Integrating respiratory microbiome and host immune response through machine learning for respiratory tract infection diagnosis

Study information

Needs review
Quality issues

study design

case-control

Citation

PMID PubMed identifier for scientific articles.

39266570

DOI Digital object identifier for electronic documents.

10.1038/s41522-024-00548-y

URI

https://www.nature.com/articles/s41522-024-00548-y

Authors

Chen H, Qi T, Guo S, Zhang X, Zhan M, Liu S, Yin Y, Guo Y, Zhang Y, Zhao C, Wang X, Wang H

Journal

NPJ biofilms and microbiomes

Year

2024

At present, the diagnosis of lower respiratory tract infections (LRTIs) is difficult, and there is an urgent need for better diagnostic methods. This study enrolled 136 patients from 2020 to 2021 and collected bronchoalveolar lavage fluid (BALF) specimens. We used metatranscriptome to analyze the lower respiratory tract microbiome (LRTM) and host immune response. The diversity of the LRTM in LRTIs significantly decreased, manifested by a decrease in the abundance of normal microbiota and an increase in the abundance of opportunistic pathogens. The upregulated differentially expressed genes (DEGs) in the LRTIs group were mainly enriched in infection immune response-related pathways. Klebsiella pneumoniae had the most significant increase in abundance in LRTIs, which was strongly correlated with host infection or inflammation genes TNFRSF1B, CSF3R, and IL6R. We combined LRTM and host transcriptome data to construct a machine-learning model with 12 screened features to discriminate LRTIs and non-LRTIs. The results showed that the model trained by Random Forest in the validate set had the best performance (ROC AUC: 0.937, 95% CI: 0.832-1). The independent external dataset showed an accuracy of 76.5% for this model. This study suggests that the model integrating LRTM and host transcriptome data can be an effective tool for LRTIs diagnosis.

Experiment 1

Edit History Delete

Flag for review

Your review change was submittedDuplicate this Experiment Add a new Signature

Reviewed Marked as Reviewed by Folakunmi on 2024-11-5

Curated date: 2024/10/20

Curator: Aleru Divine

Revision editor(s): Aleru Divine, Folakunmi

Subjects

Location of subjects: China

Host species Species from which microbiome was sampled. Contact us to have more species added.: Homo sapiens

Body site Anatomical site where microbial samples were extracted from according to the Uber Anatomy Ontology: Lung Pulmo,Lung,lung

Condition The experimental condition / phenotype studied according to the Experimental Factor Ontology: Lower respiratory tract disease disease of lower respiratory tract,disease or disorder of lower respiratory tract,disorder of lower respiratory tract,lower respiratory tract disease,lower respiratory tract disease or disorder,lower respiratory tract disorder,Lower respiratory tract disease

Group 0 name Corresponds to the control (unexposed) group for case-control studies: Non-lower respiratory tract infections (LRTIs)

Group 1 name Corresponds to the case (exposed) group for case-control studies: Lower respiratory tract infections (LRTIs)

Group 1 definition Diagnostic criteria applied to define the specific condition / phenotype represented in the case (exposed) group: Subjects who met clinical or microbiologic diagnosis criteria for lower respiratory tract infections (LRTIs).

Group 0 sample size Number of subjects in the control (unexposed) group: 27

Group 1 sample size Number of subjects in the case (exposed) group: 41

Antibiotics exclusion Number of days without antibiotics usage (if applicable) and other antibiotics-related criteria used to exclude participants (if any): N/A

Lab analysis

Sequencing type: WMS

16S variable region One or more hypervariable region(s) of the bacterial 16S gene: Not specified

Sequencing platform Manufacturer and experimental platform used for quantifying microbial abundance: Illumina

Statistical Analysis

Data transformation Data transformation applied to microbial abundance measurements prior to differential abundance testing (if any).: raw counts

Statistical test: edgeR

Significance threshold p-value or FDR threshold used for differential abundance testing (if any): 0.05

MHT correction Have statistical tests be corrected for multiple hypothesis testing (MHT)?: Yes

Alpha Diversity

Shannon Estimator of species richness and species evenness: more weight on species richness: decreased

Chao1 Abundance-based estimator of species richness: decreased

Simpson Estimator of species richness and species evenness: more weight on species evenness: decreased

Richness Number of species: decreased

Signature 1

Edit History Delete

Flag for review

Your review change was submitted

Reviewed Marked as Reviewed by Folakunmi on 2024-11-5

Curated date: 2024/10/21

Curator: Martha KJ

Revision editor(s): Martha KJ, Aleru Divine

Source: Figure 2D and Dataset S4

Description: Barplot of the differential ambulance of micro-species between LRTIs and Non-LRTIs plotted by R

Abundance in Group 1: increased abundance in Lower respiratory tract infections (LRTIs)

NCBI	Quality Control	Links
Acinetobacter calcoaceticus/baumannii complex
Aspergillus fumigatus
Corynebacterium striatum
Haemophilus influenzae
Klebsiella aerogenes
Klebsiella pneumoniae
Moraxella catarrhalis
Pneumocystis jirovecii
Pseudomonas aeruginosa
Streptococcus pneumoniae
Tropheryma whipplei