Figure 1: Bacterial community composition at family level of human stool samples analysed in this study.

The bacterial community profiles are based on operational taxonomic unit (OTU, defined at 97% genetic identity) frequency in stool samples of 44 patients with diarrhea indicative of C. difficile infection and 35 asymptomatic control individuals (n=79). One stool sample per patient was used and amplicon PCRs were performed in triplicate for this analysis. Families, which exhibited an abundance of lower than 1% in the entire dataset, were summarized as rare taxa. Relative abundance of C. difficile (Peptoclostridium difficile in SILVA database 123.1) is displayed separately and exhibited highest similarity to Clostridioides difficile strain 630 delta erm (Accession number CP016318). Occurrence of diarrhea in patents is indicated by plus (patient exhibited diarrhea) and minus (no diarrhea), results from microbiological diagnosis of C. difficile infection (C. d. m. t.) are shown below (plus, positively tested for C. difficile; minus, negatively tested for C. difficile). Presence and absence of C. difficile in amplicon data (C. d. NGS) are indicated by plus (present) and minus (absent). Data processing and employed tools are described in detail in the methods section.