| Title: | Decipher Microbial Turnover along a Gradient |
|---|---|
| Description: | Two pipelines are provided to study microbial turnover along a gradient, including the beta diversity and microbial abundance change. The 'betaturn' class consists of the steps of community dissimilarity matrix generation, matrix conversion, differential test and visualization. The workflow of 'taxaturn' class includes the taxonomic abundance calculation, abundance transformation, abundance change summary, statistical analysis and visualization. Multiple statistical approaches can contribute to the analysis of microbial turnover. |
| Authors: | Chi Liu [aut, cre], Xiangzhen Li [ctb], Minjie Yao [ctb] |
| Maintainer: | Chi Liu <[email protected]> |
| License: | GPL-3 |
| Version: | 0.3.1 |
| Built: | 2024-10-28 05:06:06 UTC |
| Source: | https://github.com/chiliubio/mecoturn |
Analyze the 'turnover' of microbial communities, i.e. beta-diversity along a gradient <doi:10.1111/j.1461-0248.2010.01552.x>. The workflow consists of the steps of dissimilarity matrix generation, matrix conversion, differential test and visualization.
new()
betaturn$new( dataset, measure = "bray", filter_thres = 0, abundance.weighted = TRUE, null.model = NULL, runs = 1000, iterations = 1000, ... )
datasetthe object of microtable class.
measuredefault "bray"; beta diversity dissimilarity metric;
must be one of c("bray", "jaccard", "wei_unifrac", "unwei_unifrac", "betaMPD", "betaMNTD", "betaNRI", "betaNTI", "ses_UniFrac", "RCbray")
or other options in parameter method of vegan::vegdist function.
If the distance matrix has been in the beta_diversity list of microtable object,
the function can ignore this step. Otherwise, the function can generate the corresponding beta diversity distance matrix in the microtable object.
bray: Bray-Curtis; RCbray: Raup–Crick based Bray-Curtis; wei_unifrac: weighted UniFrac; ses_UniFrac: standardized deviation of UniFrac.
filter_thresdefault 0; the relative abundance threshold used to filter features with low abundance.
abundance.weighteddefault TRUE; whether use abundance-weighted method for the phylogenetic metrics.
null.modeldefault NULL; one of c("taxa.labels", "richness", "frequency", "sample.pool", "phylogeny.pool", "independentswap", "trialswap"),
in which "taxa.labels" can only be used for phylogenetic analysis.
See null.model parameter of ses.mntd function in picante package for the algorithm details.
runsdefault 1000; simulation number of times for null model.
iterationsdefault 1000; iteration number for part null models to perform; see iterations parameter of picante::randomizeMatrix function.
...parameters passed to cal_betadiv function of microtable class when provided measure is not in the current vector;
parameters passed to cal_betamntd function of trans_nullmodel class when measure = "betaMNTD";
parameters passed to cal_ses_betamntd function of trans_nullmodel class when measure = "betaNTI".
dataset, stored in the object. The new dataset has a beta_diversity list and the calculated distance matrix in the list.
data(wheat_16S) b1 <- betaturn$new(wheat_16S, measure = "bray")
cal_group_distance()
Convert sample distances within groups or between groups.
betaturn$cal_group_distance( group, within_group = TRUE, by_group = NULL, ordered_group = NULL, sep = " vs ", add_cols = NULL )
groupone colname of sample_table in microtable object used for group distance convertion.
within_groupdefault TRUE; whether transform sample distance within groups? If FALSE, transform sample distances between any two groups.
by_groupdefault NULL; one colname of sample_table in microtable object.
If provided, convert distances according to the provided by_group parameter. This is especially useful for ordering and filtering values further.
When within_group = TRUE, the result of by_group parameter is the format of paired groups.
When within_group = FALSE, the result of by_group parameter is the format same with the group information in sample_table.
ordered_groupdefault NULL; a vector representing the ordered elements of group parameter; only useful when within_group = FALSE.
sepdefault TRUE; a character string to separate the group names after merging them into a new name.
add_colsdefault NULL; add several columns of sample_table to the final res_group_distance table according to the by_group column;
invoked only when within_group = FALSE.
res_group_distance stored in object.
b1$cal_group_distance(group = "Type", within_group = FALSE, by_group = "Plant_ID")
cal_group_distance_diff()
Differential test of distances among groups.
betaturn$cal_group_distance_diff(...)
...parameters passed to cal_group_distance_diff function of trans_beta class.
res_group_distance_diff stored in object.
b1$cal_group_distance_diff(method = "wilcox")
plot_group_distance()
Plot the distance between samples within or between groups.
betaturn$plot_group_distance(...)
...parameters passed to plot_group_distance function of trans_beta class.
ggplot.
b1$plot_group_distance()
clone()
The objects of this class are cloneable with this method.
betaturn$clone(deep = FALSE)
deepWhether to make a deep clone.
## ------------------------------------------------ ## Method `betaturn$new` ## ------------------------------------------------ data(wheat_16S) b1 <- betaturn$new(wheat_16S, measure = "bray") ## ------------------------------------------------ ## Method `betaturn$cal_group_distance` ## ------------------------------------------------ b1$cal_group_distance(group = "Type", within_group = FALSE, by_group = "Plant_ID") ## ------------------------------------------------ ## Method `betaturn$cal_group_distance_diff` ## ------------------------------------------------ b1$cal_group_distance_diff(method = "wilcox") ## ------------------------------------------------ ## Method `betaturn$plot_group_distance` ## ------------------------------------------------ b1$plot_group_distance()## ------------------------------------------------ ## Method `betaturn$new` ## ------------------------------------------------ data(wheat_16S) b1 <- betaturn$new(wheat_16S, measure = "bray") ## ------------------------------------------------ ## Method `betaturn$cal_group_distance` ## ------------------------------------------------ b1$cal_group_distance(group = "Type", within_group = FALSE, by_group = "Plant_ID") ## ------------------------------------------------ ## Method `betaturn$cal_group_distance_diff` ## ------------------------------------------------ b1$cal_group_distance_diff(method = "wilcox") ## ------------------------------------------------ ## Method `betaturn$plot_group_distance` ## ------------------------------------------------ b1$plot_group_distance()
Analyze the 'turnover' of taxa along a defined gradient. The workflow of taxaturn class includes the taxonomic abundance calculation, abundance transformation, abundance change summary, statistical analysis and visualization.
new()
taxaturn$new( dataset, taxa_level = "Phylum", group, ordered_group, by_ID = NULL, by_group = NULL, filter_thres = 0 )
datasetthe object of microtable class.
taxa_leveldefault "Phylum"; taxonomic rank name, such as "Genus". An integer is also acceptable.
If the provided taxa_level is not found in taxa_abund list,
the function will invoke the cal_abund function to obtain the relative abudance automatically.
groupsample group used for the selection; a colname of input microtable$sample_table.
ordered_groupa vector representing the ordered elements of group parameter.
by_IDdefault NULL; a column of sample_table used to obtain the consistent change along provided elements. So by_ID can be ID (unique repetition) or even group (with repetitions). If it denotes unique ID, consistent change can be performed across each ID. It is also especially useful for the paired wilcox test (or paired t test) in the following analysis. If it does not represent unique ID, the mean of each group will be calculated, and consistent change across groups will be obtained.
by_groupdefault NULL; NULL or other colname of sample_table of input dataset used to show the result for different groups;
NULL represents the output is the default consistent change across all the elements in by_ID;
a colname of sample_table of input dataset means the consistent change is obtained for each group instead of all the elements in by_group;
Note that the by_group can be same with by_ID, in which the final change is the result of each element in by_group.
So generally by_group has a larger scale than by_ID parameter in terms of the sample numbers in each element.
filter_thresdefault 0; the mean abundance threshold used to filter features with low abudance.
res_abund, res_change_pair and res_change in the object:
res_abundThe Mean, SD or SE of abundances for all the samples or each group. Mean: mean of abudances; SD: standard deviation; SE: standard error.
res_change_pairThe difference value of abudances between two niches, i.e. the latter minus the former.
res_changeThe summary of the abudance change results in res_change_pair.
data(wheat_16S)
t1 <- taxaturn$new(wheat_16S, taxa_level = "Phylum", group = "Type",
ordered_group = c("S", "RS", "R"), by_ID = "Plant_ID", filter_thres = 0.01)
cal_diff()
Differential test of taxonomic abundance across groups
taxaturn$cal_diff(
method = c("wilcox", "t.test", "anova", "betareg", "lme", "glmm")[1],
group2num = FALSE,
...
)methoddefault "wilcox"; see the following available options:
Wilcoxon Rank Sum and Signed Rank Tests for all paired groups
Student's t-Test for all paired groups
one-way or multi-way anova
Beta Regression based on the betareg package
lme: Linear Mixed Effect Model based on the lmerTest package
Generalized linear mixed model (GLMM) based on the glmmTMB package with the beta family function,
i.e. family = glmmTMB::beta_family(link = "logit").
For more parameters, please see glmmTMB::glmmTMB function.
In the return table, Conditional_R2 and Marginal_R2 represent total variance (explained by both fixed and random effects) and the variance explained by
fixed effects, respectively. The significance of fixed factors are tested by Chi-square test from function car::Anova.
The significance of 'Estimate' in each term of fixed factors comes from the model.
group2numdefault FALSE; whether convert ordered groups to integer numbers when method is "lme" or "glmm".
...parameters passed to trans_diff$new.
res_change or res_diff in the object.
t1$cal_diff(method = "wilcox")
plot()
Plot the line chart.
taxaturn$plot(
select_taxon = NULL,
color_values = RColorBrewer::brewer.pal(8, "Dark2"),
delete_prefix = TRUE,
plot_type = c("point", "line", "errorbar", "smooth")[1:3],
errorbar_SE = TRUE,
rect_fill = TRUE,
rect_color = c("grey70", "grey90"),
rect_alpha = 0.2,
position = position_dodge(0.1),
errorbar_size = 1,
errorbar_width = 0.1,
point_size = 3,
point_alpha = 0.8,
line_size = 0.8,
line_alpha = 0.8,
line_type = 1,
...
)select_taxondefault NULL; a taxon name.
Note that if delete_prefix is TRUE, the provided select_taxon should be taxa names without long prefix (those before |);
if delete_prefix is FALSE, the select_taxon should be full names same with those in the res_abund of the object.
color_valuesdefault RColorBrewer::brewer.pal(8, "Dark2"); colors palette for the plotting.
delete_prefixdefault TRUE; whether delete the prefix in the taxa names.
plot_typedefault c("point", "line", "errorbar", "smooth")[1:3]; a vector of visualization types. Multiple elements are available.
'smooth' denotes the fitting with geom_smooth function of ggplot2 package.
errorbar_SEdefault TRUE; TRUE: plot the errorbar with mean ± se; FALSE: plot the errorbar with mean ± sd.
rect_filldefault TRUE; Whether fill color in each rectangular area.
rect_colordefault c("grey70", "grey90"); the colors used to fill different rectangular area.
rect_alphadefault 0.2; the fill color transparency in rectangular area.
positiondefault position_dodge(0.1); Position adjustment for the points and lines, either as a string (such as "identity"), or the result of a call to a position adjustment function.
errorbar_sizedefault 1; errorbar size.
errorbar_widthdefault 0.1; errorbar width.
point_sizedefault 3; point size for taxa.
point_alphadefault 0.8; point transparency.
line_sizedefault 0.8; line size.
line_alphadefault 0.8; line transparency.
line_typedefault 1; an integer; line type.
...parameters passed to geom_smooth when 'smooth' is in plot_type parameter.
ggplot2 plot.
t1$plot()
clone()
The objects of this class are cloneable with this method.
taxaturn$clone(deep = FALSE)
deepWhether to make a deep clone.
## ------------------------------------------------ ## Method `taxaturn$new` ## ------------------------------------------------ data(wheat_16S) t1 <- taxaturn$new(wheat_16S, taxa_level = "Phylum", group = "Type", ordered_group = c("S", "RS", "R"), by_ID = "Plant_ID", filter_thres = 0.01) ## ------------------------------------------------ ## Method `taxaturn$cal_diff` ## ------------------------------------------------ t1$cal_diff(method = "wilcox") ## ------------------------------------------------ ## Method `taxaturn$plot` ## ------------------------------------------------ t1$plot()## ------------------------------------------------ ## Method `taxaturn$new` ## ------------------------------------------------ data(wheat_16S) t1 <- taxaturn$new(wheat_16S, taxa_level = "Phylum", group = "Type", ordered_group = c("S", "RS", "R"), by_ID = "Plant_ID", filter_thres = 0.01) ## ------------------------------------------------ ## Method `taxaturn$cal_diff` ## ------------------------------------------------ t1$cal_diff(method = "wilcox") ## ------------------------------------------------ ## Method `taxaturn$plot` ## ------------------------------------------------ t1$plot()
The dataset wheat_16S is structured with microtable class for the demonstration of examples.
data(wheat_16S)data(wheat_16S)
An R6 class object
sample_table: sample information table
otu_table: species-community abundance table
tax_table: taxonomic table
phylo_tree: phylogenetic tree