A1 Vertaisarvioitu alkuperäisartikkeli tieteellisessä lehdessä
Multi-omics time-series analysis in microbiome research: a systematic review
Tekijät: Sherwani, Moiz Khan; Ruuskanen, Matti O.; Feldner-Busztin, Dylan; Nisantzis Firbas, Panos; Boza, Gergely; Móréh, Ágnes; Borman, Tuomas; Putu Erawijantari, Pande; Scheuring, István; Gopalakrishnan, Shyam; Lahti, Leo
Kustantaja: Oxford University Press (OUP)
Julkaisuvuosi: 2025
Lehti: Briefings in Bioinformatics
Artikkelin numero: bbaf502
Vuosikerta: 26
Numero: 5
ISSN: 1467-5463
eISSN: 1477-4054
DOI: https://doi.org/10.1093/bib/bbaf502
Julkaisun avoimuus kirjaamishetkellä: Avoimesti saatavilla
Julkaisukanavan avoimuus : Kokonaan avoin julkaisukanava
Verkko-osoite: https://doi.org/10.1093/bib/bbaf502
Rinnakkaistallenteen osoite: https://research.utu.fi/converis/portal/detail/Publication/504681535
Recent developments in data generation have opened up unprecedented insights into living systems. It has been recognized that integrating and characterizing temporal variation simultaneously across multiple scales, from specific molecular interactions to entire ecosystems, is crucial for uncovering biological mechanisms and understanding the emergence of complex phenotypes. With the increasing number of studies incorporating multi-omics data sampled over time, it has become clear that integrated approaches are pivotal for these efforts. However, standard data analytical practices in longitudinal multi-omics are still shaping up and many of the available methods have not yet been widely evaluated and adopted. To address this gap, we performed the first systematic literature review that comprehensively categorizes, compares, and evaluates computational methods for longitudinal multi-omics integration, with a particular emphasis on four categories of the studies: (i) host and host-associated microbiome studies, (ii) microbiome-free host studies, (iii) host-free microbiome studies, and (iv) methodological framework studies. Our review highlights current methodological trends, identifies widely used and high-performing frameworks, and assesses each method across performance, interpretability, and ease of use. We further organize these methods into thematic groups-such as statistical modeling, machine learning, dimensionality reduction, and latent factor approaches-to provide a clear roadmap for future research and application. This work offers a critical foundation for advancing integrative longitudinal data science and supporting reproducible, scalable analysis in this rapidly evolving field.
Ladattava julkaisu This is an electronic reprint of the original article. |  |
Julkaisussa olevat rahoitustiedot:
This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No 952914.
