ARGminer
ARGminer facilitates curation and validation of antibiotic resistance genes (ARGs) by combining sequence alignment evidence and crowdsourced annotations to produce comprehensive, validated ARG annotations.
Key Features:
- Multi-Level Annotation: Supports annotation at gene name, antibiotic category, resistance mechanism, and evidence for mobility and occurrence in clinically-important bacterial strains.
- Sequence-alignment Evidence Aggregation: Aggregates evidence from sequence alignment and multiple sources to inform annotations.
- Crowdsourcing Strategy: Employs crowdsourcing to expand curation capacity and generate community curator annotations.
- Trust Validation Filter: Applies a trust validation filter to reject confounding or spam inputs and improve reliability of curated data.
- Validation against Expert Curation: Validated across multiple curator cohorts, achieving annotation accuracy greater than 90% compared with expert annotation.
- Comparative Efficiency: Demonstrated to be more cost-effective and less time-consuming than traditional expert curation while maintaining high accuracy.
Scientific Applications:
- ARG Database Curation: Curation and validation of comprehensive ARG databases using aggregated evidence and crowdsourced annotations.
- ARG Attribute Annotation: Annotation of gene name, antibiotic category, resistance mechanism, mobility evidence, and occurrence in clinically-important bacterial strains.
- Annotation Benchmarking: Comparison of crowdsourced annotations to expert curation to assess and quantify annotation accuracy.
Methodology:
Combines sequence alignment and aggregation of evidence from multiple sources, crowdsourced curator annotations, a trust validation filter to reject confounding inputs, and validation against expert curation cohorts.
Topics
Details
- Tool Type:
- web application
- Added:
- 1/18/2021
- Last Updated:
- 1/28/2021
Operations
Publications
Arango-Argoty GA, Guron GKP, Garner E, Riquelme MV, Heath LS, Pruden A, Vikesland PJ, Zhang L. ARGminer: a web platform for the crowdsourcing-based curation of antibiotic resistance genes. Bioinformatics. 2020;36(9):2966-2973. doi:10.1093/bioinformatics/btaa095. PMID:32058567.
PMID: 32058567
Funding: - NIFA: 2015-68003-2305, 2017-68003-26498
- PIRE: 1545756