mtDNA_mt2_XL

A simulation study estimating mitochondrial DNA (mtDNA) variance components using structural equation modeling via OpenMx. Simulated pedigree data are generated across combinations of variance components and pedigree structures, and models are fit to recover those components.

Dependencies

• R packages: OpenMx, mvtnorm

---

Project Structure

mtDNA_mt2_XL/
├── InitialData/          # Pedigree structures, variance combinations, prep scripts
├── Functions/            # Core simulation and helper functions
├── Analysis/             # Power analysis plots and figures
├── Result_*/             # Simulation output (one folder per run configuration)
│   ├── cXpY/             # Results for variance combo X, pedigree Y
│   └── Analysis_*/       # Summary analyses for that result set
├── RunThis*.R            # Top-level scripts to launch simulations
├── compress_results.ps1  # Compress results for storage/sharing (maintainers)
└── decompress_results.R  # Extract results after cloning (users)

---

Naming Convention

Result subfolders follow the pattern cXpY:

• X = row index in InitialData/VarianceComb.csv (the variance component combination)
• Y = row index in InitialData/Pedigrees.csv (the pedigree structure)

For example, c3p7 contains results for variance combination 3 run on pedigree structure 7.

---

Workflow

1. Prepare initial data

Before running simulations, generate the fixed pedigrees and load variance combinations:

source("InitialData/InitialDataPrep.R")

This reads Pedigrees.csv and VarianceComb.csv, simulates pedigree structures, and saves everything to InitialData/FixedPedVar.RData.

2. Run simulations

Use the appropriate RunThis*.R script depending on the simulation configuration:

Script	Description	Output
RunThis.R	Base simulation (n=10,000)	Result2/
RunThis_r2.R	20k simulation	Result_20k/
RunThis_a2mt2.R	A2/Mt2 model only (n=20,000)	ResultA2Mt2_20k_2/
RunThis_rdPed.R	Random pedigree, full model	Result_rdped_full/
RunThis_PedCheck.R	Pedigree structure checks	Result_Check/

Each script loops over variance combinations (rows of VarianceComb.csv) and pedigree structures (rows of Pedigrees.csv), writing one cXpY/ subfolder of results per combination.

3. Analyse results

Analysis scripts are in the Analysis/ folder and generate power plots grouped by chain/partition combinations (e.g., PowPlotc1p1234.R).

---

Input Data

InitialData/Pedigrees.csv

Defines pedigree structures used in simulations.

Column	Description
k	Kids per couple
G	Number of generations
p	Sex ratio
r	Marriage rate

InitialData/VarianceComb.csv

Defines variance component combinations to simulate.

Column	Description
a2	Additive genetic variance
cN2	Nuclear common environment
cE2	Extended common environment
mt2	Mitochondrial variance
d2	Dominance variance
j2	Additive × mitochondrial interaction
e2	Residual variance

---

File Compression

Result subfolders are stored as .tar.gz archives to keep the repository size manageable. Analysis* subfolders are excluded.

After cloning — decompress results

source("decompress_results.R")

Or from a terminal at the repo root:

Rscript decompress_results.R

This extracts all .tar.gz archives across every Result* directory.

For maintainers — compress results

Run from the repo root in PowerShell:

& ".\compress_results.ps1"

---

Seeds

Run	Seed
rdped_v1	11
rdped_v2	12
rdped_v3	13
rdped_20k	20
rdped_full	100