The dysbiosis of oral microbiome (OM) precedes the clinical signs of periodontal disease. Its simple measure thus could indicate individuals at risk of periodontitis development; however, such a tool is still missing. Up to now, numerous microbial taxa were associated with periodontal health or periodontitis. The outputs of most studies could, nevertheless, be slightly biased from following two reasons: First, the healthy group is often characterized only by the absence of the disease, but the individuals could already suffer from dysbiosis without any visible signs. Second, the healthy/diseased OM characteristics are frequently determined based on average data obtained for whole groups of periodontally healthy persons versus patients. Especially in smaller sets of tested individuals the typical individual variability can thus complicate the unambiguous assignment of oral taxa to respective state of health. In this work the taxonomic composition of OM was evaluated for 20 periodontally healthy individuals and 15 patients with chronic periodontitis. The narrowed selection set of the most diseased patients (confirmed by clinical parameters) and the most distant group of healthy individuals with the lowest probability of dysbiosis was determined by clustering analysis and used for identification of marker taxa. Based on their representation in each individual oral cavity we proposed the numeric index of periodontal health called R/G value. Its diagnostic potential was further confirmed using independent set of 20 periodontally healthy individuals and 20 patients with periodontitis with 95 percent of samples assigned correctly. We also assessed the individual temporal OM dynamics in periodontal health and we compared it to periodontitis. We revealed that the taxonomic composition of the system changes dynamically but generally it ranges within values typical for periodontal health or transient state, but far from values typical for periodontitis. R/G value tool, formulated from individually evaluated data, allowed us to arrange individual OMs into a continuous series, instead of two distinct groups, thus mimicking the gradual transformation of a virtual person from periodontal health to disease. The application of R/G value index thus represents a very promising diagnostic tool for early prediction of persons at risk of developing periodontal disease.

A robust and widely applicable method for sampling of aquatic microbial biofilm and further sample processing is presented. The method is based on next-generation sequencing of V4-V5 variable regions of 16S rRNA gene and further statistical analysis of sequencing data, which could be useful not only to investigate taxonomic composition of biofilm bacterial consortia but also to assess aquatic ecosystem health. Five artificial materials commonly used for biofilm growth (glass, stainless steel, aluminum, polypropylene, polyethylene) were tested to determine the one giving most robust and reproducible results. The effect of used sampler material on total microbial composition was not statistically significant; however, the non-plastic materials (glass, metal) gave more stable outputs without irregularities among sample parallels. The bias of the method is assessed with respect to the employment of a non-quantitative step (PCR amplification) to obtain quantitative results (relative abundance of identified taxa). This aspect is often overlooked in ecological and medical studies. We document that sequencing of a mixture of three merged primary PCR reactions for each sample and further evaluation of median values from three technical replicates for each sample enables to overcome this bias and gives robust and repeatable results well distinguishing among sampling localities and seasons.