Year: 2025
Type: article
Abstract: MOTIVATION: The development of high-throughput sequencing enabled the massive production of "omics" data for various applications in biology. By analyzing simultaneously paired datasets collected on the same samples, integrative statistical approaches allow researchers to get a global picture of such systems and to highlight existing relationships between various molecular types and levels. Here, we introduce NMFProfiler, an integrative supervised NMF that accounts for the stratification of samples into groups of biological interest. RESULTS: NMFProfiler was shown to successfully extract signatures characterizing groups with performances comparable to or better than state-of-the-art approaches. In particular, NMFProfiler was used in a clinical study on atopic dermatitis (AD) and to analyze a multi-omic cancer dataset. In the first case, it successfully identified signatures combining known AD protein biomarkers and novel transcriptomic biomarkers. In addition, it was also able to extract signatures significantly associated to cancer survival. AVAILABILITY AND IMPLEMENTATION: NMFProfiler is released as a Python package, NMFProfiler (v0.3.0), available on PyPI.
Source: Bioinformatics
Authors: Aurélie Mercadié, Éléonore Gravier, Gwendal Josse, Isabelle Fournier, Cécile Viodé, Nathalie Vialaneix, Céline Brouard
Institutions: Institut National de Recherche pour l'Agriculture, l'Alimentation et l'Environnement, Université de Toulouse, Mathématiques et Informatique Appliquées Toulouse, Inserm, Université de Lille, Centre Hospitalier Universitaire de Lille, Protéomique, Réponse Inflammatoire et Spectrométrie de Masse
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $3,989
Funder: Agence Nationale de la Recherche
OpenAlex ID: W4407295887