MuscleMeta-Exercise

Transcriptome Meta-Analysis of Acute Exercise Response in Human Skeletal Muscle

A re-analysis and visualization pipeline for the transcriptional response to acute exercise in human skeletal muscle, using the ExTraMeta preprocessed dataset (Amar et al., Nat Commun 2021) and REML random-effects meta-regression implemented in metafor.

---

Key Results

Forest plots for three canonical exercise-responsive genes across 23 study arms (endurance + resistance, multiple GSE cohorts):

Gene	Pooled logFC	95% CI	p-value	Interpretation
NR4A3	1.39	[0.75, 2.03]	<0.001	Strongly upregulated
PPARGC1A	0.58	[0.19, 0.97]	0.003	Consistently upregulated
SMAD3	0.09	[−0.04, 0.22]	0.18	Regulator; not direct target

NR4A3 — Acute Skeletal Muscle

PPARGC1A — Acute Skeletal Muscle (REML)

PPARGC1A — Stratified by Timepoint

SMAD3 — Acute Skeletal Muscle

---

Background

This pipeline re-analyses data from the ExTraMeta resource:

Amar D, Lindholm ME, Norrbom J, Wheeler MT, Rivas MA, Ashley EA. Time trajectories in the transcriptomic response to exercise — a meta-analysis. Nature Communications 12, 3471 (2021). https://doi.org/10.1038/s41467-021-23579-x

ExTraMeta integrates transcriptomic data from 43 studies (739 individuals) covering skeletal muscle and blood responses to acute and chronic exercise. This repo focuses on the acute skeletal muscle subset, applying REML random-effects meta-regression to study cross-cohort consistency of the exercise transcriptional response.

---

Pipeline Overview

ExTraMeta RData
      │
      ▼
01_load_extrameta_data.R      ← Load & inspect preprocessed data
      │
      ▼
02_run_meta_analysis.R        ← REML meta-regression (metafor::rma)
      │
      ▼
03_forest_plots.R             ← Publication-quality forest plots
      │
      ▼
04_time_stratified_analysis.R ← Stratify PPARGC1A response by timepoint
      │
      ▼
05_export_cohort_sources.R    ← Export GSE/cohort metadata tables
06_trace_yi_vi_generation.R   ← Inspect effect size / variance provenance

---

Data Sources

All data used in this analysis are from publicly available resources. No raw patient data are included in this repository.

Resource	URL	Description
ExTraMeta	https://extrameta.org	Preprocessed meta-analysis input (RData)
GEO	https://www.ncbi.nlm.nih.gov/geo/	Original expression datasets

GSE accessions used: GSE19062, GSE27285, GSE28392, GSE28998, GSE41769, GSE43219, GSE43856, GSE45426, GSE59088, GSE87749, GSE106865, GSE107934

See data/README.md for download instructions.

---

Quick Start

Prerequisites

install.packages(c("metafor", "data.table", "dplyr", "ggplot2"))
if (!require("BiocManager")) install.packages("BiocManager")
BiocManager::install(c("GEOquery", "limma", "Biobase"))

Run

# 1. Load ExTraMeta preprocessed data
source("scripts/01_load_extrameta_data.R")

# 2. Run REML meta-analysis
source("scripts/02_run_meta_analysis.R")

# 3. Generate forest plots
source("scripts/03_forest_plots.R")

# 4. Time-stratified analysis
source("scripts/04_time_stratified_analysis.R")

# 5. Export cohort metadata
source("scripts/05_export_cohort_sources.R")

---

Repository Structure

MuscleMeta-Exercise/
├── scripts/
│   ├── 01_load_extrameta_data.R        # Data loading & inspection
│   ├── 02_run_meta_analysis.R          # REML meta-regression
│   ├── 03_forest_plots.R               # Forest plot generation
│   ├── 04_time_stratified_analysis.R   # Timepoint-stratified analysis
│   ├── 05_export_cohort_sources.R      # Cohort metadata export
│   └── 06_trace_yi_vi_generation.R     # Effect size provenance tracing
├── figures/
│   ├── NR4A3_acute_muscle_forest.png
│   ├── PPARGC1A_acute_muscle_forest.png
│   ├── PPARGC1A_acute_muscle_REML_forest.png
│   ├── PPARGC1A_acute_muscle_REML_forest_by_time.png
│   └── SMAD3_acute_muscle_forest.png
├── data/
│   └── README.md                       # Data download instructions
├── docs/
│   └── pipeline.md                     # Detailed methods
└── results/                            # Output directory (gitignored)

---

Citation

If you use this pipeline, please cite the original ExTraMeta paper:

@article{amar2021extrameta,
  title   = {Time trajectories in the transcriptomic response to exercise — a meta-analysis},
  author  = {Amar, David and Lindholm, Malene E and Norrbom, Jessica and
             Wheeler, Matthew T and Rivas, Manuel A and Ashley, Euan A},
  journal = {Nature Communications},
  volume  = {12},
  pages   = {3471},
  year    = {2021},
  doi     = {10.1038/s41467-021-23579-x}
}

---

License

MIT License. See LICENSE for details.