COVID-19 Knowledge Extractor (COKE): A Tool and a Web Portal to Extract Drug - Target Protein Associations from the CORD-19 Corpus of Scientific Publications on COVID-19

• Published in: ChemRxiv
• Main Authors: Korn, Daniel, Pervitsky, Vera, Bobrowski, Tesia, Alves, Vinicius M, Schmitt, Charles, Bizon, Chris, Baker, Nancy, Chirkova, Rada, Cherkasov, Artem, Muratov, Eugene, Tropsha, Alexander
• Format: Journal Article Paper
• Language: English
• Published: United States 26.11.2020
• Edition: 1
• Subjects: Artificial Intelligence; Chemistry; Chemoinformatics - Computational Chemistry; COVID-19; COVID-19 data; Drug-target-disease associations; text mining; CORD-19; knowledge acquisition
• ISSN: 2573-2293
• DOI: 10.26434/chemrxiv.13289222.v1

The COVID-19 pandemic has catalyzed a widespread effort to identify drug candidates and biological targets of relevance to SARS-COV-2 infection, which resulted in large numbers of publications on this subject. We have built the VID-19 nowledge xtractor (COKE), a web application to extract, curate, and annotate essential drug-target relationships from the research literature on COVID-19 to assist drug repurposing efforts. SciBiteAI ontological tagging of the COVID Open Research Dataset (CORD-19), a repository of COVID-19 scientific publications, was employed to identify drug-target relationships. Entity identifiers were resolved through lookup routines using UniProt and DrugBank. A custom algorithm was used to identify co-occurrences of protein and drug terms, and confidence scores were calculated for each entity pair. COKE processing of the current CORD-19 database identified about 3,000 drug-protein pairs, including 29 unique proteins and 500 investigational, experimental, and approved drugs. Some of these drugs are presently undergoing clinical trials for COVID-19. The rapidly evolving situation concerning the COVID-19 pandemic has resulted in a dramatic growth of publications on this subject in a short period. These circumstances call for methods that can condense the literature into the key concepts and relationships necessary for insights into SARS-CoV-2 drug repurposing. The COKE repository and web application deliver key drug - target protein relationships to researchers studying SARS-CoV-2. COKE portal may provide comprehensive and critical information on studies concerning drug repurposing against COVID-19. COKE is freely available at https://coke.mml.unc.edu/ and the code is available at https://github.com/DnlRKorn/CoKE .