OPTIMIZATION OF MOLECULAR AND BIOINFORMATIC TECHNIQUES FOR AQUATIC MICROBIOME STUDIES

From sample collection andwater filtration through 16S rRNA and metagenomic analyses, different methods and protocols are used in the study of microbial communities. The methods selected are often based on the researcher’s experience, literature or simply on gut feeling. In this study, we explore to what extent methodological choices affect the observed microbiome structure and function. Three aquatic environments were used in this study to cover a wide range of aquatic communities: (i) a reservoir connected to aquaculture facilities; (ii) a bloom of Microcystis spp. in Lake Kinneret, and (iii) coastal Mediterranean bacterioplankton. We describe the impact of: a) mesh filter size and its ability to differentiate between free-living and particle-associated bacteria; b) different DNA and RNA extraction methods (including those optimized for use in a robotic system to reduce person-to-person bias); c) different 16S analysis pipelines; and d) different metagenomic library preparation protocols (insert size, size selection) and sequencing depth. Our results suggest that filter size but not extraction method could significantly affect the microbial community composition and diversity. In contrast, the effect of different library preparation techniques was less pronounced. Results of the same samples processed by different lab members were similar, while the time between sample collection and filtration impacted estimates of microbial community structure. Our results show that methodological choices, which typically receive limited attention, are important for producing representative and reproducible aquatic microbiome studies.