3.9 Writing Output

So far, we have used SLiM’s built-in functions to write output. Often, we will want to have a bit more control over our output.

The function used to write to a file in SLiM is writeFile(). We will always give writeFile() two arguments - the name of the file we want to write to, and the content we want to write. We typically will also provide the argument append = T or append = F. If append is False, then when we write to a file, we erase any existing contents in that file. If append is True, then then when we write to the file, we keep any existing contents and just add another line at the end.

For the sweep that we’ve introduced above, let’s say we want to run a few iterations of our simulation and save the following information:

Number of generations since m2 was introduced
Current allele frequency of m2
The trial number

We can save this in the format:

Generation, Allele.Frequency, Trial
0,50,2
1,57,2
2,55,2
3,47,2

Here is a version of our script from the previous page, modified to write the above information to file:

initialize(){
    initializeMutationRate(1e-7);
    // Create mutation type
    initializeMutationType("m1", 0.5, "f", 0.0);
    initializeMutationType("m2", 1, "f", 0.2);
    m2.convertToSubstitution = F;
    m1.convertToSubstitution = F;
    
    
    // Create Types of DNA 
    initializeGenomicElementType("g1", m1, 1);   
    
    // Arrange DNA into a genome
    initializeGenomicElement(g1, 0, 3000);
    
    initializeRecombinationRate(1e-4);
    
    defineConstant('trialNumber', 1);
    initializeSex("A");
}

1 early(){
    sim.addSubpop("p1", 5000);
}

300 late(){
  // Initialize generation counter
  sim.setValue("generationCount", 0);

  // Introduce sweep mutation at 5% allele frequency 
    p1_size = size(p1.genomes);
    initial_freq = asInteger(p1_size * 0.05);
    target = sample(p1.genomes, initial_freq);
    target.addNewDrawnMutation(m2, 1000);   
    
    // Write the header for your file
    line = "Generation, Allele.Frequency, Trial";
    
    // Get a file name
    // this is standard syntax for Mac; see below for non-Mac systems
    defineConstant("fname", paste("~/slim/AF_trial", trialNumber, ".csv", sep =""));
    
    // Write to file
    writeFile(fname, line, append=F); 
}

300:1000 late(){
    gen = sim.getValue("generationCount");
    // Count number of m2 mutations
    m2_count = sum(p1.genomes.countOfMutationsOfType(m2));
    // Write data to file
    line = paste(gen, m2_count, trialNumber, sep = ",");
    writeFile(fname, line, append=T);
    // Update generation counter
    sim.setValue("generationCount", gen + 1);
}


1000 late() { 
     print('Number of fixed mutations:');
     print(length(sim.substitutions));
     sim.simulationFinished();
}

Let’s break down the additions:

In the initialize statement, we add the following line: defineConstant('trialNumber', 1); This sets the trial number as a constant (persists throughout the simulation, cannot be modified) equal to 1.
In the code block at generation 300 where we add the sweep mutation, we add the following lines of code

// Initialize generation counter
sim.setValue("generationCount", 0);

This initializes a generation counter and sets it equal to 0. Because this is saved as a simulation value, we will be able to update the value of our counter throughout the simulation . This will be the leftmost column of our output. * In the code block at generation 300 where we add the sweep mutation, we add the following lines of code:

    // Write the header for your file
    line = "Generation, Allele.Frequency, Trial";
    
    // Get a file name
    defineConstant("fname", paste("~/AF_trial", trialNumber, ".csv", sep =""));
    
    // Write to file
    writeFile(fname, line, append=F);

Here, we define the local variable line to be the text of the first line of our file. Next, we define a constant containing our output file name. To do this, we use the paste() function - this function takes as input a set of variables and combines them together, separating them by the characters provided in the argument sep ="". For example, the above paste command produces the filename "~/AF_trial1.csv". Finally, we write to file. Because we do want this line to be the first line of our file, we do not append.

For non-Mac users, the syntax "~/directory/filename" won’t work. Instead, you’ll need to specify the exact path to where you want the file, starting with the system path (e.g., C for computer, D for downloads, etc. depending on your system). In the above example, everything would be the same, but you’d replace the defineConstant line with defineConstant("fname", paste("C:/Users/[user]/Downloads/AF_trial", trialNumber, ".csv", sep=""));. The “user” should be accessible via your system; for me, it’s firstnamelastname. You can also create a folder (titled “popsims”) within Downloads (or whatever directory you choose) to host the files from this class. In that case, your line would be defineConstant("fname", paste("C:/Users/[user]/Downloads/popsims/AF_trial", trialNumber, ".csv", sep=""));. This syntax will be the same for writing files in SLiM; just choose a filepath and folder that you’ll be able to access so you can view your output files and use them downstream in visualizations.

We’ve added the following code block:

300:1000 late(){
    gen = sim.getValue("generationCount");
    // Count number of m2 mutations
    m2_count = sum(p1.genomes.countOfMutationsOfType(m2));
    // Write data to file
    line = paste(gen, m2_count, trialNumber, sep = ",");
    writeFile(fname, line, append=T);
    // Update generation counter
    sim.setValue("generationCount", gen + 1);
}

Here, we access the m2 mutations using the syntax: p1.genomes.countOfMutationsOfType(m2). Then, we paste together the number of generations since m2 was introduced, the current m2 count, and the trial number, this time separating our entries by commas. We add this line to the end our file (by appending). Finally, we take our generation counter and raise it by one.