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

Predicting the structure of interacting chains is crucial for understandingbiological systems and developing new drugs. Large-scale pre-trained ProteinLanguage Models (PLMs), such as ESM2, have shown impressive abilities inextracting biologically meaningful representations for protein structureprediction. In this paper, we show that ESMFold, which has been successful incomputing accurate atomic structures for single-chain proteins, can be adaptedto predict the heterodimer structures in a lightweight manner. We proposeLinker-tuning, which learns a continuous prompt to connect the two chains in adimer before running it as a single sequence in ESMFold. Experiment resultsshow that our method successfully predicts 56.98% of interfaces on the i.i.d.heterodimer test set, with an absolute improvement of +12.79% over theESMFold-Linker baseline. Furthermore, our model can generalize well to theout-of-distribution (OOD) test set HeteroTest2 and two antibody test sets Faband Fv while being $9\times$ faster than AF-Multimer.

Linker-Tuning: Optimizing Continuous Prompts for Heterodimeric Protein Prediction