Related Experiment Video
Updated: Aug 30, 2025
07:35
A Knowledge Graph Approach to Elucidate the Role of Organellar Pathways in Disease via Biomedical Reports
Published on: October 13, 2023
1.7K
Diagnostic model based on bioinformatics and machine learning to distinguish Kawasaki disease using multiple datasets
Mengyi Zhang1,2, Bocuo Ke1,2, Huichuan Zhuo1,2
1Department of Laboratory Medicine, West China Second University Hospital, Sichuan University, No. 20, Section 3, Renmin South Road, Chengdu, 610041, PR, Sichuan Province, China.
BMC Pediatrics
|August 30, 2022
Summary
Researchers developed a machine learning diagnostic model to distinguish Kawasaki disease (KD) in children. This model identifies eight key genes, offering potential new biomarkers and therapeutic targets for KD.
Area of Science:
- Pediatric cardiology
- Immunology
- Bioinformatics
Background:
- Kawasaki disease (KD) is a primary cause of acquired heart disease in children, characterized by systemic vasculitis.
- Early and accurate diagnosis is crucial for managing KD and preventing cardiac complications.
Purpose of the Study:
- To develop a diagnostic model with prognostic capabilities to aid in distinguishing children with Kawasaki disease.
- To identify novel molecular biomarkers and potential therapeutic targets for KD.
Main Methods:
- Utilized gene expression datasets from Gene Expression Omnibus (GEO) for analysis.
- Employed differential expressed gene (DEG) screening, pathway enrichment analysis (Gene Ontology, Kyoto Encyclopedia of Genes and Genomes), random forest (RF), and artificial neural network (ANN) for model construction.
- Integrated weighted gene co-expression network analysis with DEG screening results.
Main Results:
- Identified 2,017 DEGs, with significant enrichment in immune response pathways.
- Constructed an eight-gene diagnostic model (VPS9D1, CACNA1E, SH3GLB1, RAB32, ADM, GYG1, PGS1, HIST2H2AC) using RF and ANN.
- Achieved excellent classification performance for KD diagnosis across different patient cohorts (patients, convalescents, healthy individuals) with high area under the curve values.
Conclusions:
- Successfully developed and validated a machine learning-based diagnostic model for Kawasaki disease.
- The identified gene signature shows promise as a diagnostic tool and may reveal new therapeutic strategies for KD.
