The data sets and scoring weights compiled for this analysis are listed in Table 1. A zero to five scoring system was developed that utilized both the relevance of the experimental conditions to the dormant state and the rank order of expression.
The maximum score for a particular experimental dataset was adjusted based on our judgement of relevance to the clinical occurrence of dormancy phase M. tuberculosis infections. For studies with multiple time series sampling, increasing weight was given to later time-points samples. For example Wayne and Hayes [8] have shown two phases of metabolic shiftdown: NRP-1 (characterized by tolerance to metronidazole, greatly slowed growth, and thickening of cell walls) and NRP-2 (characterized by sensitivity to metronidazole and absence of growth) [114]. In order to take into account both phases, but weigh the scoring toward NRP-2, which is likely more relevant to clinical dormancy, genes in NRP-2 received higher scores (e.g. 5 points maximum) than genes in NRP-1 (e.g. 2 or 3 points maximum).
Once the maximum score was set for a particular experiment, the genes were ordered from highest to lowest based on expression ratio (fold expression in the experimental condition versus cells in log-phase liquid culture). The highest scoring gene received the maximum score (listed in column 3 of Table 1. (e.g. 5, 4 ..., 1 point)). The score was decreased by 0.005 points for each gene in order until zero, or the end of the data set was reached. Thus when the maximum score was 4 points, the 100th ranked gene would receive a score of 3.500. For a maximum score of 5 points, 1000 genes or 25% of the MTB genome received a score. For experiments where data from multiple time points were collected, the maximum score across all time points was used as the final score.
For example the up-regulation score of Rv3133c, devR, was calculated by summing the maximum scoring time-point (underlined) from the following studies (score:timepoint):
Hampshire et al., 2004 [30] (0:18 d, 0:49 d, 3.57:62 d, 0.185:76 d);
Muttucumaru et al., 2004 [115] (1.915:7 d, 3.915:14 d);
Voskuil et al. 2004 [31] (0:4 d, 1.875:6 d, 1.895:8 d, 2.860:10 d, 2.815:12 d, 3.820:14 d, 3.07:20 d, 0:30 d, 3.325:80 d);
Schnappinger et al. 2003 [29] (0:24 h (naïve, ampl.), 0:24 h (naïve, oligo.), 0:48 h (naïve, ampl.)), 4.870:24 h (activated, ampl.), 4.800:24 h (activated, oligo.), 4.185:48 h (activated, ampl.).
In the remaining 3 studies [116-118]), this gene scored zero. Therefore, the up-regulation score of Rv3133c, devR equaled 16.175.
The scores for down-regulated expression were collected in a similar fashion, except that genes were ordered from lowest to highest and scores were negative. For the transposon-based knockout experiments, genes which had no effect on growth received a score of zero.
Answers to all your Biology Questions
