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 Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Motivation: Protein-protein interactions (PPI) are critical to the functionof proteins in both normal and diseased cells, and many critical proteinfunctions are mediated by interactions.Knowledge of the nature of theseinteractions is important for the construction of networks to analysebiological data. However, only a small percentage of PPIs captured in proteininteraction databases have annotations of function available, e.g. only 4% ofPPI are functionally annotated in the IntAct database. Here, we aim to labelthe function type of PPIs by extracting relationships described in PubMedabstracts.  Method: We create a weakly supervised dataset from the IntAct PPI databasecontaining interacting protein pairs with annotated function and associatedabstracts from the PubMed database. We apply a state-of-the-art deep learningtechnique for biomedical natural language processing tasks, BioBERT, to build amodel - dubbed PPI-BioBERT - for identifying the function of PPIs. In order toextract high quality PPI functions at large scale, we use an ensemble ofPPI-BioBERT models to improve uncertainty estimation and apply an interactiontype-specific threshold to counteract the effects of variations in the numberof training samples per interaction type.  Results: We scan 18 million PubMed abstracts to automatically identify 3253new typed PPIs, including phosphorylation and acetylation interactions, with anoverall precision of 46% (87% for acetylation) based on a human-reviewedsample. This work demonstrates that analysis of biomedical abstracts for PPIfunction extraction is a feasible approach to substantially increasing thenumber of interactions annotated with function captured in online databases.

Assigning function to protein-protein interactions: a weakly supervised BioBERT based approach using PubMed abstracts