Table 1.
Overview of datasets for chemical NLP.
| Year | Dataset | Documents | Chemical, protein, and gene mentions | Unique IDs | Relations | Application |
|---|---|---|---|---|---|---|
| 2008 | Corbett32 | 500 abstracts, 42 papers | 11,571 | — | — | NER |
| 2008 | SCAI33 | 100 abstracts | 1,206 | — | — | NER |
| 2012 | ADE39 | 300 case reports | 5,063 drugs | — | 6,821 drug adverse effects 279 drug dosage | RE |
| 2013 | DDI43 | 1,025, including texts from DrugBank and abstracts | 18,502 drugs | — | 5,028 drug-drug interactions | RE |
| 2015 | CHEMDNER34 | 10,000 abstracts | 84,355 chemicals | — | — | NER |
| 2016 | BC5CDR35 | 1,500 articles | 15,935 chemicals 12,850 diseases | 4,409 MeSH | chemically induced diseases | NER, NEN, RE |
| 2017 | N-ary drug-gene-mutation42 | — | — | — | 137,469 drug–gene 3,192 drug–mutation | RE |
| 2017 | ChemProt40 | 2,482 abstracts | 32,514 chemicals 30,922 genes | — | chemical-protein | RE |
| 2019 | DrugProt41 | 5,000 abstracts | 65,561 chemicals 61,775 proteins | — | 24,526 chemical-protein | RE |
| 2020 | EBED38 | 4,200, including abstracts, paragraphs, figure legends, and patents | 16,715 chemicals 56,059 genes | 5,161 ChEBI 12,563 Entrez | chemically induced diseases | NER, NEN, RE |
| 2021 | ChEMU 202044 | 1,500 patent extracts | 17,834 chemicals | — | chemical reaction steps | NER, RE |
| 2022 | BioRED37,75 | 1,000 abstracts | 7,021 chemicals 12,412 genes | 1,096 MeSH 2,605 Entrez | chemical-(chemical/disease/gene/variant) | NER, NEN, RE |
| 2024 | EnzChemRED (this work) | 1,210 abstracts | 18,887 chemicals 13,028 proteins | 3,155 ChEBI 2,569 UniProtKB | chemical-chemical and (chemical-chemical)-protein | NER, NEN, RE |