A package for drift across batches normalization and visualization
dbnorm contains R functions which allow visualization and
removal of technical heterogeneity from large metabolomics dataset. By including advanced statistical tools, the dbnorm package allows user to inspect the structure and quality of large metabolomics datasets both in macroscopic and microscopic scales at the sample batch level and metabolic feature level, respectively.
It allows users to efficiently correct drift across batch and to adjust large metabolomics datasets for technical variation which helps improving the estimation of the biological mechanisms underlying disease condition or medical state.
dbnorm includes 11 distinct functions for pre-processing of data and estimation of missing values,
conventional functions for batch effect correction based on statistical models, as well as functions using advanced statistical
tools to generate several diagnosis plots to help users to choose the statistical model which better fits to their data
structure. dbnorm includes several statistical models such as, ComBat(parametric and non-parametric)-model [PMID:16632515] from sva package [PMID:22257669] ,that was already in use for metabolomics data normalization, and ber function [DOI:10.1007/s12561-013-9081-1], priorly developed for microarray gene expression data, that we propose here as a new approach for correction of drift across batch in metabolomics datasets.
https://www.nature.com/articles/s41598-021-84824-3
Install package dependencies in CRAN and Bioconductor via codes bellow:
R installer
install.packages(c("ggplot2","parallel","reshape2","plyr",
"knitr","tibble","installr","fs","rmarkdown","processx","backports",
"bootstrap","boot","caret","dplyr","stringr","ggfortify","factoextra","MASS",
"RColorBrewer","RCurl","lattice","data.table","igraph","tidyr","scales",
"e1071","fpc","rlang","glue","digest"))
Call all those package by library:
e.g. library (MASS)
Bioconductor installer
if (!requireNamespace("BiocManager", quietly = TRUE))
install.packages("BiocManager")
BiocManager::install(c("pcaMethods","limma","impute","sva","BiocParallel","genefilter","Biobase","mixOmics","statTarget", "multtest"))
Call all those package by library:
e.g. library (sva)
R install package from archive
require(usethis)
require(devtools)
URL: https://cran.r-project.org/src/contrib/Archive/
devtools::install_version("ber", version = "4.0", repos = "http://cran.us.r-project.org")
devtools::install_version("NormalizeMets", version = "0.25", repos = "http://cran.us.r-project.org")
devtools::install_version("metabolomics", version = "0.1.4", repos = "http://cran.us.r-project.org")
library(ber)
library(NormalizeMets)
library(Rcpp)
library(metabolomics)
dbnorm is freely available from GitHub. The package documentation, including user manual is available within the downloaded R package file. If all package dependencies were installed, you will be able to install the dbnorm. Users can either manually download the tar.gz file or clon the GitHub.
cd ~/Downloads
R CMD INSTALL dbnorm tar.gz
Rstudio users
First, download the tar.gz file from GitHun then in Rstudio from Tools icon, select Install.packages and
then Browse to the tar.gz file.
git clone
git@github.com:NBDZ/dbnorm.git
R CMD build Mdbnorm
R CMD INSTALL dbnorm tar.gz
Rstudio users
Download the tar.gz file from GitHun then in Rstudio from Tools icon, select Install.packages and
then Browse to the tar.gz file.
In this section, we briefly introduce and explain the functions implemented in the dbnorm with the expected outcome. Data to be uploaded must be normalized and scaled on the log2 to account for the high abundance features (variables) by which technical heterogeneity might be overlooked. The input data must be in .csv format with the independent experiments in the rows and the features (variables) in the columns, with the batch levels considered in the first column.
- Example:
data<-read.csv(" path/to/directory/folde/mydata.csv",sep = ",", header = T, row.names = 1)
library(dbnorm)
- emvd
This function allows you to estimate missing values (Zero or/and NA values) by the lowest detected value in the entire experiment.
df<-data[-1] #keep data matrix by removing batch level in the first column
>emvd [df]
f<- emvd[df] # save the imputed data matrix
- emvf
This function allows you to estimate missing values (Zero or/and NA values) for each feature (variable) by the lowest value detected for the corresponding feature (variable), applied on the column.
df<-data[-1] #keep data matrix by removing batch level in the first column
> emvf [df]
f <- emvf[df] # save the imputed data matrix
- Visodbnorm ; Visualization of drift across batch normalization
This function performs batch effect adjustment via three statistical models implemented in the dbnorm, namely two-stage procedure as described by Giordan (2013)[DOI:10.1007/s12561-013-9081-1] and/or empirical Bayes methods in two setting of parametric and non-parametric as described by Johnson et al.(2007) [PMID: 16632515] and in sva package by Leek et al.(2012)[PMID: 22257669]. Meanwhile, the graphical inferences in the context of unsupervised learning algorithms create visual inspection to inform users about the spatial separation of the sample sets analyzed in the different analytical runs alongside the distribution of features (variables) in the raw and treated datasets. This function is suggested for less than 2000 features (variables).
- Value
Graphical check such as PCA plot and Scree plot compiled into a PDF (saved in the working directory) and three .csv files (saved in a folder, intiate with Rtmpe, in Users's Temporary directory: "C:\Users\ “%USERNAME”\AppData\Local\Tem") for adjusted data based on the models implemented in the package.The RLA plots are visualized in the Viewer panel in the rstudio console.
Visodbnorm(data)
- dbnormSCORE ; Adjusted coefficient of determination for a data normalized for across batch signal drift
This function gives a quick notification about the performance of the statistical models, two-stage procedure [DOI:10.1007/s12561-013-9081-1] and/or empirical Bayes methods in two setting of parametric and non-parametric as described in [PMID: 16632515] and by sva package [PMID:22257669], implemented in the dbnorm package, in accommodating technical variability. Subsequently, the adjusted coefficient of determination or Adjusted R-Squared is calculated for each variable estimated in a regression model for its dependency to the batch level in the raw data and treated data via either of those models. Immediately, the performance of applied models are presented by a score calculated based on the maximum variability explained by the batch level, notify the consistency of model performance for all detected features (variables), facilitating quick comparison of the models for selecting one of those models, which is more appropriate to the data structure. This function is suggested for less than 2000 features (variables) for better computational speed.
- Value
Graphical check such as Correlation plot and Score plot compiled into a PDF file (saved in the working directory) and .csv files (saved in a folder, intiate with Rtmpe, in Users's Temporary directory: "C:\Users\ “%USERNAME”\AppData\Local\Tem") in the two vector data matrix of Adjusted R-squared for each model and a Table of score for the maximum Adjusted R-squared detected for the applied models.
dbnormSCORE (data)
- ProfPlotraw
- ProfPlotBer
- ProfPlotBagging
- ProfPlotComPara
- ProfPlotComPara
- ProfPlotComNPara; Visualization of analytical heterogeneity on the profile of features (variables) in raw data and after correction via ber-, ber-bagging, parametric ComBat and non-parametric ComBat
These functions allow users to adjust the data for batch effect using either of models implemented in the package described earlier, and inform about the presence of across batch signal drift or batch effect in the raw and treated data represented via the shifted probability density function (PDF) plots of the features (variables) detected in an experiment.
- Value
Graphical check such as the plots compiled into a PDF file (saved in the working directory) and a .csv file (saved in a folder, intiate with Rtmpe, in Users's Temporary directory: "C:\Users\ “%USERNAME”\AppData\Local\Tem") of corrected dataset via either of applied function.
profplotraw (data)
ProfPlotBer (data)
ProfPlotBagging (data)
ProfPlotComPara (data)
ProfPlotComPara (data)
ProfPlotComNPara (data)
- dbnormBer
- dbnormBagging
- dbnormPcom
- dbnormNPcom; Data normalization for across batches signal drift using either of ber-,ber-bagging, parametric ComBat- and non parametric ComBat- models and unsupervised clustering and regression analysis of corrected data
To increase computational processing of big data, in these functions, statistical models and graphical checks implemented in “Visodborm” decomposed in to several separated functions each of these performing a unique batch effect correction with respective result and graphical checks.
- Value
Graphical check such as PCA plot, Scree plot and Correlation plot compiled into a PDF (saved in the working directory) and the .csv (saved in a folder, intiate with Rtmpe, in Users's Temporary directory: "C:\Users\ “%USERNAME”\AppData\Local\Tem") for corrected dataset based on either of applied model and the two column matrix of Adjusted-R squared. The RLA plots are visualized in the Viewer panel in the rstudio console.
dbnormBer(data)
dbnormBagging (data)
dbnormPcom(data)
dbnormNPcom(data)
- hclustdbnorm; Hierarchical clustering analysis of original data and corrected data for batch effect
This function allows users to evaluate dissimilarity between identical samples (quality control replicates or analytical replicates) analyzed in different batches, prior and after correction using, ber, ber_bagging and parametric and non-parametric ComBat . Pearson distance and average method for clustering were considered.Bagging model is performed using partial bagging with n=150 bootstrap samples
hclustdbnorm (data)
Distributed under the GPL license. See LICENSE for details.