Benchmarking Machine Learning for Therapeutic Deep Eutectic Solvent Screening: Reviewer Data and Code Package

Engineering notes

Engineering notes will be added by the aipentium editorial team.

Chinese explanation / 中文解读

中文解读待补充：本站会优先为大语言模型、生成式AI、ChatGPT相关技术、计算机视觉、深度学习等高价值论文补充中文说明。

Original abstract

Confidential reviewer-access package accompanying the manuscript “Benchmarking Machine Learning for Therapeutic Deep Eutectic Solvent Screening: Dataset Construction, Cross-Source Generalization Failure, and Applicability-Boundary Analysis”, prepared for submission to Digital Discovery. The package contains the curated 621-record THEDES dataset, descriptor-enriched data, the final n=340 random-forest benchmark model and metadata, frozen blind and prospective score vectors, analysis and figure-generation scripts, publication-v3 figures, Supporting Information, and machine- readable outputs underlying the reported results. The historical fitted object that generated the June 2026 external score vectors was not retained; those archived score vectors are preserved and analysed unchanged. This draft is provided solely for confidential peer review. No public redistribution or reuse licence is granted at this stage. A versioned public archive with author-approved licences and a persistent DOI will be released in coordination with publication.

Links and sources

Comments