Simulating a population that mimics Yuroba (YRI) from the 1000 genomes project

[Albert Doughan]

Hi Ben,

I study genetic and phenotypic diversity from population to population using ancestry informative markers (AIMs, i.e. SNPs with high allele frequency differentials among populations). Our collaborator developed a new method, myESL, that can identify these AIMs and I am benchmarking the new method against existing ones like FST, Delta and Informativeness.

 
I am doing this by simulating a simple population that mimics the Yuroba (YRI) population from the 1000 Genomes Project. For simplicity, I am focusing on chromosome 22.

 
Below is the SLiM script I am using to generate the population and it runs without any issues. Am I on the right track? Once this simple model is successful, I will incorpoate more parameters and make it complex.

 
Thank you!


initialize() {
defineConstant("N_YRI", 12500);  // Effective population size for YRI
defineConstant("MU", 1.2e-8);     // Human mutation rate per bp per generation
defineConstant("R", 1e-8);        // Recombination rate (approximately 1 cM/Mb)
defineConstant("CHR_LENGTH", 50818468);  // Length of chr22 in bp
defineConstant("BURN_IN", 10 * N_YRI);   // Burn-in period
defineConstant("SAMPLE_TIME", BURN_IN + 1000);  // When to take samples

// Initialize mutation rate
initializeMutationRate(MU);

// Define mutation types
initializeMutationType("m1", 0.5, "f", 0.0);

// m2: deleterious mutations (optional, can be added for realism)
// initializeMutationType("m2", 0.5, "g", -0.01, 0.2);

// Define genomic element type
initializeGenomicElementType("g1", m1, 1.0);

// Define the chromosome
initializeGenomicElement(g1, 0, CHR_LENGTH - 1);

// Set recombination rate
initializeRecombinationRate(R);
}

// Create the YRI population
1 early() {
sim.addSubpop("p1", N_YRI);

}

// Burn-in period: let population reach mutation-drift equilibrium
1:BURN_IN late() {
// Monitor progress every 1000 generations during burn-in
if (sim.cycle % 1000 == 0) {
cat("Burn-in generation " + sim.cycle + " / " + BURN_IN + "\n");
cat("  Segregating mutations: " + size(sim.mutations) + "\n");
cat("  Mean heterozygosity: " + calcHeterozygosity(p1.haplosomes) + "\n\n");
}
}

// After burn-in, run for additional generations
(BURN_IN+1):SAMPLE_TIME late() {
if (sim.cycle % 100 == 0) {
cat("Post-burn-in generation " + sim.cycle + "\n");
}
}

// Sample output at end
SAMPLE_TIME late() {
cat("\n=== SIMULATION COMPLETE ===\n");
cat("Final generation: " + sim.cycle + "\n");
cat("Population size: " + p1.individualCount + "\n");
cat("Segregating mutations: " + size(sim.mutations) + "\n");
cat("Fixed mutations: " + size(sim.substitutions) + "\n");

// Calculate summary statistics
cat("\nSummary Statistics:\n");
cat("  Mean heterozygosity: " + calcHeterozygosity(p1.haplosomes) + "\n");

// Output a sample of individuals (similar to 1000G sample size)
cat("\n=== OUTPUTTING SAMPLE ===\n");
cat("Sampling 108 individuals (216 haplotypes) to match 1000G YRI...\n");

// Sample 108 individuals randomly
sampledIndividuals = sample(p1.individuals, 108);

// Output in VCF format 
sampledIndividuals.haplosomes.outputHaplosomesToVCF(filePath="yri_chr22_sample.vcf");

cat("Sample output complete. VCF written to: yri_chr22_sample.vcf\n");

sim.simulationFinished();
}

[Ben Haller]

Hi Albert!

SLiM has hundreds of users, if not perhaps thousands.  I'm sorry, but I can't review everybody's scripts for correctness.  If you have a specific SLiM question, feel free to ask it on here.  If you just want to know whether your script is working correctly, I'd suggest that you employ common techniques like code review with a collaborator, analysis of the results of the script to see that they match expectations, and comparison of results against a separate, independent model written without the use of SLiM.  Good luck!

Cheers,
-B.

Benjamin C. Haller
Messer Lab
Cornell University


Albert Doughan wrote on 10/6/25 11:59 AM:

--
SLiM forward genetic simulation: http://messerlab.org/slim/
Before posting, please read http://benhaller.com/slim/bugs_and_questions.html
---
You received this message because you are subscribed to the Google Groups "slim-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to slim-discuss...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/slim-discuss/96cbc972-a8b8-42af-b8bf-0d3349da22dfn%40googlegroups.com.

[Albert Doughan]

Acknowledged! Thank you

[Andrew Kern]

Hi Albert-- if you are interested in simulating YRI easily, I'd recommend taking a look at the stdpopsim project: https://popsim-consortium.github.io/stdpopsim-docs/stable/index.html

Andy

To view this discussion visit https://groups.google.com/d/msgid/slim-discuss/684e0752-6e83-48b0-b0cf-7485dd471b15n%40googlegroups.com.