3.15 Week 4 Assignment

3.15.1 Visualizing Simulations

So far in class and for homework we’ve simulated a variety of evolutionary dynamics and visualized their results with built-in plots in SLiM. In some cases, the things you’ll simulate will be best visualized by custom plots, which we’ll start writing in R. This introduction to R will be useful as you move into futher coding projects as the fundamentals of plotting will remain the same.

3.15.2 R basics

A few coding terms:

function: code that is used to execute something (e.g., the print() function is the same in R as it is in SLiM)
package: a set of functions that already exist and can be loaded into your workspace
comment: in R, you indicate a comment with the pound (or hashtag) #
script: a file (typically with the suffix .R) that contains your R code

3.15.3 Reading data in to R

To start your plot, you’ll need to have your data in your R workspace. Recall that you wrote the output of your allele frequency simulations to file; you created 3 different files, each exploring a unqiue set of parameters, so you could compare the dynamics resulting from changing those parameters (e.g., population size, selection coefficient). You can read those files (or any other file you’d like to investigate in R).

# read in the data from your first allele frequency simulation

af1 <- read.csv("~/slim/AF_trial1.csv")

In the above example, you’re using the read.csv function (which is built in to R) to save your allele frequency file with the variable name af1. If you did head(af1) or print(af1) you would see the contents of this file. We use the read.csv function becase your file was saved in csv (comma separated values) format, as you can see from the suffix of your filename.

3.15.4 Using published packages

We will plot with the package ggplot2. This library offers a set of functions designed for visualizations in R. You can read more about it here.

When using a new package, you’ll first need to install it into your workspace’s library. Each time you write a script you will load any package you wish to use from your library.

# the first time you're using ggplot (or any package), install it
install.packages("ggplot2")

# each time you want to use the package, load it in your script 
library(ggplot2)

Note that the install.packages() function requires its argument to be in quotes (a string) but the library() function accepts a variable name without quotes.

3.15.5 Psuedocode

It is often helpful to outline in plain language (not code) what you intend to do in your script. This allows you to clearly outline your plans and design your plots without getting bogged down by the syntax of your code. Comments (which begin with the # in R) are a useful way to do this.

3.15.6 `ggplot2` basics

The main function in ggplot2 is very simply ggplot(). It then takes the major arguments data (which will be the data you read in from your file, i.e., af1 in the above example) and aes(). Aes() is short for aesthetics; it usually includes the data column needed for the x axis and the data column needed for the y axis. It can also include the data column used for the color.

Additional information is then applied to that function via +. To plot the frequency trajectory of an allele in a population, we’ll use a line plot; this is added through the geom_line() function. We’ll also specify which data column to label the information by via the labs() function.

Check out the ggplot2 website above for examples of additional information and customization you can apply to your plots. You might consider theme_classic() or setting a limit on the x axis via xlim(), in addition to others.

The syntax would therefore be something like:

# create your plot 
ggplot(data = my_data, aes(x = column_for_x_axis, y = column_for_y_axis, color = as.factor(column_to_base_colors_on))) + 
  geom_line() + # create a line plot 
  theme_classic() + # apply a theme
  labs(colour="Column that color is based on") # indicate that the color is based on whatever column you chose

3.15.7 Saving a plot

When you’re done with your plot, you can save it by going to the file menu and exporting it as you would any other file. Or you can do so in R code with the function ggsave():

# assign your plot to an object 
plot1 <- ggplot(data = my_data, aes(x = column_for_x_axis, y = column_for_y_axis, color = as.factor(column_to_base_colors_on))) + 
  geom_line() + # create a line plot 
  theme_classic() + # apply a theme
  labs(colour="Column that color is based on")

ggsave(plot1, "filepath/filename.pdf")

The function ggsave() is built-in to the ggplot2 package. It expects two arguments: the variable name of the plot you’re interested in printing, and then the name of the file (along with full file path) you’d like to save the image to.

3.15.8 Assignment

Work through the ggplot2 tutorial.
Plot the frequency trajectory of one allele in the population.

Write psuedocode for each step (think about the x axis, the y axis, what style (e.g., scatterplot, line plot) would best represent your data and answer your question).
Fill in the name of the data as well as the aesthetics (x and y). Remember that you’re plotting allele frequency, which is a measure of the number of a given allele in the population divided by the number of possible instances at that site. For example, if you have 10 copies of the allele A at a given position and there are 20 individuals in your population, there are 40 possible instances at that site (humans are diploid - two copies of each chromosome), so the frequency of the A allele is 10 / (20 * 2) = 0.25.

Update your code to include three additional allele frequency trajectories on a single plot. To do so, merge the data from the four allele frequency files into a single data structure.

Add any additional psuedocode necessary.
Consider using the rbind() function to combine the information from the four trials into a single data structure.
Add an aesthetic to color each allele frequency trajectory by the value in the Trial column. Consider using the as.factor() function to ensure each trial is treated as a discrete variable and plotted in a distinct color.
Limit the x axis to include only values from 0 to 4000.
Explore various themes available in ggplot2 and apply one to your plot.

Email your code and the resulting pdf files for the images.