Package 'zitools'

Title: Analysis of zero-inflated count data
Description: zitools allows for zero inflated count data analysis by either using down-weighting of excess zeros or by replacing an appropriate proportion of excess zeros with NA. Through overloading frequently used statistical functions (such as mean, median, standard deviation), plotting functions (such as boxplots or heatmap) or differential abundance tests, it allows a wide range of downstream analyses for zero-inflated data in a less biased manner. This becomes applicable in the context of microbiome analyses, where the data is often overdispersed and zero-inflated, therefore making data analysis extremly challenging.
Authors: Carlotta Meyring [aut, cre]
Maintainer: Carlotta Meyring <[email protected]>
License: BSD_3_clause + file LICENSE
Version: 1.1.0
Built: 2024-10-31 06:44:19 UTC
Source: https://github.com/bioc/zitools

Help Index

Arithmetic Operators

Description

Arithmetic operators for a Zi-class object

Usage

## S4 method for signature 'Zi,ANY'
e1 * e2

Arguments

e1

Zi-class object, matrix or number
e2

Zi-class object, matrix or number

Value

a Zi-class object after a specific arithmetic operation is performed

Examples

data(mtx)
Zi <- ziMain(mtx)
Zi*Zi
Zi*2

Arithmetic Operators

Description

Arithmetic operators for a Zi-class object

Usage

## S4 method for signature 'Zi,ANY'
e1 / e2

Arguments

e1

Zi-class object, matrix or number
e2

Zi-class object, matrix or number

Value

a Zi-class object after a specific arithmetic operation is performed

Examples

data(mtx)
Zi <- ziMain(mtx)
Zi/Zi
Zi/2

Arithmetic Operators

Description

Arithmetic operators for a Zi-class object

Arithmetic operators for a Zi-class object

Usage

## S4 method for signature 'Zi,ANY'
e1 + e2

## S4 method for signature 'Zi,ANY'
e1 - e2

Arguments

e1

Zi-class object, matrix or number
e2

Zi-class object, matrix or number

Value

a Zi-class object after a specific arithmetic operation is performed

a Zi-class object after a specific arithmetic operation is performed

Examples

data(mtx)
Zi <- ziMain(mtx)
Zi+Zi
Zi+2
data(mtx)
Zi <- ziMain(mtx)
Zi+Zi
Zi+2

Access assays

Description

access assays of an Zi-class object if the inputdata is an object of the class SummarizedExperiment

Usage

## S4 method for signature 'Zi'
assays(x, withDimnames = TRUE, ...)

Arguments

x

Zi-class object
withDimnames

A logical, indicating whether the dimnames of the SummarizedExperiment object should be applied (i.e. copied) to the extracted assays. see assays
...

see assays

Value

list

Examples

data(mtx)
colData <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3',
    'Sample4', 'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9',
    'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
rowData <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
se <- SummarizedExperiment::SummarizedExperiment(assays = list(counts = mtx),
    colData = colData, rowData = rowData)
Zi <- ziMain(se)
assays(Zi)

Create boxplots of a 'Zi'-class object

Description

Create boxplots of a 'Zi'-class object.

Usage

## S3 method for class 'Zi'
boxplot(x, ...)

Arguments

x

'Zi'-class object
...

see boxplot.default

Value

A List with all information to create a boxplot see boxplot.default

See Also

boxplot.default

Examples

data(mtx)
Zi <- ziMain(mtx)
boxplot(Zi)
boxplot(log1p(Zi))

Access the col Data

Description

access the colData of an Zi-class object if the inputdata is an object of the class SummarizedExperiment

Usage

## S4 method for signature 'Zi'
colData(x, ...)

Arguments

x

Zi-class object
...

colData

Value

DFrame

See Also

colData

Examples

data(mtx)
colData <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3', 'Sample4',
    'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9', 'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
rowData <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
se <- SummarizedExperiment::SummarizedExperiment(assays = list(counts = mtx),
    colData = colData, rowData = rowData)
Zi <- ziMain(se)
colData(Zi)

Calculate the row or column means of zero-inflated count data

Description

Calculate row and column means of zero-inflated count data taking weights for structural zeros into account.

Usage

## S4 method for signature 'Zi'
colMeans2(x, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

## S4 method for signature 'Zi'
rowMeans2(x, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

Arguments

x

A Zi-class object
rows, cols

A vector indicating the subset of rows and/or columns to operate over. If NULL (default), no subsetting is done
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE
useNames

logical If TRUE (default), names attributes of result are set. Else if FALSE, no naming support is done.

Value

a numeric vector of row/column length

Examples

data(mtx)
Zi <- ziMain(mtx)
colMeans2(Zi)
rowMeans2(Zi)

Calculate the row or column median of zero-deinflated count data

Description

Calculate the row or column median of zero-deinflated data of a Zi-class object. To calculate the median, the deinflatedcounts matrix will be extracted.

Usage

## S4 method for signature 'Zi'
colMedians(x, rows = NULL, cols = NULL, na.rm = TRUE, ..., useNames = TRUE)

## S4 method for signature 'Zi'
rowMedians(x, rows = NULL, cols = NULL, na.rm = TRUE, ..., useNames = TRUE)

Arguments

x

Zi-class object
rows, cols

A vector indicating the subset of rows (and/or columns) to operate over. If NULL, no subsetting is done
na.rm

logical If TRUE NAs are excluded, otherwise not. default = TRUE
...

see colMedians
useNames

logical. If TRUE (default), names attributes of result are set. Else if FALSE, no naming support is done.

Value

returns a numeric vector of row/column length

Examples

data(mtx)
Zi <- ziMain(mtx)
colMedians(Zi, useNames = TRUE)
rowMedians(Zi, useNames = TRUE)

Calculate weighted Pearson Correlation coeffiecients

Description

calculate the weighted pearson correlation coefficients of a count matrix of an Zi object taking weights for zero counts into account

Usage

## S4 method for signature 'Zi,ANY'
cor(x, y = NULL, use = "everything", method = "pearson")

Arguments

x

'Zi'-class object
y

'Zi'-class object
use

'everything' see cor
method

default = 'pearson', weighted correlation only implemented for person correlation

Value

correlation matrix

Examples

data(mtx)
Zi <- ziMain(mtx)
cor(Zi)

Calculate weighted Covariance

Description

calculate the weighted covariance of the columns of the count matrix of an Zi object taking weights for possible structural zero counts into account

Usage

## S4 method for signature 'Zi,ANY'
cov(x, y = NULL, use = "everything")

Arguments

x

'Zi'-class object
y

'Zi'-class object
use

'everything'

Value

covariance matrix

Examples

data(mtx)
Zi <- ziMain(mtx)
cov(Zi)

Access the model

Description

access the deinflatedcounts of an Zi-class object

Usage

deinflatedcounts(x)

## S4 method for signature 'Zi'
deinflatedcounts(x)

deinflatedcounts(x) <- value

## S4 replacement method for signature 'Zi'
deinflatedcounts(x) <- value

Arguments

x

Zi-class object
value

deinflatedcounts object

Value

deinflatedcounts

Examples

data(mtx)
Zi <- ziMain(mtx)
deinflatedcounts(Zi)

Draw a Heat Map

Description

draw a heatmap of a given 'Zi'-class object, heatmap.Zi uses the deinflatedcounts matrix (drawn structural zeros) to produce a heatmap. NA values are white

Usage

## S3 method for class 'Zi'
heatmap(x, ...)

Arguments

x

'Zi'-class object
...

see heatmap

Value

heatmap

Examples

data(mtx)
Zi <- ziMain(mtx)
#heatmap(Zi) # Error, clustering not possible
heatmap(Zi, Rowv=NA) # no clustering of rows
heatmap(Zi, Rowv=NA, Colv=NA) # no clustering of rows and cols

Access the inputcounts

Description

access the inputcounts of an Zi-class object

Usage

inputcounts(x)

## S4 method for signature 'Zi'
inputcounts(x)

inputcounts(x) <- value

## S4 replacement method for signature 'Zi'
inputcounts(x) <- value

Arguments

x

Zi-class object
value

inputcounts object

Value

inputcounts

Examples

data(mtx)
Zi <- ziMain(mtx)
inputcounts(Zi)

Access and Set the inputdata

Description

access the inputdata of an Zi-class object

Usage

inputdata(x)

## S4 method for signature 'Zi'
inputdata(x)

inputdata(x) <- value

## S4 replacement method for signature 'Zi'
inputdata(x) <- value

Arguments

x

Zi-class object
value

inputdata object

Value

inputdata

Examples

data(mtx)
Zi <- ziMain(mtx)
inputdata(Zi)

log(1+x)

Description

Calculate log(1+x) of all 'matrix' objects of a 'Zi'-class object, log calculates by default natural logarithms

Usage

## S4 method for signature 'Zi'
log1p(x)

Arguments

x

Zi-class object

Value

a Zi-class object where the log(1+x) values of inputcounts, deinflatedcounts and weights are calculated.

See Also

log1p, log2p

Examples

data(mtx)
Zi <- ziMain(mtx)
log1p(Zi)

log2p(x+1)

Description

Calculate log2(x+1) of all 'matrix' objects of a 'Zi'-class object

Usage

log2p(x)

Arguments

x

Zi-class object

Value

a Zi-class object where the log2(1+x) values of inputcounts, deinflatedcounts and weights are calculated.

Examples

data(mtx)
Zi <- ziMain(mtx)
log2p(Zi)

Arithmetic Mean

Description

Calculate the arithmetic mean of zero inflated data taking weights for structural zeros into account

Usage

## S3 method for class 'Zi'
mean(x, ...)

Arguments

x

A Zi-class object
...

mean.default

Value

mean value

See Also

weighted.mean, colMeans2, rowMeans2

Examples

data(mtx)
Zi <- ziMain(mtx)
mean(Zi)

Calculate the median of zero-deinflated count data

Description

Caluclate the median of zero-deinflated data of a 'Zi'-class object. To calculate the median, the deinflatedcounts matrix will be extracted

Usage

## S3 method for class 'Zi'
median(x, na.rm = TRUE, ...)

Arguments

x

Zi-class object
na.rm

logical If TRUE NAs are excluded, otherwise not. default = TRUE
...

see median.default

Value

median value

See Also

median, colMedians, rowMedians

Examples

data(mtx)
Zi <- ziMain(mtx)
median(Zi)

Missing Value Heatmap

Description

Missing Value Heatmap

Usage

MissingValueHeatmap(ZiObject, title = "", xlab = "", ylab = "")

Arguments

ZiObject

ZiObject, result of the ziMain function
title

Title of the plot .
xlab

Title of the x axis.
ylab

Title of the y axis.

Value

heatmap

Examples

data(mtx)

Access the model

Description

access the model of an Zi-class object

Usage

model(x)

## S4 method for signature 'Zi'
model(x)

model(x) <- value

## S4 replacement method for signature 'Zi'
model(x) <- value

Arguments

x

Zi-class object
value

model object

Value

model

Examples

data(mtx)
Zi <- ziMain(mtx)
model(Zi)

Matrix Data

Description

Zero-inflated matrix data

Usage

mtx

Format

'mtx'

A matrix with 100 rows and 10 columns

Value

a data matrix

Access the otu table

Description

access the otu_table of an Zi-class object if the inputdata slot is a phyloseq object

Usage

## S4 method for signature 'Zi'
otu_table(object)

Arguments

object

Zi-class object

Value

otu_table

Examples

data(mtx)
OTU <- otu_table(mtx, taxa_are_rows = TRUE)
sample_data <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3',
    'Sample4', 'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9',
    'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
SAM <- sample_data(sample_data)
tax_table <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
TAX <- tax_table(tax_table)
ps <- phyloseq::phyloseq(OTU, TAX, SAM)
Zi <- ziMain(ps)
otu_table(Zi)

Plotting

Description

plot

Usage

## S4 method for signature 'Zi,ANY'
plot(x, y, ...)

Arguments

x

Zi-class object
y

the y coordinates of points in the plot, optional if x is an appropriate structure
...

Arguments to be passed to plot

Value

returns plot object

Examples

data(mtx)
Zi <- ziMain(mtx)
plot(Zi)

Calculate the quantiles of zero-deinflated count data

Description

Calculate the quantiles of zero-deinflated data of a Zi-class object. To calculate the quantiles, the deinflatedcounts matrix will be extracted.

Usage

## S3 method for class 'Zi'
quantile(x, probs = seq(0, 1, 0.25), na.rm = TRUE, ...)

Arguments

x

A Zi-class object
probs

A numeric vector of J probabilities in [0,1]
na.rm

logical If TRUE NAs are excluded, otherwise not. default = TRUE
...

quantile

Value

quantile value

See Also

quantile, rowQuantiles, colQuantiles

Examples

data(mtx)
Zi <- ziMain(mtx)
quantile(Zi)

Resample a Zi-class object

Description

Resample the deinflatedcounts matrix of an Zi-class object. Resampling is done by drawing from a binomial distribution with a given probability that a count value (zero and non-zero) is a structural zero.

Usage

resample_deinflatedcounts(x)

Arguments

x

Zi-class object

Value

a Zi-class object where the deinflatedcounts are resampled

Examples

data(mtx)
Zi <- ziMain(mtx)
resample_deinflatedcounts(Zi)

Access the row data

Description

access the rowData of an Zi-class object if the inputdata is an object of the class SummarizedExperiment

Usage

## S4 method for signature 'Zi'
rowData(x, useNames = TRUE, ...)

Arguments

x

Zi-class object
useNames

returns a rowData dataframe with rownames
...

rowData

Value

DFrame

Examples

data(mtx)
colData <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3', 'Sample4',
    'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9', 'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
rowData <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
se <- SummarizedExperiment::SummarizedExperiment(assays = list(counts = mtx),
    colData = colData, rowData = rowData)
Zi <- ziMain(se)
rowData(Zi)

Calculate the row or column quantiles of zero-deinflated count data

Description

Calculate the row or column quantiles of zero-deinflated data of a Zi-class object. To calculate the quantiles, the deinflatedcounts matrix will be extracted

Usage

## S4 method for signature 'Zi'
rowQuantiles(
  x,
  rows = NULL,
  cols = NULL,
  probs = seq(from = 0, to = 1, by = 0.25),
  na.rm = TRUE,
  type = 7L,
  ...,
  useNames = TRUE,
  drop = TRUE
)

## S4 method for signature 'Zi'
colQuantiles(
  x,
  rows = NULL,
  cols = NULL,
  probs = seq(from = 0, to = 1, by = 0.25),
  na.rm = TRUE,
  type = 7L,
  ...,
  useNames = TRUE,
  drop = TRUE
)

Arguments

x

A Zi-class object
rows, cols

A vector indicating the subset of rows and/or columns to operate over. If NULL (default), no subsetting is done.
probs

A numeric vector of J probabilities in [0,1]
na.rm

logical If TRUE NAs are excluded, otherwise not. default = TRUE
type

An integer specifying the type of estimator
...

Additional arguments passed to specific methods rowQuantiles
useNames

logical If TRUE (default), names attributes of result are set. Else if FALSE, no naming support is done.
drop

If TRUE a vector is returned if J == 1.

Value

a numeric vector of row/column length

Examples

data(mtx)
Zi <- ziMain(mtx)
rowQuantiles(Zi, useNames = TRUE)
colQuantiles(Zi, useNames = TRUE)

Row and Column Standard Deviations of zero inflated count data

Description

Calculate row and column standard deviations of zero inflated count data taking weights for structural zeros into account

Usage

## S4 method for signature 'Zi'
rowSds(x, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

## S4 method for signature 'Zi'
colSds(x, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

Arguments

x

A Zi-class object
rows, cols

A vector indicating the subset of rows and/or columns to operate over. If NULL (default), no subsetting is done
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE
useNames

logical If TRUE (default), names attributes of result are set. Else if FALSE, no naming support is done.

Value

a vector of row/column length

Examples

data(mtx)
Zi <- ziMain(mtx)
rowSds(Zi)
colSds(Zi)

Row and Column Variances of zero inflated count data

Description

Calculate row and column variances of zero inflated count data taking weights for structural zeros into account.

Usage

## S4 method for signature 'Zi'
rowVars(x, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

## S4 method for signature 'Zi'
colVars(x, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

Arguments

x

A Zi-class object
rows, cols

A vector indicating the subset of rows and/or columns to operate over. If NULL (default), no subsetting is done
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE
useNames

logical If TRUE (default), names attributes of result are set. Else if FALSE, no naming support is done.

Value

a vector of row/col length

Examples

data(mtx)
Zi <- ziMain(mtx)
rowVars(Zi)
colVars(Zi)

Row and Column weighted means of zero inflated count data

Description

Calculate row and column weighted means of zero inflated count data, additionally taking weights for structural zeros into account.

Usage

## S4 method for signature 'Zi'
rowWeightedMeans(
  x,
  w,
  rows = NULL,
  cols = NULL,
  na.rm = FALSE,
  useNames = TRUE
)

## S4 method for signature 'Zi'
colWeightedMeans(
  x,
  w,
  rows = NULL,
  cols = NULL,
  na.rm = FALSE,
  useNames = TRUE
)

Arguments

x

A Zi-class object
w

a numerical vector of weights either of length = rows or length = cols giving the weights to use for elements of x
rows, cols

A vector indicating the subset of rows and/or columns to operate over. If NULL (default), no subsetting is done
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE
useNames

logical If TRUE (default), names attributes of result are set. Else if FALSE, no naming support is done.

Value

a numeric vector of length N(K)

Examples

data(mtx)
Zi <- ziMain(mtx)
rowWeightedMeans(Zi, w = runif(ncol(inputcounts(Zi)), 0.1,1))
colWeightedMeans(Zi, w = runif(nrow(inputcounts(Zi)), 0.1,1))

Row and column weighted standard deviations or variances of zero inflated count data

Description

Calculate row and column standard deviations or variances of zero inflated count data, additionally taking weights for structural zeros into account.

Usage

## S4 method for signature 'Zi'
rowWeightedSds(x, w, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

## S4 method for signature 'Zi'
colWeightedSds(x, w, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

## S4 method for signature 'Zi'
rowWeightedVars(x, w, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

## S4 method for signature 'Zi'
colWeightedVars(x, w, rows = NULL, cols = NULL, na.rm = FALSE, useNames = TRUE)

Arguments

x

A Zi-class object
w

a numerical vector of weights either of length = rows or length = cols giving the weights to use for elements of x
rows, cols

A vector indicating the subset of rows and/or columns to operate over. If NULL (default), no subsetting is done
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE
useNames

logical If TRUE (default), names attributes of result are set. Else if FALSE, no naming support is done.

Value

a numeric vector of length N(K)

Examples

data(mtx)
Zi <- ziMain(mtx)
rowWeightedSds(Zi, w = runif(ncol(inputcounts(Zi)), 0.1,1))
colWeightedSds(Zi, w = runif(nrow(inputcounts(Zi)), 0.1,1))
rowWeightedVars(Zi, w = runif(ncol(inputcounts(Zi)), 0.1,1))
colWeightedVars(Zi, w = runif(nrow(inputcounts(Zi)), 0.1,1))

Access the sample data

Description

access the sample_data of an Zi-class object if the inputdata slot is a phyloseq object

Usage

## S4 method for signature 'Zi'
sample_data(object)

Arguments

object

Zi-class object

Value

sample_data

Examples

data(mtx)
OTU <- otu_table(mtx, taxa_are_rows = TRUE)
sample_data <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3',
    'Sample4', 'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9',
    'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
SAM <- sample_data(sample_data)
tax_table <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
TAX <- tax_table(tax_table)
ps <- phyloseq::phyloseq(OTU, TAX, SAM)
Zi <- ziMain(ps)
sample_data(Zi)

Standard Deviation of zero inflated count data

Description

Calculate the standard deviation of zero inflated count data taking weights for structural zeros into account.

Usage

## S4 method for signature 'Zi'
sd(x, na.rm = FALSE)

Arguments

x

A Zi-class object
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE

Value

standard deviation value

See Also

weightedSd, rowSds, colSds

Examples

data(mtx)
Zi <- ziMain(mtx)
sd(Zi)

Show summary of Zi object

Description

Message printed at command line

Usage

## S4 method for signature 'Zi'
show(object)

Arguments

object

Zi-class object

Value

returns a numeric vector of row/column length

Examples

data(mtx)
Zi <- ziMain(mtx)
Zi
show(Zi)

Subset a Zi-class object based on feature data

Description

Subset a Zi-class object based on tax_table of a phyloseq object or on rowData of a SummarizedExperiment object

Usage

subset_feature(Zi, ...)

Arguments

Zi

Zi-class object
...

The subsetting expression that should be applied, see subset for more details

Value

a Zi-class object after subsetting is done

Examples

data(mtx)
OTU <- otu_table(mtx, taxa_are_rows = TRUE)
sample_data <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3',
    'Sample4', 'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9',
    'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
SAM <- sample_data(sample_data)
tax_table <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
TAX <- tax_table(tax_table)
ps <- phyloseq::phyloseq(OTU, TAX, SAM)
Zi <- ziMain(ps)
subset_Zi_phylo <- subset_feature(Zi, ta2 == 'Bacteroidetes')

Subset a Zi-class object based on sample data

Description

Subset a Zi-class object based on sample_data of an phyloseq object or on colData based on a SummarizedExperiment object

Usage

subset_sample(Zi, ...)

Arguments

Zi

Zi-class object
...

The subsetting expression that should be applied, see subset for more details

Value

a Zi-class object after subsetting is done

Examples

data(mtx)
OTU <- otu_table(mtx, taxa_are_rows = TRUE)
sample_data <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3',
    'Sample4', 'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9',
    'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
SAM <- sample_data(sample_data)
tax_table <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
TAX <- tax_table(tax_table)
ps <- phyloseq::phyloseq(OTU, TAX, SAM)
Zi <- ziMain(ps)
subset_Zi <- subset_sample(Zi, SampleID %in% c('Sample1','Sample2'))

Transpose a Zi-class object

Description

transpose all matrizes of a Zi-class object

Usage

## S4 method for signature 'Zi'
t(x)

Arguments

x

Zi-class object

Value

Zi-class object

Examples

data(mtx)
Zi <- ziMain(mtx)
t(Zi)

Access the taxonomy table

Description

access the taxonomy table (tax_table) of an Zi-class object if the inputdata slot is a phyloseq object

Usage

## S4 method for signature 'Zi'
tax_table(object)

Arguments

object

Zi-class object

Value

tax_table

Examples

data(mtx)
OTU <- otu_table(mtx, taxa_are_rows = TRUE)
sample_data <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3',
    'Sample4', 'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9',
    'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
SAM <- sample_data(sample_data)
tax_table <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
TAX <- tax_table(tax_table)
ps <- phyloseq::phyloseq(OTU, TAX, SAM)
Zi <- ziMain(ps)
tax_table(Zi)

Variance of zero inflated count data

Description

Calculate the variance of zero inflated count data taking weights for structural zeros into account.

Usage

## S4 method for signature 'Zi,ANY'
var(x, na.rm = FALSE)

Arguments

x

A Zi-class object
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE

Value

variance value

See Also

weightedVar, rowVars, colVars

Examples

data(mtx)
Zi <- ziMain(mtx)
var(Zi)

Weighted Arithmetic Mean of zero inflated count data

Description

Calculate a weighted mean of zero inflated count data, additionally taking weights for structural zeros into account

Usage

## S4 method for signature 'Zi'
weighted.mean(x, w, ...)

Arguments

x

A Zi-class object
w

a numerical vector of weight the same length as x giving the weights to use for elements of x
...

weighted.mean

Value

weighted mean value

See Also

weighted.mean, rowWeightedMeans, colWeightedMeans

Examples

data(mtx)
Zi <- ziMain(mtx)
weight <- runif(length(inputcounts(Zi)), 0.1, 1)
weighted.mean(Zi, w= weight)

Weighted Variance and weighted Standard Deviation

Description

Calculate a weighted variance and standard deviation of zero inflated count data, additionally taking weights for structural zeros into account

Usage

weightedSd(x, w = NULL, idxs = NULL, na.rm = FALSE, center = NULL, ...)

weightedVar(x, w = NULL, idxs = NULL, na.rm = FALSE, center = NULL, ...)

Arguments

x

A Zi-class object
w

a numerical vector of weight the same length as x giving the weights to use for elements of x
idxs

A vector indicating subset of elements to operate over. If NULL, no subsetting is done.
na.rm

logical If TRUE NAs are excluded, otherwise not. default = FALSE
center

numeric scalar specifying the center location of the data. If NULL, it is estimated from data.
...

weightedVar

Value

a numeric scalar

See Also

weightedVar, rowWeightedVars, colWeightedVars

Examples

data(mtx)
Zi <- ziMain(mtx)
weight <- runif(length(inputcounts(Zi)), 0.1, 1)
weightedVar(Zi, w= weight)
weightedSd(Zi, w = weight)

Access the weights

Description

access the weights of an Zi-class object

Usage

weights(x)

## S4 method for signature 'Zi'
weights(x)

weights(x) <- value

## S4 replacement method for signature 'Zi'
weights(x) <- value

Arguments

x

Zi-class object
value

weights object

Value

weights

Examples

data(mtx)
Zi <- ziMain(mtx)
weights(Zi)

Class Zi

Description

Objects of this class store all the results of the ZiMain function to continue zero inflated data analysis

Value

Zi-class object

Slots

inputdata

a matrix, phyloseq or SummarizedExperiment object.

inputcounts

matrix. The count matrix, features as rows, samples as columns

model

list. The result of fitting a zero inflated model using zeroinfl

deinflatedcounts

matrix. The matrix where predicted structural zeros are omitted and stored as NA values

weights

matrix. A matrix containing weights for zero counts

Convert a Zi-class object to a DESeq2 dds object

Description

A Zi-class object is converted to a DESeqDataSet object, which can be used for DESeq2 analysis. Both, weight and count matrices will be stored in assays of the DESeqDataSet.

Usage

zi2deseq2(ZiObject, design, colData, ...)

Arguments

ZiObject

Zi-class object
design

A formula which specifies the design of the experiment, taking the form formula(~ x + y + z). That is, a formula with right-hand side only. By default, the functions in this package and DESeq2 will use the last variable in the formula (e.g. z) for presenting results (fold changes, etc.) and plotting. When considering your specification of experimental design, you will want to re-order the levels so that the NULL set is first.
colData

if the inputdata of the Zi-class object is a matrix: a DataFrame or data.frame with at least a single column. Rows of colData correspond to columns of countData
...

phyloseq::phyloseq_to_deseq2 if the inputdata of the 'Zi'-object is a phyloseq object DESeq2::DESeqDataSet if the inputdata the ' Zi'-object is a SummarizedExperiment object

Value

a dds class object

Examples

data(mtx)
Zi <- ziMain(mtx)
colData <- data.frame(group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
zi2deseq2(Zi, ~group, colData)

Replace the otu table of a phyloseq object

Description

Replace the OTU table of a phyloseq object with the OTU table of zero de-inflated count data

Usage

zi2phyloseq(ZiObject)

Arguments

ZiObject

Zi-class object with a phyloseq object as input

Value

a 'phyloseq'-class object

Examples

data(mtx)
OTU <- otu_table(mtx, taxa_are_rows = TRUE)
sample_data <- data.frame(SampleID = c('Sample1', 'Sample2', 'Sample3',
    'Sample4', 'Sample5', 'Sample6', 'Sample7', 'Sample8', 'Sample9',
    'Sample10'),
    Group = factor(x = c(1,1,1,1,1,2,2,2,2,2)))
SAM <- sample_data(sample_data)
tax_table <- data.frame(Kingdom = c(rep('Bacteria', times = 100)),
    Phylum = c(rep('Bacteroidetes', times = 50),
    rep('Firmicutes', times = 50)))
TAX <- tax_table(tax_table)
ps <- phyloseq::phyloseq(OTU, TAX, SAM)
Zi <- ziMain(ps)
new_ps <- zi2phyloseq(Zi)
new_ps

ziMain - main function to fit a zero inflation model and calculate weights for structural zeros

Description

This function fits a zero-inflated mixture model (either Poisson or negative binomial distribution) to count data and calculates weights for all zeros indicating whether a zero is a real count (weight close to 1) or whether it is a structural zero (weight close to 0). The default model is a zero inflated negative binomial model.

The input inputdata of the ziMain function is either a phyloseq object, SummarizedExperiment object or count matrix.

In order to reduce calculation times, the count matrix is divided into blocks of around 5000 count values. Then, a zero inflation model (either Poisson or negative binomial distribution) is fitted to the data. The response variable count is estimated using the predictor variables sample(columns) and feature(rows). Using the fitted zero inflated model, probabilities given that a zero in the count matrix is a structural zero are predicted. Those probabilities are used in two ways: 1) A zero-deinflated count matrix is generated where a appropriate proportion of zeros are randomly replaced by NA. This count matrix can be used for analysis methods which cannot deal with weights. 2) Weights

w=(1−π)fNB(y;μ,θ)fZINB(y;μ,θ,π).w = \frac{\left(1 - \pi\right) f_{\text{NB}}\left(y; \mu, \theta \right)}{f_{\text{ZINB}}\left(y;\mu, \theta, \pi\right)}.

(see Van den Berge, K., Perraudeau, F., Soneson, C. et al.) are calculated in order to down-weight structrual zeros in analyses which can account for weighting of individual data points.

all zero counts are calculated given the following formula:

The result of the ziMain function can be used to analyze zero inflated count data.

Usage

ziMain(
  inputdata,
  feature = "feature",
  formula = count ~ sample + feature,
  dist = "negbin",
  link = "logit",
  zeroRows.rm = FALSE,
  ...
)

Arguments

inputdata

phyloseq object, SummarizedExperiment object, or matrix (rows =features, columns=samples)
feature

'feature', 'gene', 'OTU', 'phylum', etc. By default, rownames are labelled as feature1, feature2, ...
formula

formula to fit the zero inflated model y ~ x1 + x2 + ..., default = count ~ sample + feature. A different set of regressors can be specified using y ~ x1 +x2 + ...|z1 + z2

  • ... where the first part describes the count data model and the second part describes the zero inflation model

dist

= distribution, either poisson ('poisson'), negative binomial ('negbin')
link

= link function, either 'logit', 'probit', 'cloglog', 'cauchit'
zeroRows.rm

= logical, indicating whether rows that only contain zeros should be removed (TRUE) or not (FALSE) (they are removed to fit a zero inflated model and will be added afterwards count matrix per default = 0 and weights = 1)
...

additional parameters to describe the model, see zeroinfl

Value

Zi-class object

Slots

inputdata

a matrix, phyloseq or SummarizedExperiment object.

inputcounts

matrix. The count matrix, features as rows, samples as columns

model

list. The result of fitting a zero inflated model using zeroinfl

deinflatedcounts

matrix. A matrix where zero counts are randomly replaced according to the estimated probability of being a structural zero

weights

matrix. A matrix containing weights for zero counts

References

Van den Berge, K., Perraudeau, F., Soneson, C. et al. Observation weights unlock bulk RNA-seq tools for zero inflation and single-cell applications. Genome Biol 19, 24 (2018). https://doi.org/10.1186/s13059-018-1406-4

See Also

zeroinfl

Examples

# zero-inflated count matrix
data(mtx)
# calling ziMain function:
Zi <- ziMain(mtx)
#Example Data Sets from other R packages
#data(enterotype)
#data(GlobalPatterns)
#data(esophagus)
#ziMain(esophagus)
#data(soilrep)