Table 1.

Microscopy techniques used to study the gut phageome, including the major advantages and limitations of each technique.

Technique	Example gut discoveries	Advantages	Limitations
TEM	Tailed phages were the most abundant VLPs in human faeces (Flewett et al., 1974).	Visualisation of phage morphology.	Biased towards identifying tailed phages due to potential loss of tail structures in sample preparation (Williamson et al., 2008).
	Faecal samples from patients were found to share no VLPs (Hoyles et al., 2014).		Limited to observations of morphologies.
			Time-consuming.
EFM	Up to 5.58 x 109 VLPs were observed per gram of faeces (Hoyles et al., 2014).	Enumeration of VLPs in samples.	VLP counts are conservative estimates of true viral abundances, given the imprecision of visualising single fluorescent dots.
		Can validate viral purification procedures.	Loss of VLPs during preparation and filtration of samples, e.g. large VLPs of the order Megavirales (Colson et al., 2010).
		Greater accuracy and speed compared to TEM.	Viability of the VLP to infect and lyse bacterial cells is unknown.
			VLPs may be membrane vesicles, gene transfer agents or cell debris containing nucleic acids (Forterre et al., 2013).
phageFISH	The viability of VLPs can be determined through single cell dynamic measurements, as shown with marine phages (Allers et al., 2013).	PhageFISH is the only non-genetic method to implicate lytic, lysogenic, and chronic phage infection modes (Dang and Sullivan, 2014).
CLSM	The complex microenvironment and spatiotemporal succession can be studied in multispecies biofilms, as shown with non-phage viruses (Røder et al., 2016).	Non-destructive sampling. Can be used to visualise the biofilm infection over time.	Limited to biofilms of bacterial species which can be fluorescently labelled.
CryoEM	Phage capsids of Bortadella phages were visualised at angstrom resolution, discovering unique protein folds, as shown with non-gut phages (Zhang et al., 2013).	Very high resolution.	Destructive sampling.