Metagenomic next-generation sequencing (mNGS)-based characterization of the viral spectrum in clinical pulmonary and peripheral blood samples of patients

Ning Zheng¹Hai-Long Yu²Bing-Jie Zhang¹Dan Wang¹Ya-Liang Ji¹Lu-Lu Dai¹Wen Li¹Sheng-Hui Li²Zhi-Liang Hu^3,4*Yi-Shan Zheng^1*

• ¹Department of critical care medicine, The Second Hospital of Nanjing, Affiliated to Nanjing University of Chinese Medicine., Nanjing, China
• ²Puensum Genetech Institute, Wuhan, Hebei Province, China
• ³Department of Infectious Disease, the Second Hospital of Nanjing, Nanjing University of Chinese Medicine, Nanjing, China
• ⁴Center for Global Health, School of Public Health, Nanjing Medical University, Nanjing, China

Background: Metagenomic next-generation sequencing (mNGS) enables comprehensive profiling of viral communities in clinical samples. However, comparative analyses of the virome across anatomical compartments and disease states remain limited. This study aims to characterize the virome in bronchoalveolar lavage fluid (BALF) and peripheral blood samples from patients with various clinical conditions using mNGS.Methods: A total of 338 clinical samples-including 240 BALF and 69 blood samples for DNA sequencing, and 18 BALF and 11 blood samples for RNA sequencing-underwent shotgun metagenomic sequencing. Following removal of host-derived reads, high-quality non-human sequences were aligned to a viral reference database. Virome composition was assessed through alpha and beta diversity metrics. Principal coordinates analysis was used to evaluate disease-related variation, and virus-bacteria associations in BALF were investigated via Spearman correlation.Results: Sequencing yielded an average of 51 million raw reads per sample, resulting in approximately 8 million non-human reads after host filtering. Distinct virome profiles were observed between BALF and blood samples. Bacteriophages dominated all groups, with Siphoviridae and Myoviridae as the most abundant families, although only 13.6% of viral abundance could be assigned to known families. Diversity analyses revealed significant differences between BALF and peripheral blood, and DNA-sequenced BALF samples showed disease-specific viral signatures in pulmonary infections. In contrast, tumor presence had no significant effect on virome composition in either BALF or blood. Network analysis identified complex virus-bacteria correlations in BALF, with genera such as Haemophilus, Megasphaera, and Treponema as key bacterial hosts.This study reveals pronounced differences in virome composition between the respiratory and circulatory systems and highlights the specific influence of pulmonary disease-but not tumors-on the pulmonary virome. The observed virus-bacteria networks provide novel insights into pulmonary microbial ecology and underscore the importance of integrating host and disease context in virome studies.