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

Machine learning-assisted retrosynthesis prediction models have been gainingwidespread adoption, though their performances oftentimes degrade significantlywhen deployed in real-world applications embracing out-of-distribution (OOD)molecules or reactions. Despite steady progress on standard benchmarks, ourunderstanding of existing retrosynthesis prediction models under the premise ofdistribution shifts remains stagnant. To this end, we first formally sort outtwo types of distribution shifts in retrosynthesis prediction and construct twogroups of benchmark datasets. Next, through comprehensive experiments, wesystematically compare state-of-the-art retrosynthesis prediction models on thetwo groups of benchmarks, revealing the limitations of previous in-distributionevaluation and re-examining the advantages of each model. More remarkably, weare motivated by the above empirical insights to propose two model-agnostictechniques that can improve the OOD generalization of arbitrary off-the-shelfretrosynthesis prediction algorithms. Our preliminary experiments show theirhigh potential with an average performance improvement of 4.6%, and theestablished benchmarks serve as a foothold for further retrosynthesisprediction research towards OOD generalization.

RetroOOD: Understanding Out-of-Distribution Generalization in Retrosynthesis Prediction