Gene expression matrix and clinical metadata for COPD, NSCLC, and NSCLC with COPD

Raw RNA-seq profiles were downloaded from NCBI-GEO (accession numbers GSEXXXXX, GSEYYYYY, …), uniformly processed (batch-effect removal, duplicate removal, outlier filtering) and quantile-normalized to TPM followed by log2 transformation.

The matrix contains 19 848 protein-coding genes (including 1 133 mitochondria-related genes) as rows and samples as columns. Clinical variables comprise 26 fields such as age, sex, smoking pack-years, FEV1 %, tumor stage, histology, etc. The clean, ready-to-use data enable differential expression, WGCNA, machine-learning classification/regression, and mitochondria-focused functional analyses aimed at identifying shared or disease-specific mitochondrial gene signatures underlying COPD–NSCLC comorbidity.