Simplified detection of genetic background admixture using artificial intelligence
Clinical Genetics · 2024
DOI: 10.1111/cge.14527 ↗Kurz gefasst
Fachartikel in der Zeitschrift Clinical Genetics, 2024. Verfasst von Rini Pauly und 1 weiteren Autor:innen. Beteiligte Einrichtung: Biomedical Data Science & Informatics Program, Clemson University, Clemson, South Carolina, USA. Hauptthemen laut MEDLINE: Artificial Intelligence. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 52 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Admixture refers to the mixing of genetic ancestry from different populations. Admixture is important for genomic medicine because it can affect how an individual responds to certain medications, how they metabolize drugs, and susceptibility to certain diseases. For example, some genetic variants associated with drug metabolism and response may be more common in certain populations, and individuals with admixed ancestry may have a different frequency of these variants than individuals from the ancestral populations. Understanding the patterns of admixture in a population can also help researchers identify new genetic variants associated with diseases or traits and develop more personalized and targeted treatments. In this study, we compared and classified the known and self‐reported genetic backgrounds from 1000 Genomes Project and admixed samples from GTEx projects using supervised, unsupervised and statistical classification methodologies. We developed a novel tool called Admix‐AI that uses a one‐dimensional convolutional neural network to understand and classify admixed genetic backgrounds using 213 DNA‐marker based genetic background labels. Admix‐AI can be used to discover admixed proportions in samples and ultimately aid personalized genomic medicine by identifying specific biomarker systems. We compared Admix‐AI to the existing admixture categorization software and found our tool to be computationally faster with 2× speedup and streamlined usage. Admix‐AI is available as open‐source code under GPL version 3.0 license at https://github.com/rpauly/Admix-AI .
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Biomedical Data Science & Informatics Program, Clemson University, Clemson, South Carolina, USA.
- Genetics and Biochemistry Department, Clemson University, Clemson, South Carolina, USA.
- Center for Human Genetics, Clemson University, Greenwood, South Carolina, USA.
Zitieren
APA
Pauly, R., & Alexander Feltus, F. (2024). Simplified detection of genetic background admixture using artificial intelligence. Clinical Genetics. https://doi.org/10.1111/cge.14527
Vancouver
Pauly R, Alexander Feltus F. Simplified detection of genetic background admixture using artificial intelligence. Clinical Genetics. 2024. doi:10.1111/cge.14527
BibTeX
@article{Pauly2024,
author = {Pauly, Rini and Alexander Feltus, Frank},
title = {Simplified detection of genetic background admixture using artificial intelligence},
journal = {Clinical Genetics},
year = {2024},
doi = {10.1111/cge.14527}
}
EUVIMED