Fig. 7

Read classification of skin and stool mWGS samples. a The proportion of skin (left) and stool (right) reads that are classified using reprDB, panDB, and a previous database described in [4]. b The most abundant microbial groups in selected skin (upper) and stool (lower) samples based on reprDB and panDB. Microbial species are grouped by their lowest taxonomic level at or above the genus level. Tentative classifications and misplaced classifications that are waiting for revisions are shown in square brackets according to the convention of the NCBI Taxonomy database [40]. Uncultured species are classified into an independent group [41]. c The proportion of reads mapped to the pan-genome or all sequenced strain genomes of selected stool (upper) and skin (lower) species, normalized to the proportion of reads mapped to the representative genome(s) of the same species (dashed line). d The proportions of reads aligned to different genera (right) or species (left) based on reprDB (x-axis) and panDB (y-axis). Results are shown for a randomly selected skin sample (MET0284, upper) and a random stool sample (SRS011061, lower). In the upper left panel, three different types of species that recruit inconsistent proportion of reads based on reprDB and panDB are shown in colored boxes. Blue box: species with increased genome space in panDB; green box: species with reads re-assigned to a different genome in panDB; orange box: species that are present in panDB but not in reprDB