471 and inc = 0.217, whereas the corresponding optimal topology resulted in res = 0.176 and inc = 0.000. The average bootstrap support of the optimised topologies compared to the average bootstrap of random marker topologies was significantly higher for congruence at the 5 marker level
(bootopt = 88.33 vs. p38 MAPK apoptosis bootrand = 86.38, p < 0.001), 6 marker level (bootopt = 88.67 vs. bootrand = 87.81, p < 0.001), and 7 marker level (bootopt = 88.92 vs. bootrand = 88.29, p < 0.001), as well as for resolution at Fludarabine in vivo the 6 marker level (bootopt = 90.71 vs. bootrand = 87.81, p < 0.001). Figure 4 The impact of the number of markers on phylogenetic parameters. The effect of concatenating sequence markers on topology (of the Francisella genus) in comparison with the LY3039478 mw whole-genome tree for (A) incongruence score, (B) resolution score, and (C) average bootstrap support from 1000 replicates. The results of the optimised topology comparisons are shown as crosses. Table 4 Summary of the optimisation procedure for resolution (res) and congruence (inc) in the Francisella genus where the consensus set of markers are highlighted according to how often they are selected in the optimal partitions of markers; position 1 corresponds to the most represented marker Position 1 2 3 4 5 6 7 No of markers Metric 2 res 08-fabH
35-tpiA inc 08-fabH 35-tpiA 3 res 08-fabH 35-tpiA 24-lpnB inc 08-fabH 35-tpiA 02-16 s 4 res Idoxuridine 08-fabH 35-tpiA 24-lpnB 27-parC inc
35-tpiA 08-fabH 01-16S 02-16 s 5 res 08-fabH 35-tpiA 24-lpnB 27-parC 22-lpnA inc 35-tpiA 08-fabH 24-lpnB 27-parC 33-rpoB 6 res 08-fabH 24-lpnB 35-tpiA 27-parC 22-lpnA 25-mdh inc 35-tpiA 08-fabH 24-lpnB 04-16 s 01-16S 33-rpoB 7 res 08-fabH 35-tpiA 24-lpnB 26-mutS 27-parC 18-groEL 22-lpnA inc 35-tpiA 08-fabH 01-16S 04-16 s 24-lpnB 27-parC 25-mdh Markers 02-16 s + ItS + 23 s and 04-16 s + ItS + 23 s are abbreviated as 02-16 s and 04-16 s, respectively. Discussion Knowledge about theoretical limitations of marker assays is important for the successful detection and identification of bacteria in research as well as public health contexts. Existing methods for detection and identification of Francisella were developed with limited knowledge about the genetic diversity within the Francisella genus. From a clinical perspective, the lack of knowledge of diversity in the environment may be of minor importance since diagnostic sampling is performed on humans or animals suspected of having the disease. In contrast, use of the same detection assays for environmental sampling can lead to problems with false positive results. The recent increase in publicly available genome sequences enables development of improved detection and identification methods for both purposes.