A tree-based kernel for densities and its applications in clustering DNase-seq profiles
Biometrics · 2026
DOI: 10.1093/biomtc/ujag154 ↗Kurz gefasst
Fachartikel in der Zeitschrift Biometrics, 2026. Verfasst von Yuliang Xu und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Applied Mathematics and Statistics, Johns Hopkins University, Baltimore, MD 21218, United States. Hauptthemen laut MEDLINE: Models, Statistical, Deoxyribonucleases, Sequence Analysis, DNA. Die Arbeit nennt 39 Literaturangaben. Gefördert von: NIGMS NIH HHS, NIGMS.
Abstract (Originalsprache)
Modeling multiple sampling densities within a hierarchical framework enables borrowing of information across samples. These “density random effects” can act as kernels in latent variable models to represent exchangeable subgroups or clusters. A key feature of these kernels is the (functional) covariance they induce, which determines how densities are grouped in mixture models. Our motivating problem is clustering chromatin accessibility profiles from high-throughput DNase-seq experiments to detect transcription factor (TF) binding. TF binding typically produces footprint profiles with spatial patterns, creating long-range dependency across genomic locations. Existing nonparametric hierarchical models impose restrictive covariance assumptions and cannot accommodate such dependencies, often leading to biologically uninformative clusters. We propose a nonparametric density kernel that is flexible enough to capture diverse covariance structures and adapts to various spatial patterns of TF footprints. The kernel specifies dyadic tree splitting probabilities via a multivariate logit-normal model with a sparse precision matrix. Bayesian inference for latent variable models using this kernel is implemented through Gibbs sampling with Pólya–Gamma augmentation. Extensive simulations show that our kernel substantially improves clustering accuracy. We apply the proposed mixture model to DNase-seq data from the Encyclopedia of DNA Elements project, which results in biologically meaningful clusters corresponding to binding events of two common TFs.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- Department of Applied Mathematics and Statistics, Johns Hopkins University, Baltimore, MD 21218, United States.
- Department of Human Genetics, University of Chicago, Chicago, IL 60637, United States.
- Department of Statistics & Data Science Institute, University of Chicago, Chicago, IL 60637, United States.
ORCID-Profile: Yuliang Xu, Kaixuan Luo, Li Ma
Förderung
- NIGMS NIH HHS
- NIGMS
Zitieren
APA
Xu, Y., Luo, K., & Ma, L. (2026). A tree-based kernel for densities and its applications in clustering DNase-seq profiles. Biometrics. https://doi.org/10.1093/biomtc/ujag154
Vancouver
Xu Y, Luo K, Ma L. A tree-based kernel for densities and its applications in clustering DNase-seq profiles. Biometrics. 2026. doi:10.1093/biomtc/ujag154
BibTeX
@article{Xu2026,
author = {Xu, Yuliang and Luo, Kaixuan and Ma, Li},
title = {A tree-based kernel for densities and its applications in clustering DNase-seq profiles},
journal = {Biometrics},
year = {2026},
doi = {10.1093/biomtc/ujag154}
}
EUVIMED