EUVIMEDOpdag EUVIMED
Open access · CC BY 4.0

Score Matching for Differential Abundance Testing of Compositional High‐Throughput Sequencing Data

Johannes Ostner iD, Hongzhe Li, Christian L Müller

Statistics in Medicine · 2026-04 · 45 (8-9)

DOI: 10.1002/sim.70534 ↗

At a glance

Journal article in Statistics in Medicine, published online on 2026-04-07. Written by Johannes Ostner and 2 co-authors. Institution: Computational Health Center, Helmholtz Munich, Neuherberg, Germany. Major MEDLINE topics: High-Throughput Nucleotide Sequencing, Models, Statistical. 53 references. The full text is freely available (open access, CC BY 4.0). Funded by: NIGMS NIH HHS, NIH HHS, Helmholtz-Gemeinschaft.

AbstractSubjectsBibliografiske oplysningerCiteRelated studies

Abstract (original language)

The class of a‐b power interaction models, proposed by [1], provides a general framework for modeling sparse compositional data with pairwise feature interactions. This class includes many distributions as special cases and enables modeling of zero entries through power transformations, making it particularly suitable for modern high‐throughput sequencing data with excess zeros, including single‐cell RNA‐Seq and microbial amplicon data. Here, we present an extension of this class of models that allows inclusion of covariate information, thus enabling accurate characterization of covariate dependencies in heterogeneous populations. Combining this model with a tailored differential abundance (DA) test leads to a novel DA testing scheme, cosmoDA , that can reduce the false positive detection rate caused by correlated features. cosmoDA uses penalized generalized score matching for parsimonious model fitting. We show on simulated benchmarks that cosmoDA can accurately estimate feature interactions in the presence of population heterogeneity and significantly reduces the false discovery rate when testing for differential abundance of correlated features. Using single‐cell and amplicon data, we illustrate cosmoDA 's ability to estimate data‐adaptive Box–Cox‐type data transformations and assess the impact of zero replacement and power transformations on downstream differential abundance results. cosmoDA is available at https://github.com/bio‐datascience/cosmoDA .

Subjects

MeSH terms (MEDLINE)

High-Throughput Nucleotide SequencingModels, StatisticalComputer SimulationHumansSingle-Cell Analysis

Author keywords

compositional datadifferential abundancegenerative modelmicrobiomescore matchingsingle‐cell RNA sequencing

Authors and institutions

  • Computational Health Center, Helmholtz Munich, Neuherberg, Germany.
  • Institut für Statistik, Ludwig-Maximilians-Universität München, Munich, Germany.
  • Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, USA.
  • Center for Computational Mathematics, Flatiron Institute, New York, New York, USA.

ORCID profiles: Johannes Ostner

Funding

  • NIGMS NIH HHS
  • NIH HHS
  • Helmholtz-Gemeinschaft
  • National Institutes of Health

Cite

APA

Ostner, J., Li, H., & Müller, C. L. (2026). Score Matching for Differential Abundance Testing of Compositional High‐Throughput Sequencing Data. Statistics in Medicine, 45(8-9). https://doi.org/10.1002/sim.70534

Vancouver

Ostner J, Li H, Müller CL. Score Matching for Differential Abundance Testing of Compositional High‐Throughput Sequencing Data. Statistics in Medicine. 2026;45(8-9). doi:10.1002/sim.70534

BibTeX
@article{Ostner2026,
  author = {Ostner, Johannes and Li, Hongzhe and Müller, Christian L},
  title = {Score Matching for Differential Abundance Testing of Compositional High‐Throughput Sequencing Data},
  journal = {Statistics in Medicine},
  year = {2026},
  volume = {45},
  number = {8-9},
  doi = {10.1002/sim.70534}
}
Start for free →