To get the tool clone the git repository:
git clone https://github.com/dsmutin/samovar.gitIn your R session, source all functions:
source(PATH/TO/SAMOVAR/scripts/source.R)Functions for data preparation and their descriptions are avialable in /functions Manipulations to visualize data is avialable in /plots Test pipeline via /test repositorium
Enjoy beauty of R plots!
get conut and amount table from kraken2 report
path, #path to k2 reports
keep_unclassified = T, #keep or not unclassified reads from report as new taxa
pattern = "", #optional: pattern of file names to use
trim_char = F, #optional: character to split file name in final table
clade = F, #optional: which clade (like "O", "F", "G", "S" to use)
type = k2, #character, kraken2 or kaiju
legend = F #optional: path to legend file in csv format:
#legend$name == name of kraken report (OR slited file name!), rest of columns - legend itself
result is data.frame with columns:
#taxa,
#clade,
#sample,
#number of reads,
#ammount of all reads,
#amount of classified reads
#and several columns of applied legend
get untidy table for base::heatmap or future processing
df, #result tibble from get_counts.R
counts = T, #use counts or amount data
classified_only = F, #only if counts = F, get amount by classified data
drop_unclassified = F, #drop all unclasified levels
keep_sample_name = F #warning! if T translate legend to separated format in name of each column :(
trim taxa by their amount
df, #result tibble from get_counts.R
top_taxa = 10 #how many taxa to trim
#attention! function remove all [taxa]*, keep only "S" and similar taxa
get top taxa from result data frame of different clade
df, #result tibble from get_counts.R
clade, #clade
top, #count of top taxa to use
rop all unclassified tables
df, #result tibble from get_counts.R
rescale all amounts to 1. neccessary if you drop some taxa levels
df, #result tibble from get_counts.R
get parents taxas to df
df, #result tibble from get_counts.R
additional function to get all unclassified level down
df, #result tibble from get_counts.R
Calculate all alpha-diversity metrix. To use only part of them, use df2alpha
Calculate all beta-diversity matrix
Calculate PCA based on beta-diversity matrix
(image below based on another dataset)
donut plot of the composition
bar plot of the composition
box plot of the composition
cluster plot. untidy table input
now depricated
2D cluser visualisation using legend. untidy table input
heatmap plot. untidy table input
corrplot. untidy table input. scaling required
ggraph and circlize connection plot
#
tsne plot. untidy table input
#
volcano plot based on log amount change between groups
(image below based on another dataset)
#
pca plot for samples. untidy table input
#
pca plot for taxa. untidy table input
# Diversity:
df2alpha(df, split_by = F, add_legend = F, ...)
df2alpha_summary(df, split_by = F, add_legend = F,
alpha_function_list = list("Shannon" = shannon, "Simpson" = simpson), ...)
df2beta <- function(df, clade = F, dist_function = bray_curtis,
treshhold_up = 1,treshhold_down = 0,
add_legend = F, add_labels= F, print_df = F,
heatmap_type = "heatmap", ...)
df2beta_pcoa <- function(df,
dist_function = bray_curtis,
treshhold_up = 1, treshhold_down = 0,
add_legend = F, add_ellipse = F,...)
# Basic view:
df2donut(df, ...) #ggplot2 donut plot
df2composition(df, ...) #ggplot2 bar plot
df2barplot(df, ...) #ggplot2 box plot. c:
# Interspecies crosstalk
df2cluster(df, k_means = 2, use = "sp", ...) #base R cluster plot
df2clust2d(dfU, legend_detect, clade = F, k_means = 5, counts = F, ...) #2D cluser visualisation using legend
df2heatmap(dfU,clade, trim = F, counts = F, ...) #base R heatmap plot
df2corrplot(df, clade = F, counts = F, ...) #corrplot. using scale
df2chord(df, clade = F, k_means = 5, counts = F,
coenf_level = F, coenf, line_as_clusters = F, ...) #ggraph and circlize connection plot
df2tsne(df, clade_trim, clade_color = "P", counts = T,
taxa = NA, trim_taxa = T, only_input_taxa = F) #ggplot2 tsne plot
df2pca_sample(df, clade = F, scale = T, ...) #factoextra pca plot for samples
df2pca_sp (df, clade = F, scale = T, ...) #factoextra pca plot for taxa
df - data.frame in format above
dfU - data.frame after df_untidy transformation. numeric matrix with samples in column and species in row names
split_by: column number to split the plot
add_legend: column number(s) to add color legend
add_label: column number(s) to add label legend
add_ellipse: column number(s) to add ellipse legend
treshhold_up: remove all values above the level from df
treshhold_down: remove all values below the level from df
dist_function: function to calculate diversity measure. one from abdiv package
print_df: return data.frame insted of plot
clade: character. which clade to use(e.g. "S", "G", "F")
clade_color: character. which clade to usefor coloring (e.g. "S", "G", "F")
drop_unclassified: logical. drop unclassified
counts: logical. use counts or amount
k_means: numeric. number of k-means to use
use: character. one of "sp" and "sample"
legend_detect: character. what to detect in legend to divied by two groups
coenf_level: logical or numeric. trim connections according to level
coenf: character. one of "upper", "lower" or "both" to trim in coenf_level
only_input_taxa: logical. use or not only inputed taxa for labels
taxa: character. what taxa to show on labels
trim_taxa: logical. get only 1 word from the taxa to label. better use for species
scale: logical. scale or not the data to get z-scores
To cite the tool, please use
Smutin D, Taldaev A, Lebedev E, Adonin L. Shotgun Metagenomics Reveals Minor Micro“bee”omes Diversity Defining Differences between Larvae and Pupae Brood Combs. International Journal of Molecular Sciences. 2024; 25(2):741. https://doi.org/10.3390/ijms25020741
or import BibText:
@Article{ijms25020741,
AUTHOR = {Smutin, Daniil and Taldaev, Amir and Lebedev, Egor and Adonin, Leonid},
TITLE = {Shotgun Metagenomics Reveals Minor Micro“bee”omes Diversity Defining Differences between Larvae and Pupae Brood Combs},
JOURNAL = {International Journal of Molecular Sciences},
VOLUME = {25},
YEAR = {2024},
NUMBER = {2},
ARTICLE-NUMBER = {741},
URL = {https://www.mdpi.com/1422-0067/25/2/741},
PubMedID = {38255816},
ISSN = {1422-0067},
DOI = {10.3390/ijms25020741}
}
Work on libraries:
- tidyverse
- ggrepel
- viridis
- corrplot
- ggraph
- igraph
- circlize
- factoextra
- Rtsne