Transcriptomic meta-analysis enhances biological veracity and reproducibility by integrating diverse studies, yet prevailing P-value or effect-size integration approaches exhibit limited power to resolve subtle signatures. We present AWmeta, an adaptively-weighted framework that unifies both paradigms. Benchmarking across 35 Parkinson's and Crohn's disease datasets spanning diverse tissues and adaptively down-weighting underpowered studies, AWmeta yields higher-fidelity differentially expressed genes (DEGs) with markedly reduced false positives, and establishes superior gene differential quantification convergence at both gene and study levels over state-of-the-art random-effects model (REM) and original studies. AWmeta requires fewer samples and DEGs from original studies to achieve substantial gene differential estimates, lowering experimental costs. We demonstrate AWmeta's remarkable stability and robustness against external and internal perturbations. Crucially, AWmeta prioritizes disease tissue-specific mechanisms with higher functional coherence than those from REM and original studies. By bridging statistical rigor with mechanistic interpretability, AWmeta harmonizes heterogeneous transcriptomic data into actionable insights, serving as a transformative tool for precision transcriptomic integration.
Clone this repository and rename to your project name.
git clone https://github.com/YanshiHu/AWmeta.git
mv AWmeta my_project_name
cd my_project_name
wget https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-Linux-x86_64.sh
bash Miniforge3-Linux-x86_64.shTip
1. Miniforge holds the minimal installers for Conda and Mamba specific to conda-forge.
2. Mamba is a reimplementation of the Conda package manager in C++, thus much more faster.
sudo apt install zlib1g-dev libncurses5-dev libbz2-dev liblzma-devmamba create -n AWmeta
mamba activate AWmetamamba install conda-forge::r-base==4.1.0
mamba install bioconda::bioconductor-awfisher
mamba install bioconda::bioconductor-metavolcanorinstall.packages("devtools")
devtools::install_github("metaOmics/MetaDE")source("AWmeta.R")2. Run following command to do adaptively-weighted transcriptomic meta-analysis (step-by-step tutorial):
AWmeta(raw.data.dir, raw.clin.dir, raw.sep, DE.method, compare.group, ref.level, paired, core.num)
Arguments:
raw.data.dir A path to the raw expression data files.
raw.clin.dir A path to the clinical/phenotype data files.
raw.sep A field separator character. Values on each line of the raw expression
and clinical/phenotype data file are separated by this character.
DE.method A character vector specifying the DE analysis method(s).
- Single Method (length 1): e.g., `"limma"`.
Applies this method to all studies.
- Two Methods (length 2): e.g., `c("limma", "DESeq2")`.
Provide one continuous method ("limma", "sam") for microarray
data and one discrete method ("edgeR", "DESeq2", "limmaVoom")
for RNA-seq data. The function will auto-detect if a study's
data is integer-based (discrete) or decimal-based (continuous)
and apply the corresponding method.
compare.group A character vector of length 2 specifying the names of the two groups
to compare in the clinical/phenotype data (e.g., c("control", "PD")).
ref.level A character string specifying which one of the two groups in
`compare.group` is the reference level (e.g., "control").
paired A logical value (TRUE/FALSE) indicating whether the samples are paired.
core.num An integer specifying the number of CPU cores to use for parallel
computation.
To cite AWmeta:
Yanshi Hu$, Zixuan Wang, Yueming Hu, Cong Feng, Qiuyu Fang, Ming Chen$ (2025). AWmeta empowers adaptively-weighted transcriptomic meta-analysis. bioRχiv, 2025.05.06.650408.
If you have any questions or thoughts, please feel free to file an issue or direct to yanshihu [at] zju [dot] edu [dot] cn.