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

Advances in next-generation metagenome sequencing have the potential torevolutionize the point-of-care diagnosis of novel pathogen infections, whichcould help prevent potential widespread transmission of diseases. Given thehigh volume of metagenome sequences, there is a need for scalable frameworks toanalyze and segment metagenome sequences from clinical samples, which can behighly imbalanced. There is an increased need for learning robustrepresentations from metagenome reads since pathogens within a family can havehighly similar genome structures (some more than 90%) and hence enable thesegmentation and identification of novel pathogen sequences with limitedlabeled data. In this work, we propose Metagenome2Vec - a contextualizedrepresentation that captures the global structural properties inherent inmetagenome data and local contextualized properties through self-supervisedrepresentation learning. We show that the learned representations can helpdetect six (6) related pathogens from clinical samples with less than 100labeled sequences. Extensive experiments on simulated and clinical metagenomedata show that the proposed representation encodes compositional propertiesthat can generalize beyond annotations to segment novel pathogens in anunsupervised setting.

Metagenome2Vec: Building Contextualized Representations for Scalable Metagenome Analysis