EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Molecular group and correlation guided structural learning for multi-phenotype prediction

Xueping Zhou, Manqi Cai, Molin Yue, Juan C Celedón, Jiebiao Wang, Ying Ding, Wei Chen, Yanming Li

Briefings in Bioinformatics · 2024

DOI: 10.1093/bib/bbae585 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Briefings in Bioinformatics, 2024. Verfasst von Xueping Zhou und 7 weiteren Autor:innen. Beteiligte Einrichtung: Department of Biostatistics, University of Pittsburgh, Pittsburgh, PA 15216, United States. Hauptthemen laut MEDLINE: Phenotype, Computational Biology. Die Arbeit nennt 34 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: NHLBI NIH HHS, NIH HHS, NSF.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

We propose a supervised learning bioinformatics tool, Biological gRoup guIded muLtivariate muLtiple lIneAr regression with peNalizaTion (Brilliant), designed for feature selection and outcome prediction in genomic data with multi-phenotypic responses. Brilliant specifically incorporates genome and/or phenotype grouping structures, as well as phenotype correlation structures, in feature selection, effect estimation, and outcome prediction under a penalized multi-response linear regression model. Extensive simulations demonstrate its superior performance compared to competing methods. We applied Brilliant to two omics studies. In the first study, we identified novel association signals between multivariate gene expressions and high-dimensional DNA methylation profiles, providing biological insights for the baseline CpG-to-gene regulation patterns in a Puerto Rican children asthma cohort. The second study focused on cell-type deconvolution prediction using high-dimensional gene expression profiles. Using Brilliant, we improved the accuracy for cell-type fraction prediction and identified novel cell-type signature genes.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

PhenotypeComputational BiologyHumansDNA MethylationAsthmaLinear ModelsCpG IslandsGenomics

Stichwörter der Autor:innen

DNA methylationassociation studycell-type deconvolutionfeature selectiongenomic grouping structuremulti-type prediction

Autor:innen und Institutionen

  • Department of Biostatistics, University of Pittsburgh, Pittsburgh, PA 15216, United States.
  • Department of Pediatrics, University of Pittsburgh Medical Center Children's Hospital of Pittsburgh, Pittsburgh, PA 15224, United States.
  • Department of Biostatistics & Data Science, University of Kansas Medical Center, Kansas, KS 66160, United States.

Förderung

  • NHLBI NIH HHS
  • NIH HHS
  • NSF
  • National Science Foundation
  • NIH
  • National Institutes of Health

Zitieren

APA

Zhou, X., Cai, M., Yue, M., Celedón, J. C., Wang, J., Ding, Y., Chen, W., & Li, Y. (2024). Molecular group and correlation guided structural learning for multi-phenotype prediction. Briefings in Bioinformatics. https://doi.org/10.1093/bib/bbae585

Vancouver

Zhou X, Cai M, Yue M, Celedón JC, Wang J, Ding Y, et al. Molecular group and correlation guided structural learning for multi-phenotype prediction. Briefings in Bioinformatics. 2024. doi:10.1093/bib/bbae585

BibTeX
@article{Zhou2024,
  author = {Zhou, Xueping and Cai, Manqi and Yue, Molin and Celedón, Juan C and Wang, Jiebiao and Ding, Ying and Chen, Wei and Li, Yanming},
  title = {Molecular group and correlation guided structural learning for multi-phenotype prediction},
  journal = {Briefings in Bioinformatics},
  year = {2024},
  doi = {10.1093/bib/bbae585}
}
Kostenlos starten →