The 16S rRNA gene is present in all prokaryotes, and consists of highly conserved regions interspersed by nine ‘hypervariable’ regions. Hypervariable regions are of variable sequence and length between different bacteria.

As such, PCR amplification and DNA sequencing of different bacterial 16S rRNA genes within a biofluid can allow identification of the different bacteria within samples.

This is now a well-established, high-throughput technique, which is relatively cheap compared with alternatives.

This technique may give detail down to bacterial genus level, but rarely gives any greater resolution than this.

There are a number of factors that can bias the results obtained, for example:

• There is variable use of primers, PCR conditions and analytic approach between different centres.

• Different bacteria have different copy numbers of the 16S rRNA gene, influencing the apparent relative abundance of certain bacteria within a sample.

This technique involves fragmentation and random sequencing of DNA from the collection of genomes and genes within the sample, and the use of advanced bioinformatic techniques to analyse the sequencing data obtained.

Shotgun sequencing regularly gives characterisation down to species (and even possibly strain) level resolution.

Since this technique sequences bacterial genes, it also gives insight into the microbial functionality of a sample.

Gaps in knowledge in the accuracy and completeness of reference microbial genomic databases means large portions of sequencing data may be difficult to interpret.

Datasets are large and complex, and experiments are much more expensive than metataxonomics.

As per metataxonomics, there are still outstanding issues related to factors which may bias results.