EUVIMEDEUVIMED entdecken
Crossref

A two-part investigation into atopic asthma: severity prediction using machine learning and genomic architecture analysis of significant SNPs

Ghinwa Ababidi, Dima Joujeh, Mohamad Taher Anan

Trakya University Journal of Natural Sciences · 2026

DOI: 10.23902/trkjnat.1976250 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Trakya University Journal of Natural Sciences, 2026. Verfasst von Ghinwa Ababidi und 2 weiteren Autor:innen. Die Arbeit nennt 16 Literaturangaben. Gefördert von: The authors declared that this study has received no financial support..

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Asthma is a widespread chronic condition characterized by fluctuating respiratory symptoms and limited airflow. It affects approximately 334 million people globally.

Aims

This study investigated the genetic and clinical factors influencing atopic asthma severity in a cohort of 138 patients.

Methods

The research employed a dual-pathway parallel framework. It utilized a cross-sectional design to investigate 138 patients diagnosed with atopic asthma, implementing a clinical machine learning approach to predict disease severity based strictly on patient-level data. In a separate, parallel in silico component, linkage disequilibrium (LD) and haplotype block analysis was performed using the LDpair tool on published GWAS single nucleotide polymorphisms to explore the independent genomic architecture associated with the disease.

Results

First, a predictive model was developed to forecast asthma severity using clinical data, including patient age, gender, duration of illness, and family relatedness. Both a Neural Network (NN) and a Support Vector Machine (SVM) were trained, with the NN achieving an overall accuracy of 92% and the SVM achieving a a high level of discriminative accuracy on the test set. These results highlight the significant predictive power of familial and clinical variables. The genetic analysis revealed a complex genetic architecture, with heterogeneous LD patterns across chromosomes and the presence of strong haplotype blocks, particularly on Chromosome 9. These findings suggest that GWAS signals may be driven by co-inherited genetic blocks rather than individual variants.

Conclusion

The study concludes that a parallel, complementary framework combining independent bioinformatics insights with clinical predictive modeling is highly effective for understanding and assessing atopic asthma severity, offering a robust foundation for future clinical risk assessment tools.

Förderung

  • The authors declared that this study has received no financial support.

Zitieren

APA

Ababidi, G., Joujeh, D., & Anan, M. T. (2026). A two-part investigation into atopic asthma: severity prediction using machine learning and genomic architecture analysis of significant SNPs. Trakya University Journal of Natural Sciences. https://doi.org/10.23902/trkjnat.1976250

Vancouver

Ababidi G, Joujeh D, Anan MT. A two-part investigation into atopic asthma: severity prediction using machine learning and genomic architecture analysis of significant SNPs. Trakya University Journal of Natural Sciences. 2026. doi:10.23902/trkjnat.1976250

BibTeX
@article{Ababidi2026,
  author = {Ababidi, Ghinwa and Joujeh, Dima and Anan, Mohamad Taher},
  title = {A two-part investigation into atopic asthma: severity prediction using machine learning and genomic architecture analysis of significant SNPs},
  journal = {Trakya University Journal of Natural Sciences},
  year = {2026},
  doi = {10.23902/trkjnat.1976250}
}
Kostenlos starten →