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

The ability to accurately model the fitness landscape of protein sequences iscritical to a wide range of applications, from quantifying the effects of humanvariants on disease likelihood, to predicting immune-escape mutations inviruses and designing novel biotherapeutic proteins. Deep generative models ofprotein sequences trained on multiple sequence alignments have been the mostsuccessful approaches so far to address these tasks. The performance of thesemethods is however contingent on the availability of sufficiently deep anddiverse alignments for reliable training. Their potential scope is thus limitedby the fact many protein families are hard, if not impossible, to align. Largelanguage models trained on massive quantities of non-aligned protein sequencesfrom diverse families address these problems and show potential to eventuallybridge the performance gap. We introduce Tranception, a novel transformerarchitecture leveraging autoregressive predictions and retrieval of homologoussequences at inference to achieve state-of-the-art fitness predictionperformance. Given its markedly higher performance on multiple mutants,robustness to shallow alignments and ability to score indels, our approachoffers significant gain of scope over existing approaches. To enable morerigorous model testing across a broader range of protein families, we developProteinGym -- an extensive set of multiplexed assays of variant effects,substantially increasing both the number and diversity of assays compared toexisting benchmarks.

Tranception: protein fitness prediction with autoregressive transformers and inference-time retrieval