Fig. 2From: RefSeq database growth influences the accuracy of k-mer-based lowest common ancestor species identificationFraction of correct species classifications (right) decreases in more recent RefSeq database versions and instead are classified at the genus level (left). Kraken classification results of simulated reads from known genomes against nine versions of the bacterial RefSeq database and the MiniKraken database. Misclassifications at the genus and species levels remain consistently low across database versionsBack to article page