R For Plant Breeders: Introduction to R For Plant Breeding

Installing & Setting Up R

To use R, you should download R and RStudio.

I have some resources linked on my resources page that will help you get R and RStudio set up. You will also find some resources on getting yourself familiar with R Studio’s Interface.

When you are ready to start working on your data, you should set your working directory to a specific folder. It should be where you have your data saved. Setting your working directory to this folder will also allow you to easily save any output files (like plots) to the same folder.

#set working directory by adding file path like below 
#setwd("C:/Users/eruph/Documents/Technical-Blog_EG/Technical-Blog/_posts/2022-03-18-introduction-to-r-for-plant-breeding")

#check working directory
getwd()

[1] "C:/Users/eruph/Documents/Technical Blog_EG/Technical-Blog/_posts/2022-03-18-introduction-to-r-for-plant-breeding"

All code lines that start with a hash/number(#) sign are comment lines. Comments help you and others who read your code understand what that code line or section is doing. If you put the # in front of a code line, that code will not run. Putting # in front of a code line is helpful when testing what different lines of code do.

Packages

Packages in R Studio are how you will access the functions you want to use/run. There are two main things you have to do when you want to use an R package:

Check if the package is installed; if not, install it
Once installed, load the package

There are two ways to install a package in R. The first way is by going to the ‘Packages’ tab on the bottom right window. The second way is to do it using code. Below is an example of installing and loading tidyverse, a collection of R packages.

#installing the tidyverse package
#install.packages("tidyverse")
# Loading tidverse for the session
#library(tidyverse)

Some of the packages included in core tidyverse are:

ggplot2 : package useful for creating graphics
dplyr : good for data manipulation
tidyr : cleaning data
readr : reading rectangular data (each column is a variable and row is a record)

Learn more about the tidyverse here .

Plant Breeding Packages you should consider installing

metan for multi-environmental data
statgenGxE
statgenSTA
lme4 for multi-level models
gapit for Genome Wide Association Study (GWAS) and Genomic Selection (GS)

Programming with R

Basic Arithmetic

In R, you can run mathematical operations & functions, logarithms, exponentials, etc. Below are some examples. When you run each line of code, the answer will be shown in the console window (bottom left window) of RStudio’s interface.

12 + 2 #addition

[1] 14

12 - 2 #subtraction

[1] 10

12 * 2 #multiplication

[1] 24

12 / 2 #division

[1] 6

12 ^ 2 #exponentiation

[1] 144

log2(16) #log

[1] 4

sqrt(16) #square root

[1] 4

Assigning variables

In R, you will often need to store values or an object because you want to use it in another function. This is called assigning. Assigning a value or object to a variable will save it in your environment to be used later.

#Assigning value 4 to x
x <- 4
#Assigning value 6 to y
y <- 6
#Assigning z and using x & y
z <- x + y

#Assigning an example dataset from datasets base r to a variable 
plant_growth <- datasets::PlantGrowth

Data Types and Structures

Basic Data Types in R

character: “text”
numeric: 12, 4.5
integer: 2, 3
logical: TRUE, FALSE

Use the function class() to check the data type of a saved object, column, or columns in the entire dataset.

#calling first 6 rows of dataset loaded earlier
head(plant_growth)

  weight group
1   4.17  ctrl
2   5.58  ctrl
3   5.18  ctrl
4   6.11  ctrl
5   4.50  ctrl
6   4.61  ctrl

#class of dataset
class(plant_growth)

[1] "data.frame"

#class of specific column in dataset called group 
class(plant_growth$group)

[1] "factor"

Data Structures

Vectors are one-dimensional arrays that can only store one data type.
Matrices are two-dimensional tables.
Factors are used to store categorical data like eye color or plant variants/lines.
Data frames are data stored in table format where columns are variables and rows are records.
Lists: can hold different data types

Use the function str() to check the data structure of an object.

#vector 
alphabet <- c("a", "b", "c", "d", "e", "f", "g")
str(alphabet)

 chr [1:7] "a" "b" "c" "d" "e" "f" "g"

#matrix
matrix_1 <- matrix(c(5,6,7,8,9,10), nrow = 2, ncol = 3)
str(matrix_1)

 num [1:2, 1:3] 5 6 7 8 9 10

#factor
primary_colors <- factor( c("red", "yellow", "blue"))
str(primary_colors)

 Factor w/ 3 levels "blue","red","yellow": 2 3 1

#data.frame
data_frame_1 <- data.frame (
  Variety = c("SCN 11", "Embean 11", "Jesca"),
  Yield = c(1200, 1550, 1320),
  Seed_weight = c(60, 30, 45)
)
str(data_frame_1)

'data.frame':   3 obs. of  3 variables:
 $ Variety    : chr  "SCN 11" "Embean 11" "Jesca"
 $ Yield      : num  1200 1550 1320
 $ Seed_weight: num  60 30 45

#list
environments <- list("Arusha", "Lyamungu", "Uyole")
str(environments)

List of 3
 $ : chr "Arusha"
 $ : chr "Lyamungu"
 $ : chr "Uyole"

R is a great tool that more plant breeders can use to conduct their analysis. R can be used to explore your data before performing more specific analysis. You can easily get descriptive statistics or make exploratory plots to show data distribute or variance. You can also conduct Analysis of Variance (ANOVA), Principal Component Analysis (PCA), and create AMMI Biplots & GGE Biplots.