Introduction Studies of chronic obstructive pulmonary disease (COPD) and asthma typically focus on the diseases separately, and there is a lack of a unique COPD gene expression signature distinguishing it from asthma. This study aims to find a unique gene set for COPD by integrating three airway datasets in patients with COPD and asthma.

Methods Three RNA-sequencing datasets, GLUCOLD (n=56, COPD), Indurain (n=107/77, Asthma/Healthy), and STOP (n=20/14, COPD/Healthy), were integrated using the COMBAT package. This method adjusted for batch effects of three different bronchial biopsy studies, making it the largest bronchial biopsy cohort (n=274) for RNA-sequencing. Unique COPD signatures were created by differential gene expression analysis.

Results Our analysis identified 136 genes that were higher in expression in COPD compared to Healthy participants and 86 genes that were lower in expression (FDR<0.05). These signatures were significantly different from Healthy and Asthma participants (p<0.05), indicating that they are unique to COPD. Pathway analysis showed activation of inflammatory pathways, including NF-?B and interferon-? signalling pathways in COPD, while a decrease in DNA-binding transcription factors. These signatures were replicated in a publicly available COPD study (n=151/87, COPD/Healthy, GSE37147) and the downregulated signature was associated with the severity of the disease.

Conclusion Our study successfully created unique COPD-associated genes, which was validated in another COPD-related airway study. These findings improve our understanding of disease mechanisms and may help to better personalized treatment by identifying clinically relevant disease phenotypes.