Fig. 4: Performance evaluation with sequenced mock communities.

Expected and estimated ARG abundances (cpg). Sample statistics, including the total number of genomes (no. gen.), mean quality scores (q), and mean read lengths (l), are shown in the upper left tables. Precision (pr.) and recall (re.) for ARG types (typ.) and subtypes (sub.) are displayed in the lower right tables. Dashed gray lines indicate 1:1 ratio lines. Species names were manually adjusted to reflect updates in taxonomy and resolve inconsistencies between NCBI and GTDB. Schaalia odontolytica was renamed to Pauljensenia odontolytica (MSA1003), Lactobacillus fermentum to Limosilactobacillus fermentum (D6300 and D6331), Bacillus subtilis to Bacillus spizizenii (D6300), Fusobacterium nucleatum to Fusobacterium animalis (D6331), and Clostridium perfringens to Sarcina perfringens (D6331). MLS: macrolide-lincosamide-streptogramin.