Automatic predictions of Drug-Protein relations database¶

This file repository contains our model predictions of Drug-Protein relations on DrugProt Large track test dataset.¶


Yoon, Wonjin, Sean Yi, Richard Jackson, Hyunjae Kim, Sunkyu Kim, and Jaewoo Kang.
Biomedical relation extraction with knowledge base refined weak-supervision (Submitted to DATABASE). 2022
* Presented under the name Using Knowledge Base to Refine Data Augmentation for Biomedical Relation Extraction at the BioCreative VII workshop

Source code repository: Here¶

Resource link:¶

Predictions: http://nlp.dmis.korea.edu/projects/drugprot-yoon-et-al-2022/runID2-predictions.tar.gz
(105M) SHA1SUM fbce7aab9e9a370f59ceb62f316e9ebd22f8e7f7

Test dataset (includes plain texts): http://nlp.dmis.korea.edu/projects/drugprot-yoon-et-al-2022/test_large.tar.gz
(2.1G) SHA1SUM f5c52b621ed25555a9eccbbf2aa4e5cdaad5f892

Citation info:¶

Our model:

Yoon, Wonjin, Sean Yi, Richard Jackson, Hyunjae Kim, Sunkyu Kim, and Jaewoo Kang. Biomedical relation extraction with knowledge base refined weak-supervision. (Submitted to DATABASE). 2022

Dataset:

Antonio Miranda, Farrokh Mehryary, Jouni Luoma, SampoPyysalo, Alfonso Valencia, and Martin Krallinger. Overview of  drugprot  biocreative  vii  track:  quality  evaluation  andlarge  scale  text  mining  of  drug-gene/protein  relations.   InProceedings of the seventh BioCreative challenge evaluation workshop, 2021.

Reference:¶

This work partially used Comparative Toxicogenomics Database (CTD). We appreciate their contribution towards BioNLP/Bioinformatics society :

Patrick  Verga,  Emma  Strubell,  and  Andrew  McCallum.Simultaneously   self-attending   to   all   mentions   for   full-abstract  biological  relation  extraction.InProceedingsof  the  2018  Conference  of  the  North  American  Chapterof  the  Association  for  Computational  Linguistics:  HumanLanguage  Technologies,   Volume  1  (Long  Papers),  pages872–884,  New  Orleans,  Louisiana,  June  2018.  Association for Computational Linguistics.

Contacts:¶

For other inquiries, please contact


Wonjin Yoon
<my first name>.info @ gmail.com

Author's note¶

Users of our resources must understand that resources we share, including automatic annotations andmodels, are not designed for real-world use nor reviewed from domain experts. Using the resource should be at your own risks.

License¶

TBA: Will be one of open source license