AMRFinderPlus and CARD tools

Question

[Baka]

Hello Biostars,

I want to use AMRFinderPlus (https://github.com/ncbi/amr/wiki) and CARD (https://card.mcmaster.ca/) tools for my research and have below questions:

How are the AMRFinderPlus and CARD different from each other for predication of AMR genes from bacterial genomic sequences?

How much overlap do AMRFinderPlus and CARD database have?

If anyone working in this area, it would be very helpful to get some information.

Thank you.

[Ram]

Baka why did you delete this post?

Answer 1

[LChart]

The most salient difference is that CARD requires a license for commercial use, while AMRFinderPlus is Public. However, according to a recent benchmark (https://www.medrxiv.org/content/10.1101/2022.08.11.22278689v2.full) you are probably better off with ARDaP.

Note especially this from the benchmark: "Due to its inability to predict AMR towards individual antibiotics, and a very high rate of false-positive predictions in the Global Dataset, CARD was deemed unsuitable for P. aeruginosa AMR analysis and was excluded from further assessment"

CARD is somewhat nicer because it's hand-curated and organized; but it's not really a substitute for a dedicated AMR finding and scoring tool.