EleMi: A Robust Method to Infer Soil Ecological Networks with Better Community Structure

Chen, Nan; Bucur, Doina

doi:10.1007/978-3-031-57515-0_13

Cited by 1 publication

(1 citation statement)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Firstly, we want to provide the user with different options to normalize their data using methods that take compositionality into account such as CLR and ILR transformations. Secondly, we would like to implement more advanced network inference methods such as SparCC (Kishore et al, 2023) and EleMi (Chen & Bucur, 2024). Thirdly, we want to implement different subcommunity detection algorithms as alternatives to k-means (e.g.…”

Section: Discussionmentioning

confidence: 99%

SpeSpeNet: An interactive and user-friendly tool to create and explore microbial correlation networks

van Eijnatten,

van Zon,

Manousou

et al. 2024

Preprint

View full text Add to dashboard Cite

Correlation networks are commonly used to explore microbiome data. In these networks, nodes are taxa and edges represent correlations between their abundance patterns across samples. As clusters of correlating taxa (co-abundance clusters) often indicate a shared response to environmental drivers, network visualization contributes to system understanding. Currently, most tools for creating and visualizing co-abundance networks from microbiome data either require the researcher to have coding skills, or they are not user-friendly, with high time expenditure and limited customizability. Furthermore, existing tools lack focus on the relationship between environmental drivers and the structure of the microbiome, even though many edges in correlation networks can be understood through a shared relationship of two taxa with the environment. For these reasons we developed SpeSpeNet (Species-Species Network, https://tbb.bio.uu.nl/SpeSpeNet), a practical and user-friendly R-shiny tool to construct and visualize correlation networks from taxonomic abundance tables. The details of data preprocessing, network construction, and visualization are automated, require no programming ability for the web version, and are highly customizable, including associations with user-provided environmental data. Here, we present the details of SpeSpeNet and demonstrate its utility using three case studies.

show abstract

Section: Discussionmentioning

confidence: 99%