# Curate this paper into an ORD dataset

Build an Open Reaction Database (ORD) dataset of the reactions reported in:

- **Paper:** Evaluating Predictive Accuracy in Asymmetric Catalysis: A Machine Learning Perspective on Local Reaction Space
- **Authors:** Isaiah O. Betinol, Aleksandra Demchenko, Jolene P. Reid
- **Published:** ACS Catalysis, 2025-03-31
- **DOI:** [10.1021/acscatal.5c01051](https://doi.org/10.1021/acscatal.5c01051)
- **ORD wishlist card:** [open on GitHub](https://github.com/orgs/open-reaction-database/projects/11?pane=issue&itemId=126976295)

The reference input `card` holds these facts, the source-data locations and the notes below in machine-readable form.

## Scope

Every reaction the paper reports with its data (the wishlist estimates about 10,000 reactions; literature review): inputs with amounts and roles, conditions, workups, outcomes and provenance, following the `ord-curation` Skill. Submit one ORD Dataset for the paper; if its data come from several screens or methods, keep them together and say so in the dataset description.

## Source data

Literature-derived asymmetric catalysis datasets compiled for ML benchmarking (University of British Columbia), ACS Catalysis 2025, DOI 10.1021/acscatal.5c01051. The SI zip (cs5c01051_si_002.zip on ACS figshare) contains spreadsheets for several case studies taken from earlier publications (files named Nature_2019, Science_2019, DD 2022, ACS_Catal_2024, Chem_Sci_2024, JACS_2023, Angew_2024), several with SMILES. The olefin-hydrogenation (Angew 2021) case study data are not included. Because the reactions come from other papers, each ORD dataset should cite the original publication(s) as well as this compilation. Split by source publication, and check whether any source set is already in ORD before curating it. The spreadsheets likely hold reaction components, catalysts and enantioselectivity (ee or ΔΔG‡); some also hold computed descriptors (e.g. Nature_DFT.xlsx), which should be excluded.

Known locations:

- [https://doi.org/10.1021/acscatal.5c01051](https://doi.org/10.1021/acscatal.5c01051): Article landing page (ACS Catalysis) (license: closed (OpenAlex); ACS TDM policy per Crossref)
- [https://acs.figshare.com/articles/dataset/Evaluating_Predictive_Accuracy_in_Asymmetric_Catalysis_A_Machine_Learning_Perspective_on_Local_Reaction_Space/28694944](https://acs.figshare.com/articles/dataset/Evaluating_Predictive_Accuracy_in_Asymmetric_Catalysis_A_Machine_Learning_Perspective_on_Local_Reaction_Space/28694944): SI zip cs5c01051_si_002.zip (10.8 MB), listed by range request. It holds a code repo whose repo/data/ contains Nature_2019.xlsx, Nature_SMILES.xlsx, Nature_DFT.xlsx, DD_data.xlsx, DD_reactants.xlsx, DD_out-of-bag/Set-1..3.csv, ACS_Catal_2024.xlsx, Chem_Sci_2024.xlsx, JACS_2023.xlsx… (license: CC BY-NC 4.0 (figshare label); MIT LICENSE.txt inside covers the repository software; the data derive from other publications)
- [https://acs.figshare.com/articles/journal_contribution/28694941](https://acs.figshare.com/articles/journal_contribution/28694941): SI PDF cs5c01051_si_001.pdf (18.7 MB) describing source references. Listing only. (license: CC BY-NC 4.0 (figshare label))

## Pitfalls

Split across roughly 7 source datasets. Outcomes are mainly enantioselectivity (ee or ΔΔG‡), not yields. Literature-compiled entries may lack full conditions. Descriptor columns must be separated from experimental data. The Angew_2024.csv file is large (24 MB). Some source sets may already be in ORD.

## Deliver

- `dataset`: the ORD Dataset, compressed (`.binpb.gz`, `.pb.gz`, `.txtpb.gz`, `.pbtxt.gz` or `.parquet`), citing the DOI above in every reaction's provenance, with no dataset or reaction IDs.
- `sources`: a `.zip` with a `README.md` (which tables or files each part came from, how structures were checked), the template and spreadsheet or the script or notebook that rebuild the dataset, and `validation.txt`, the output of ORD's validator (ord-schema 0.8.3) on the `dataset` file.

## After starting your Attempt

Open the wishlist card and check other Attempts. If the work is completed or already well underway, record the reason before ending your Attempt as abandoned. Get the source data only through legitimate channels (the publisher's Supporting Information, the repositories above, or the authors). Submitting the finished dataset to ORD, and ORD's separate Dataset Rewards program, follow ORD's own rules.


## Evaluation and public curator identity

TFP first runs the pinned automatic ORD validator on the exact dataset file, then requires two independent agent reviews; both must pass. Review agents follow the required ord-curation Skill and review.md, use the published model policy, inspect primary sources and independently rebuild the result. Warnings and scientific ambiguities require review even when validation passes. ORD maintainers decide final acceptance separately.

For work commissioned by Tokens for Progress, use public curator name `Tokens for Progress` and email `hello@tokensforprogress.org`; independent contributors use their own confirmed public identity. Do not use reserved example email addresses or fabricate contact details. Include a source coverage manifest and disclose all unresolved data.