TABLE 2 .

Potential mosaic genes detected within each msg family^a

msg family	No. of msg genes					No. of potential msg mosaic genes				% mosaic
	Full length	Partial	Pseudogenes	Total	Nonmosaicb	Full lengthc	Partial	Pseudogenesd	Totalb
I	11	16	16	43	25	8	1	9	18	42
II	11	3	4	18	13	4	1	0	5	28
III	7	2	1	10	6	3	0	1	4	40
IV	6	1	2	9	7	1	0	1	2	22
Ve	8	6	1	15	14	1	0	0	1	7
VIe	6	1	0	7	7	0	0	0	0	0

^aDetected using the Recombination Analysis Tool and/or Bellerophon numerical screening methods among three different sets of genes of each msg family: full-length, full-length plus partial genes, or full-length plus pseudogenes.

^bThe number of potential mosaic genes among the msg families was almost significantly different (P = 0.06, chi-square test).

^cSix full-length mosaic genes were detected twice but with different pairs of putative full-length parent genes according to the set of genes analyzed (four of family I, one of family II, and one of family III). One mosaic gene of family I was detected twice: once with one full-length gene and one pseudogene as parents and once with two partial genes as parents. All 10 remaining genes were detected only once with a pair of full-length parents.

^dSix mosaic pseudogenes of family I had two pseudogenes as parents. Two of family I had one full-length gene and one a pseudogene as parents. The three remaining had a pair of full-length parents.

^eSeveral potential recombination events were detected for these two families using the more sensitive method TOPALi based on the hidden Markov model (see Fig. S9 at http://www.chuv.ch/microbiologie/en/imu_home/imu-recherche/imu-research-groups/imu-research-phauser/imu-supplementary_data.htm).