CARD*Shark: automated prioritization of literature curation for the Comprehensive Antibiotic Resistance Database

Database (Oxford). 2023 Apr 20;2023:baad023. doi: 10.1093/database/baad023.

ABSTRACT

Scientific literature is published at a rate that makes manual data extraction a highly time-consuming task. The Comprehensive Antibiotic Resistance Database (CARD) utilizes literature to curate information on antimicrobial resistance genes and to enable time-efficient triage of publications we have developed a classification algorithm for identifying publications describing first reports of new resistance genes. Trained on publications contained in the CARD, CARD*Shark downloads, processes and identifies publications recently added to PubMed that should be reviewed by biocurators. With CARD*Shark, we can minimize the monthly scope of articles a biocurator reviews from hundreds of articles to a few dozen, drastically improving the speed of curation while ensuring no relevant publications are overlooked. Database URL http://card.mcmaster.ca.

PMID:37079891 | PMC:PMC10118295 | DOI:10.1093/database/baad023