摘要目的:研究2型及非2型支气管哮喘（简称哮喘）患者呼吸道共生微生物网络的作用特征。方法:基于哮喘患者队列的前瞻性研究。采集2021年5月至2022年5月，首都医科大学附属北京朝阳医院呼吸与危重症学科收治的55例哮喘患者［男25例，女30例，中位年龄47.7岁（年龄范围34.3~63.0岁）］及来自首都医科大学附属北京朝阳医院体检中心的12名健康对照者的呼吸道诱导痰样本；根据呼出气一氧化氮（FeNO）水平，将哮喘患者分为高FeNO组22例（FeNO≥40 ppb，即2型哮喘组）、低FeNO组33例（FeNO<40 ppb，即非2型哮喘组）。对所有呼吸道诱导痰样本进行宏基因组二代测序并进行微生物群落多样性、组成特征、共生网络特征和代谢功能预测等生物信息学分析。采用Kruskal-Wallis秩和检验进行组间比较，采用线性判别分析（LEfSe）方法比较各组间的菌群组成差异。R语言用于微生物网络分析。此外，还采用PICRUSt预测微生物群落的代谢功能特征。结果:健康对照组微生物群落中，厚壁菌门和变形菌门的占比（分别为29%和21%）均少于哮喘组患者（低FeNO组分别占37%和33%，高FeNO组分别占42%和26%）。低FeNO组的微生物网络有64对边，形成了16个群落；约75%的节点特征向量中心度的数值在0~0.05，25%的节点的特征向量中心度的数值在0.10~0.45；κ-核分解有4层，大约42%的顶点在中心的2层。高FeNO组的微生物网络有80对边，形成了18个群落；81% 的节点特征向量中心度的数值在0~0.05，19% 的节点的特征向量中心度的数值在0.10~0.35；κ-核分解有8层，21%的顶点位于中心的2层。低FeNO组和高FeNO组的主要功能差异表现在代谢途径（包括糖类、脂类、氨基酸和能量代谢）、抗药性、生物膜传输、信号传导、胞间通讯、细胞修复中。结论:与非2型哮喘患者相比，2型哮喘患者的呼吸道微生物菌群的α-多样性更高，变形菌门的微生物含量更低，微生物网络更为聚集；两种内型哮喘患者的代谢功能预测结果有显著差异。

abstractsObjective:To study the characteristics of the airway microbiome co-occurrence network in patients with type 2 and non-type 2 asthma.Methods:In a prospective study based on a cohort of asthma patients, respiratory induced sputum samples were collected from 55 asthma patients [25 males and 30 females, with a median age of 47.7 years (age range 34.3-63.0 years)] admitted to the Department of Respiratory and Critical Care, Beijing Chaoyang Hospital, Capital Medical University and 12 healthy controls from the Physical Examination Centre of Beijing Chaoyang Hospital, Capital Medical University, from May 2021 to May 2022. According to the level of exhaled breath nitric oxide (FeNO), the asthma patients were divided into 22 cases in the high FeNO group (FeNO≥40 ppb, i.e., type 2 asthma group) and 33 cases in the low FeNO group (FeNO<40 ppb, i.e., non-type 2 asthma group). All induced sputum samples were subjected to second-generation macrogenomic sequencing and bioinformatic analyses of microbial community diversity, compositional characteristics, symbiotic network characteristics and metabolic function prediction. The Kruskal-Wallis rank sum test was used for between-group comparisons, and the linear discriminant analysis (LEfSe) method was used to compare the differences in flora composition between groups. The R language was used for microbial network analysis. In addition, PICRUSt was used to predict the metabolic-functional characteristics of the microbial communities. Results:The microbial communities in the healthy control group had a lower proportion of p_Firmicutes and p_Proteobacteria than asthma patients, 29% and 21%, respectively; 37% and 33% in the low FeNO group and 42% and 26% in the high FeNO group. The microbial network in the low FeNO group had 64 pairs of edges forming 16 communities, and about 75% of the nodes had eigenvector centrality values between 0 and 0.05, and 25% of the nodes had eigenvector centrality values between 0.10 and 0.45. There were four layers of κ-nucleosynthesis, and about 42% of the vertices were in the centre of the two layers. The microbial network of the high-FeNO group had 80 pairs of edges forming 18 clusters, and 81% of the nodes had eigenvector centrality values between 0 and 0.05, and 19% of the nodes had eigenvector centrality values between 0.10 and 0.35. The κ-nucleus decomposition had eight layers, and 21% of the vertices were located in the centre′s two layers. The main functional differences between the low and high FeNO groups were shown in metabolic pathways (including sugar, lipid, amino acid, and energy metabolism), drug resistance, biofilm transport, signalling, intercellular communication, and cellular repair. Conclusions:Compared with non-type 2 asthmatics, type 2 asthmatics had a higher alpha diversity of respiratory microbiota, lower levels of microorganisms in the p_Proteobacteria, and a more aggregated microbial network. There was a significant difference in the predicted metabolic function of the two endotypes of asthmatics.