Fig. 3
From: HumGut: a comprehensive human gut prokaryotic genomes collection filtered by metagenome data
An overview of the genomes used to build HumGut. a The pie charts show the proportion of genomes from each collection (UHGG above, RefSeq below) included in HumGut. To qualify for HumGut inclusion, genomes had to have at least 0.95 MASH screen identity with at least one healthy metagenome, as did most of the UHGG and half of the RefSeq genomes. Histograms show the distribution of the mean identity shared between the qualified genomes and healthy metagenomes. A high average identity means that the qualified genome has been found contained in most of the screened samples. b The genome sources for HumGut clusters. The upper pie chart shows data for 30,691 clusters belonging to HumGut_97.5 (genomes grouped based on 97.5% genome sequence identity); the bottom one presents data for 5,170 HumGut_95 clusters (95% sequence identity—species level threshold). The majority of clusters in both HumGut collections are comprised of only UHGG genomes, while 6% and 9% of the clusters consist of only RefSeq genomes (HumGut_97.5 and HumGut_95, respectively)