Fine-Tuning of DADA2 Parameters for Multiregional Metabarcoding Analysis of 16S rRNA Genes from Activated Sludge and Comparison of Taxonomy Classification Power and Taxonomy Databases

Taxonomic classification using metabarcoding is a commonly used method in microbiological studies of environmental samples and during monitoring of biotechnological processes. However, it is difficult to compare results from different laboratories, due to the variety of bioinformatics tools that have been developed and used for data analysis. This problem is compounded by different choices regarding which variable region of the 16S rRNA gene and which database is used for taxonomic identification. Therefore, this study employed the DADA2 algorithm to optimize the preprocessing of raw data obtained from the sequencing of activated sludge samples, using simultaneous analysis of three frequently used regions of 16S rRNA (V1-V3, V3-V4, V4-V5). Additionally, the study evaluated which variable region and which of the frequently used microbial databases for taxonomic classification (Greengenes2, Silva, RefSeq) more accurately classify OTUs into taxa. Adjusting the values of selected parameters of the DADA2 algorithm, we obtained the highest possible numbers of OTUs for each region. Regarding biodiversity within regions, the V3-V4 region had the highest Simpson and Shannon indexes, and the Chao1 index was similar to that of the V1-V3 region. Beta-biodiversity analysis revealed statistically significant differences between regions. When comparing databases for each of the regions studied, the highest numbers of taxonomic groups were obtained using the SILVA database. These results suggest that standardization of metabarcoding of short amplicons may be possible.