Table 1

*Sources, experimental models, and scoring criteria for Mycobacterium tuberculosis* DNA microarray gene expression and genome-wide gene knock-out (growth phase essentiality) data used in this study.**
Reference	Experimental model	Timepoint: Maximum score^a

Betts et al., 2002 [116]	Starvation under controlled O₂	96 h: 3 24 h: 2 4 h: 1
Hampshire et al., 2004 [30]	Nutrient depletion under controlled O₂	62 and 75 d: 5 49 d: 4 18 d: 2
Muttucumaru et al., 2004 [115]	Wayne model of hypoxia [8]	14 d (NRP-2): 4 7 d (NRP-1): 2
Voskuil et al. 2004 [31]	Wayne model of hypoxia [8]	30 and 80 d: 5 14 and 20 d: 4 10 and 12 d: 3 6 and 8 d: 2
Schnappinger et al. 2003 [29]	Infection of mouse macrophages, +/- γ-INF	24 and 48 h: 5
Karakousis et al., 2004 [117]	Hollow fiber subcutaneous implant in mice	10 d: 3
Talaat et al., 2004 [118]	Infection of mice. MTB harvested from lung^b	28 d: 3
Sassetti et al. 2003a [13]	TraSH mutated libraries grown on solid media	14 d:5
Rengarajan et al. 2005 [12]	Infection of mouse macrophages, +/- γ-INF with TraSH mutated libraries of M. tuberculosis	7 d: 5
Sassetti et al. 2003b [14]	C57BL/6J mice infected with TraSH mutated libraries of M. tuberculosis	7, 14, 28 and 56 d: 5

Footnotes ^aMaximum score based on relevance as a dormancy model. ^bRatio of M. tuberculosis from Balb/c lung to MTB in aerated culture for 28 d.
Murphy and Brown BMC Infectious Diseases 2007 7:84 doi:10.1186/1471-2334-7-84