PROKAR‐SEQ: an analysis and visualization framework for next‐generation sequencing based quantification of prokaryotic communities

Background: 16S ribosomal DNA (rDNA) based PCR followed by sequencing of these PCR fragments is the preferred method of quantifying prokaryotic microbial communities. Limitations of the sequencing technology (mainly the throughput) were until recently the limiting factor to quantify and compare large amounts of different samples or communities. However, recent developments in high-throughput sequencing techniques allow easier quantification of microbial communities and have led to a spectacular increase of insights into the composition and functionalities of microbial communities. The development of these recent sequencing techniques went hand in hand with the development of new analysis tools. In the last couple of years, several tools such as MOTHUR, PyroNoise/AmpliconNoise , VITCOMIC and Qiime have been developed to analyse pyrosequencing data sets and to assess microbial diversity. Methods: PROKAR-SEQ is a modular framework consisting of several different modules, which uses a MySQL database to retrieve raw data, to store intermediary data, and to store final analysis results. Final results are visualized by the visualization module, which fetches all required data from the database as well. Sequences are processed using AmpliconNoise and mapped onto the reference using BWA-SW. 16S rDNA reference sequences where downloaded from the Ribosomal Database Project (RDP) website. A complete taxonomic overview of all prokaryotes present in RDP was downloaded from the RDP website. Each taxonomic entry is quantified (using a custom perl pipeline) on each taxonomic level (fylum, order, genus…). An interactive interface was made using HTML/PHP and the Google Maps API. The web-based toolkit includes a prokaryote taxonomical browser, a geographical visualisation tool and various graphing tools. Results: For a user-selected taxonomic clade (domain, phylum, class, order, family or genus according to RDP), the relative frequency in each sample is visually depicted by a circle on the geographic location of the sample. For each sample, the rRNA content is visualized with a chart depicting the relative abundance of the selected taxonomic group from high to low. Thus users can compare different samples with each other (on the required taxonomic level) or analyze the composition of a single sample in detail. Conclusions: We developed PROCAR-SEQ, a tool to quantify and visualize microbial communities using GS-FLX next-generation sequencing (NGS) data. An example of the possibilities is available at http://athos.ugent.be/metagenom-x?p=map_test