English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Graph embedding learning that aims to automatically learn low-dimensionalnode representations, has drawn increasing attention in recent years. To date,most recent graph embedding methods are evaluated on social and informationnetworks and are not comprehensively studied on biomedical networks undersystematic experiments and analyses. On the other hand, for a variety ofbiomedical network analysis tasks, traditional techniques such as matrixfactorization (which can be seen as a type of graph embedding methods) haveshown promising results, and hence there is a need to systematically evaluatethe more recent graph embedding methods (e.g. random walk-based and neuralnetwork-based) in terms of their usability and potential to further thestate-of-the-art.  We select 11 representative graph embedding methods and conduct a systematiccomparison on 3 important biomedical link prediction tasks: drug-diseaseassociation (DDA) prediction, drug-drug interaction (DDI) prediction,protein-protein interaction (PPI) prediction; and 2 node classification tasks:medical term semantic type classification, protein function prediction. Ourexperimental results demonstrate that the recent graph embedding methodsachieve promising results and deserve more attention in the future biomedicalgraph analysis. Compared with three state-of-the-art methods for DDAs, DDIs andprotein function predictions, the recent graph embedding methods achievecompetitive performance without using any biological features and the learnedembeddings can be treated as complementary representations for the biologicalfeatures. By summarizing the experimental results, we provide generalguidelines for properly selecting graph embedding methods and setting theirhyper-parameters for different biomedical tasks.

Graph Embedding on Biomedical Networks: Methods, Applications, and Evaluations