README.md

In donaldtmcknight/microDecon: A package for removing contaminant sequencing reads from metabarcoding studies

microDecon

An R package for removing contamination from metabarcoding (e.g., microbiome) datasets post-sequencing

Description

Contamination is a serious and widespread problem in metabarcoding studies (particularly bacterial microbiome studies). microDecon provides an easy method of identifying and removing contaminant reads post-sequencing. It uses the data in one or more blank samples that were carried throughout the entire project (collection, extraction, amplification, and sequencing) to identify and remove the reads in each actual sample that are from contamination. Because it removes reads rather than entire OTUs, it is a large improvement over existing methods. It was designed specifically for microbiome sequencing studies (e.g., 16S bacterial studies and ITS fungal studies), but it should be applicatble to metabarcoding studies more generally. A more detailed explanation of the algorithms it uses can be found in the User's Guide.

Installation

Simply copy, paste, and run the following code in R.

install.packages("devtools") #Installs devtools (if not already installed)
devtools::install_github("donaldtmcknight/microDecon") #Installs microDecon

Running the package

Load microDecon

library(microDecon)

Load a data frame with a column of OTU IDs, at least one column from a blank sample, at least one column from actual samples, and (optional) a column with taxonomic information. For example, the following OTU data frame has three blanks, three samples (from two populations), and a taxa column (the columns must be in that order: OTU IDs, blanks, samples [samples should be grouped by population, species, etc.], taxa).

example <- cbind.data.frame(c("OTU1","OTU2","OTU3","OTU4","OTU5","OTU6"),
                        c(0,200,1000,50,0,25),
                        c(0,220,800,30,0,10),
                        c(0,180,1300,70,0,30),
                        c(60,660,1440,70,2400,30),
                        c(64,520,1000,48,1900,20),
                        c(40,480,700,35,2100,15),
                        c("K_Bacteria; P_Actinobacteria","K_Bacteria; P_Proteobacteria","K_Bacteria; P_Proteobacteria","K_Bacteria; P_Bacteroidetes","K_Bacteria","K_Bacteria"))
colnames(example) <- c("OTU_ID","Blank1","Blank2","Blank3","Pop1_Sample1","Pop1_Sample2","Pop2_Sample3","Taxa")

The decon() function is the primary function of the package (it is a warpper for all other functions). To decontaminate these samples on default settings, use the following code (data is your data frame, numb.blanks specifies the number of blanks in your data frame, numb.ind specifies the number of individuals per group (e.g., population), taxa specifies whether your data frame includes a taxa column).

decontaminated <- decon(data = example,numb.blanks=3,numb.ind=c(2,1),taxa=T)

For details on additional options and when you should use settings other than the defaults, see the User's Guide and help files in R.

donaldtmcknight/microDecon documentation built on Oct. 23, 2023, 10:57 a.m.