Title: Feature-Wise Normalization for Microbiome Sequencing Data
Version: 2.4
Maintainer: Dillon Lloyd <dtlloyd@ncsu.edu>
Description: A novel feature-wise normalization method based on a zero-inflated negative binomial model. This method assumes that the effects of sequencing depth vary for each taxon on their mean and also incorporates a rational link of zero probability and taxon dispersion as a function of sequencing depth. Ziyue Wang, Dillon Lloyd, Shanshan Zhao, Alison Motsinger-Reif (2023) <doi:10.1101/2023.10.31.563648>.
License: GPL-3
Depends: R (≥ 4.0.0), microbiome,
Imports: phyloseq, stats, S4Vectors, BiocGenerics, vegan, methods, MASS, future, future.apply, matrixStats, pscl, parallelly, ggplot2, utils
URL: https://github.com/wangziyue57/TaxaNorm
biocViews: Sequencing, Microbiome, Metagenomics, Normalization, Visualization
Suggests: rmarkdown, knitr
VignetteBuilder: knitr
Encoding: UTF-8
RoxygenNote: 7.2.3
BugReports: https://github.com/wangziyue57/TaxaNorm/issues
NeedsCompilation: no
Packaged: 2023-12-11 15:19:47 UTC; Dillon
Author: Ziyue Wang [aut], Dillon Lloyd [aut, cre, cph], Shanshan Zhao [aut, ctb], Alison Motsinger-Reif [aut, ctb]
Repository: CRAN
Date/Publication: 2023-12-12 18:30:05 UTC

TaxaNorm: Feature-Wise Normalization for Microbiome Sequencing Data

Description

A novel feature-wise normalization method based on a zero-inflated negative binomial model. This method assumes that the effects of sequencing depth vary for each taxon on their mean and also incorporates a rational link of zero probability and taxon dispersion as a function of sequencing depth. Ziyue Wang, Dillon Lloyd, Shanshan Zhao, Alison Motsinger-Reif (2023) <arXiv:https://www.biorxiv.org/content/10.1101/2023.10.31.563648v1.abstract>.

Author(s)

Maintainer: Dillon Lloyd dtlloyd@ncsu.edu [copyright holder]

Authors:

See Also

Useful links:

TaxaNorm data objects

Description

Objects included in the TaxaNorm package, loaded with utils::data

Usage

data(TaxaNorm_Example_Input, package = "TaxaNorm")

data(TaxaNorm_Example_Output, package = "TaxaNorm")

TaxaNorm_Example_Input

Example data #'

TaxaNorm_Example_Output

Example output

Examples

data(TaxaNorm_Example_Input, package = "TaxaNorm")
          data(TaxaNorm_Example_Output, package = "TaxaNorm")

TaxaNorm package generics

Description

TaxaNorm package generics; see class man pages for associated methods

Usage

input_data(x, ...)

input_data(x, ...) <- value

rawdata(x, ...)

rawdata(x, ...) <- value

normdata(x, ...)

normdata(x, ...) <- value

ecdf(x, ...)

ecdf(x, ...) <- value

model_pars(x, ...)

model_pars(x, ...) <- value

converge(x, ...)

converge(x, ...) <- value

llk(x, ...)

llk(x, ...) <- value

final_df(x, ...)

final_df(x, ...) <- value

coefficients(x, ...)

coefficients(x, ...) <- value

mu(x, ...)

mu(x, ...) <- value

theta(x, ...)

theta(x, ...) <- value

pi(x, ...)

pi(x, ...) <- value

Arguments

x

TaxaNorm S4 object

...

Included for extendability; not currently used

value

Replacement value

Value

TaxaNorm generic functions return the specified slot of the TaxaNorm object given to the function

TaxaNorm_Model_Parameters

Description

S4 class to store TaxaNorm Parameters

Usage

TaxaNorm_Model_Parameters(coefficients, mu, theta, pi)

## S4 method for signature 'TaxaNorm_Model_Parameters'
coefficients(x)

## S4 replacement method for signature 'TaxaNorm_Model_Parameters'
coefficients(x) <- value

## S4 method for signature 'TaxaNorm_Model_Parameters'
mu(x)

## S4 replacement method for signature 'TaxaNorm_Model_Parameters'
mu(x) <- value

## S4 method for signature 'TaxaNorm_Model_Parameters'
theta(x)

## S4 replacement method for signature 'TaxaNorm_Model_Parameters'
theta(x) <- value

## S4 method for signature 'TaxaNorm_Model_Parameters'
pi(x)

## S4 replacement method for signature 'TaxaNorm_Model_Parameters'
pi(x) <- value

Arguments

coefficients

Passed to coefficients slot

mu

Passed to mu slot

theta

Passed to theta slot

pi

Passed to pi slot

x

TaxaNorm_Model_Parameters object

value

Replacement value

Details

Parameters for TaxaNorm Method

Functions

Slots

coefficients

matrix coefficients

mu

matrix mu

theta

matrix theta

pi

matrix pi

Examples

coefficients <-  matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
mu <- matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
theta <-  matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
pi <- matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
TaxaNorm_Model_Parameters(coefficients = coefficients,mu = mu,theta = theta,pi = pi)

Function to QC TaxNorm algorithm

Description

Function to QC TaxNorm algorithm

Usage

TaxaNorm_Model_QC(TaxaNormResults)

Arguments

TaxaNormResults

Input data; Results from TaxaNorm normalization

Value

a list containing qc taxnorm object

Examples

data("TaxaNorm_Example_Output", package = "TaxaNorm")
TaxaNorm_Model_QC(TaxaNormResults = TaxaNorm_Example_Output)

Function for TaxNorm NMDS

Description

Function for TaxNorm NMDS

Usage

TaxaNorm_NMDS(TaxaNormResults, group_column)

Arguments

TaxaNormResults

(Required) Input data; should be either a phyloseq object or a count matrix

group_column

column to cluster on

Value

NMDS Plot

Examples

data("TaxaNorm_Example_Output", package = "TaxaNorm")
TaxaNorm_NMDS(TaxaNorm_Example_Output, group_column = "body_site")

Function to run TaxaNorm algorithm

Description

Function to run TaxaNorm algorithm

Usage

TaxaNorm_Normalization(
  data,
  depth = NULL,
  group = NULL,
  meta.data = NULL,
  filter.cell.num = 10,
  filter.taxa.count = 0,
  random = FALSE,
  ncores = NULL
)

Arguments

data

(Required) Input data; should be either a phyloseq object or a count matrix

depth

sequencing depth if pre-calculated. It should be a vector with the same length and order as the column of the count data

group

condition variables if samples are from multiple groups; should be correpsond to the column of the count data. default is NULL, where no grouping is considered

meta.data

meta data for Taxa

filter.cell.num

taxa with "filter.cell.num" in more than the value provided will be filtered

filter.taxa.count

"filter.taxa.count" samples will be removed before testing. default is keep taxa appear in at least 10 samples within each group

random

calculate randomized normal quantile residual

ncores

whether multiple cores is used for parallel computing; default is max(1, detectCores() - 1)

Value

a TaxaNorm Object containing the normalized count values and accessory information

Examples


data("TaxaNorm_Example_Input", package = "TaxaNorm")
Normalized_Data <- TaxaNorm_Normalization(data= TaxaNorm_Example_Input,
                                         depth = NULL,
                                         group = sample_data(TaxaNorm_Example_Input)$body_site,
                                         meta.data = NULL,
                                         filter.cell.num = 10,
                                         filter.taxa.count = 0,
                                         random = FALSE,
                                         ncores = 1)

Function for TaxNorm input data

Description

Function for TaxNorm input data

Usage

TaxaNorm_QC_Input(data)

Arguments

data

(Required) Input data; should be either a phyloseq object or a count matrix

Value

QC PLots

Examples

data("TaxaNorm_Example_Input", package = "TaxaNorm")
qc_data <- TaxaNorm_QC_Input(TaxaNorm_Example_Input)

TaxaNorm Results

Description

S4 class to store TaxaNorm Results

Usage

TaxaNorm_Results(
  input_data,
  rawdata,
  normdata,
  ecdf,
  model_pars,
  converge,
  llk,
  final_df
)

## S4 method for signature 'TaxaNorm_Results'
input_data(x)

## S4 replacement method for signature 'TaxaNorm_Results'
input_data(x) <- value

## S4 method for signature 'TaxaNorm_Results'
rawdata(x)

## S4 replacement method for signature 'TaxaNorm_Results'
rawdata(x) <- value

## S4 method for signature 'TaxaNorm_Results'
normdata(x)

## S4 replacement method for signature 'TaxaNorm_Results'
normdata(x) <- value

## S4 method for signature 'TaxaNorm_Results'
ecdf(x)

## S4 replacement method for signature 'TaxaNorm_Results'
ecdf(x) <- value

## S4 method for signature 'TaxaNorm_Results'
model_pars(x)

## S4 replacement method for signature 'TaxaNorm_Results'
model_pars(x) <- value

## S4 method for signature 'TaxaNorm_Results'
converge(x)

## S4 replacement method for signature 'TaxaNorm_Results'
converge(x) <- value

## S4 method for signature 'TaxaNorm_Results'
llk(x)

## S4 replacement method for signature 'TaxaNorm_Results'
llk(x) <- value

## S4 method for signature 'TaxaNorm_Results'
final_df(x)

## S4 replacement method for signature 'TaxaNorm_Results'
final_df(x) <- value

Arguments

input_data

passed to input_data slot

rawdata

Passed to rawdata slot

normdata

Passed to normdata slot

ecdf

Passed to ecdf slot

model_pars

Passed to model_pars slot

converge

Passed to converge slot

llk

Passed to llk slot

final_df

Passed to final_df slot

x

TaxaNorm_Results object

value

Replacement value

Details

All results from the TaxaNorm method and what was used to get those results

Functions

Slots

input_data

ANY phyloseq input data

rawdata

data.frame Data frame of counts to use

normdata

data.frame Normalized Data

ecdf

data.frame ecdf

model_pars

TaxaNorm_Model_Parameters list of model parameters

converge

⁠vector(<logical>)⁠ converge

llk

ANY llk

final_df

ANY final_df

Examples

coefficients <-  matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
mu <- matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
theta <-  matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
pi <- matrix(c(1,2,3,4,5,6,7,8,9),nrow=3,ncol=3,byrow=TRUE)
model_pars <- TaxaNorm_Model_Parameters(coefficients = coefficients,mu = mu,theta = theta,pi = pi)
data("TaxaNorm_Example_Input", package = "TaxaNorm")
rawdata <- data.frame(Taxa1 = c(1,2,3),Taxa2 = c(3,4,5),Taxa3 = c(6,7,8))
normdata <- data.frame(Taxa1 = c(-1.4,-1.09,-0.73),
Taxa2 = c( -0.36,0,0.36), Taxa3 = c(0.73,1.09,1.46))
ecdf <- data.frame(0.05,0.23,0.89)
converge <- c(TRUE,TRUE,FALSE)
llk <- c(1,1.5,0.5)
final_df <- data.frame(Taxa1 = c(1,2,3),Taxa2 = c(3,4,5),Taxa3 = c(6,7,8))
TaxaNorm_Results(input_data = TaxaNorm_Example_Input,
                                  rawdata = rawdata,
                                   normdata = normdata,
                                   ecdf = ecdf,
                                   model_pars = model_pars,
                                   converge = converge,
                                   llk = llk,
                                   final_df = final_df)

Function to run TaxNorm algorithm

Description

Function to run TaxNorm algorithm

Usage

TaxaNorm_Run_Diagnose(Normalized_Results, prev = TRUE, equiv = TRUE, group)

Arguments

Normalized_Results

(Required) Input results from from run_norm()

prev

run prev test

equiv

run equiv test

group

group used for taxanorm normalization

Value

a list containing the normalized count values

Examples

data("TaxaNorm_Example_Input", package = "TaxaNorm")
data("TaxaNorm_Example_Output", package = "TaxaNorm")
Diagnose_Data <- TaxaNorm_Run_Diagnose(Normalized_Results = TaxaNorm_Example_Output,
                                        prev = TRUE,
                                        equiv = TRUE,
                                        group = sample_data(TaxaNorm_Example_Input)$body_site)