Fig. 1

HumGut overview. HumGut represents a collection of genomes and MAGs contained in 3,534 healthy human gut metagenomes. To be considered as contained, a genome shared at least 0.95 sequence identity with at least one of the metagenomes (inferred by the number of shared hashes). The qualified genomes were scored based on the average sequence identity across all the metagenomes. Next, they were ranked based on their scores: the higher the score, the higher the position on the list. Subsequently, the genomes were clustered based on MASH and fastANI distance (D). The top-ranked genome formed a cluster centroid. Around 30,600 clusters were formed applying a D = 0.025-threshold. The use of HumGut as a reference set helps the process of taxonomic assignments by drastically reducing the number of unclassified human gut metagenomic reads