Table 2.

High-throughput sequencing platforms and their potential metagenomic applications in public health

Manufacturer	Platform	Read length (bp)	Output	Run time	Reads	Advantages	Limitations	Applications
Second generation sequencers
Ilumina	HiSeq 2500	36-150	600 Gb	11 days	2.4 × 109	Very high depth	Long run time	High sensitivity to detect pathogens that are present at very low concentrations in metagenomic samples
						Low error rate	Short read lengths
						Lower cost per bp	Errors in regions following GGC motifs [65]
						Paired-end reads	Decreasing read quality toward ends [66]
	MiSeq	36-250	8.5 Gb	39 hours	34 × 106	Desktop machine	Short read lengths	Able to detect pathogens at low levels rapidly
						Lowest error rate of desktop sequencers	Errors in regions following GGC motifs [65]	Can be deployed locally Useful for diseases of unknown etiology
						Lower cost per bp	Decreasing read quality toward ends [66]
						Paired-end reads
Roche	Genome Sequencer (GS) FLX Titanium	1,000	1 Gb	23 hours	1 × 106	Long read lengths	Errors in homopolymeric regions	Able to de novo assemble genomes of novel pathogens from metagenomic samples
	GS Junior System	500	35 Mb	10 hours	1 × 105	Desktop machine	Errors in homopolymeric regions	Able to sequence novel genomes rapidly
						Longest read length of desktop sequencers	Lower depth compared to GS FLX	Can be deployed locally
								Useful for outbreak investigations
Life Technologies	Ion Torrent with Personal Genome Machine (PGM) 318 Chip	400	1-2 Gb	7 hours	3-5 × 106	Desktop machine	Errors in homopolymeric regions	Fastest output is helpful for urgent public health issues
						Fastest run time of desktop sequencers	Biased coverage in AT rich regions [67]	Can be deployed locally
	Proton	200	10 Gb	2-4 hours	6-8 × 107	Desktop machine	Short read length	Able to detect pathogens at low levels rapidly
						Very fast run time	Errors in homopolymeric regions	Can be deployed locally
								Useful for diseases of unknown etiology
Third generation sequencers
Pacific Biosystems	PacBio RS	2,000-15,000	100 Mb	2 hours	50,000	Very fast run time	High error rate	Able to assemble genomes for novel pathogens rapidly
						Very long read lengths	Sub-reads often shorter than quoted read lengths	Complementary to other methods
							Requires higher DNA input [67]
Oxford Nanopore	MinION	48,000	10s of Gb per 24 hours	Run until complete	Not applicable	Very fast run time	Not yet available
						No sample preparation required

Statistics for Illumina, Roche and maximum values for each category in each system are shown as of 2012. bp base pair, Mb megabases, Gb gigabases.