> 6 Tenericutes
#> 7 TM6
Created on 2020-11-02 by the reprex package (v0.3.0.9001)
If you want to retain the records, but just classify those taxa to "Other"
suppressPackageStartupMessages({
library(dplyr)
library(stringr)
})
pattern <- "[:punct:]"
control_otu <- data.frame(phyla = c("[Thermi]", "Actinobacteria", "Bacteroides", "Chlamydiae", "Chloroflexi", "Cyanobacteria", "Firmicutes", "Planctomyces", "Proteobacteria", "Tenericutes", "TM6", "Verrucomicrobia"))
control_otu %>% mutate(phyla = str_remove_all(phyla, pattern)) -> control_otu
the_others <- c("Thermi", "Chlamydiae", "Chloroflexi", "Cyanobacteria", "Verrucomicrobia")
control_otu %>% mutate(phyla = ifelse(phyla %in% the_others,"Other",phyla))
#> phyla
#> 1 Other
#> 2 Actinobacteria
#> 3 Bacteroides
#> 4 Other
#> 5 Other
#> 6 Other
#> 7 Firmicutes
#> 8 Planctomyces
#> 9 Proteobacteria
#> 10 Tenericutes
#> 11 TM6
#> 12 Other
Created on 2020-11-02 by the reprex package (v0.3.0.9001)
In this step the data is not only selected, but also converted:
control_otu <- control_otu %>%
t(.) %>%
as.data.frame(.) %>%
mutate(Other = Thermi + Chlamydiae + Chloroflexi + Cyanobacteria + Verrucomicrobia) %>%
dplyr::select(-Thermi, -Chlamydiae, -Chloroflexi, -Cyanobacteria, -Verrucomicrobia)
head(control_otu)
# [Thermi] Chloroflexi Actinobacteria Cyanobacteria Proteobacteria Chlamydiae
1 0 0 0 6 155 0
2 0 0 24 0 44 0
3 0 0 3 0 133 0
After mutating and selecting comes the HMP test:
#HMP test
group_data <- list(control_otu, cocci_otu, ST_otu, both_otu)
xdc <- Xdc.sevsample(group_data)
xdc
1 - pchisq(.2769004, 5)
This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.
If you have a query related to it or one of the replies, start a new topic and refer back with a link.