The Comprehensive Antibiotic Resistance Database (CARD; http://arpcard.mcmaster.ca) is a manually curated resource containing high quality reference data on the molecular basis of antimicrobial resistance (AMR), with an emphasis on the genes, proteins and mutations involved in AMR. CARD is ontologically structured, model centric, and spans the breadth of AMR drug classes and resistance mechanisms, including intrinsic, mutation-driven and acquired resistance. It is built upon the Antibiotic Resistance Ontology (ARO), a custom built, interconnected and hierarchical controlled vocabulary allowing advanced data sharing and organization. Its design allows the development of novel genome analysis tools, such as the Resistance Gene Identifier (RGI) for resistome prediction from raw genome sequence. Recent improvements include extensive curation of additional reference sequences and mutations, development of a unique Model Ontology and accompanying AMR detection models to power sequence analysis, new visualization tools, and expansion of the RGI for detection of emergent AMR threats. CARD curation is updated monthly based on an interplay of manual literature curation, computational text mining, and genome analysis.

The Comprehensive Antibiotic Resistance Database (CARD; card.mcmaster.ca) combines the Antibiotic Resistance Ontology (ARO) with curated AMR gene (ARG) sequences and resistance-conferring mutations to provide an informatics framework for annotation and interpretation of resistomes. As of version 3.2.4, CARD encompasses 6627 ontology terms, 5010 reference sequences, 1933 mutations, 3004 publications, and 5057 AMR detection models that can be used by the accompanying Resistance Gene Identifier (RGI) software to annotate genomic or metagenomic sequences. Focused curation enhancements since 2020 include expanded β-lactamase curation, incorporation of likelihood-based AMR mutations for Mycobacterium tuberculosis, addition of disinfectants and antiseptics plus their associated ARGs, and systematic curation of resistance-modifying agents. This expanded curation includes 180 new AMR gene families, 15 new drug classes, 1 new resistance mechanism, and two new ontological relationships: evolutionary_variant_of and is_small_molecule_inhibitor. In silico prediction of resistomes and prevalence statistics of ARGs has been expanded to 377 pathogens, 21,079 chromosomes, 2,662 genomic islands, 41,828 plasmids and 155,606 whole-genome shotgun assemblies, resulting in collation of 322,710 unique ARG allele sequences. New features include the CARD:Live collection of community submitted isolate resistome data and the introduction of standardized 15 character CARD Short Names for ARGs to support machine learning efforts.

The International Pseudomonas aeruginosa Consortium is sequencing over 1000 genomes and building an analysis pipeline for the study of Pseudomonas genome evolution, antibiotic resistance and virulence genes. Metadata, including genomic and phenotypic data for each isolate of the collection, are available through the International Pseudomonas Consortium Database (http://ipcd.ibis.ulaval.ca/). Here, we present our strategy and the results that emerged from the analysis of the first 389 genomes. With as yet unmatched resolution, our results confirm that P. aeruginosa strains can be divided into three major groups that are further divided into subgroups, some not previously reported in the literature. We also provide the first snapshot of P. aeruginosa strain diversity with respect to antibiotic resistance. Our approach will allow us to draw potential links between environmental strains and those implicated in human and animal infections, understand how patients become infected and how the infection evolves over time as well as identify prognostic markers for better evidence-based decisions on patient care.

Metagenomic methods enable the simultaneous characterization of microbial communities without time-consuming and bias-inducing culturing. Metagenome-assembled genome (MAG) binning methods aim to reassemble individual genomes from this data. However, the recovery of mobile genetic elements (MGEs), such as plasmids and genomic islands (GIs), by binning has not been well characterized. Given the association of antimicrobial resistance (AMR) genes and virulence factor (VF) genes with MGEs, studying their transmission is a public-health priority. The variable copy number and sequence composition of MGEs makes them potentially problematic for MAG binning methods. To systematically investigate this issue, we simulated a low-complexity metagenome comprising 30 GI-rich and plasmid-containing bacterial genomes. MAGs were then recovered using 12 current prediction pipelines and evaluated. While 82–94 % of chromosomes could be correctly recovered and binned, only 38–44 % of GIs and 1–29 % of plasmid sequences were found. Strikingly, no plasmid-borne VF nor AMR genes were recovered, and only 0–45 % of AMR or VF genes within GIs. We conclude that short-read MAG approaches, without further optimization, are largely ineffective for the analysis of mobile genes, including those of public-health importance, such as AMR and VF genes. We propose that researchers should explore developing methods that optimize for this issue and consider also using unassembled short reads and/or long-read approaches to more fully characterize metagenomic data.

MotivationMetagenomic methods have emerged as a key tool in public-health microbiology for surveillance of virulence factor (VF) and antimicrobial resistance (AMR) genes. However, metagenomic data, even when assembled, typically results in complex, mixed sets DNA sequence fragments rather than fully resolved individual genomes. Recently, metagenome-assembled genomes (MAGs) have emerged as a promising approach that groups sequences into bins that are likely derived from the same underlying genome. However, MAGs have not been well assessed for their ability to identify some of the key sequences of interest for infectious disease surveillance purposes: AMR and VFs associated with mobile genetic elements (MGEs) such as plasmids and genomic islands (GIs). We hypothesized that due to the di erent copy number and sequence composition of plasmids and GIs compared to core genome sequence, such sequences will be under-represented in MAG-based approaches. ResultsTo evaluate the impact of MAG recovery methods on recovery of AMR genes and MGEs, we generated a simulated metagenomic dataset comprised of 30 genomes with up to 16.65% of the chromosomal DNA consisting of GIs and 65 associated plasmids. MAGs were then recovered from this data using 12 di erent MAG pipelines and evaluated for recovery accuracies. Across all pipelines, 81.9-94.3% of chromosomes were recovered and binned. However, only 37.8-44.1% of GIs and 1.5-29.2% of plasmids were recovered and correctly binned at >50% coverage. In terms of AMR and VF genes associated with MGEs, 0-45% of GI-associated AMR genes and 0-16% of GI-associated VF genes were correctly assigned. More strikingly, 0% of plasmid-borne VF or AMR genes were recovered. This work shows that regardless of the MAG recovery approach used, plasmid and GI dominated sequences will disproportionately be left unbinned or incorrectly binned. From a public-health perspective, this means MAG approaches are less suited for analysis of mobile genes, especially key groups such as AMR and VF genes. This underlines the utility of read-based and long-read approaches to thoroughly evaluate the resistome in metagenomic data.