Kolmogorov-Smirnov tests on junction type distributions from 454 pyrosequencing of spleen VHCμ expressed rearrangements.¹

1The procedure to compute KS tests and to aggregate JST from different fish is detailed in Figure S8. Aggregated distributions are denoted by «_» linking the corresponding fish number. Control fish are 1–3 and infected fish 4–7. 2KS test was applied to different combinations of distributions: (1) All control (aggregated) versus all infected (aggregated) (2) One Ctrl vs One Inf (all combinations): each control versus each infected individual distribution; (3) All combinations: each individual or aggregated distribution versus all other individual or aggregated distributions, within control or infected respectively. For example {1 vs 2; 1 vs 3; 2 vs 3; 1 vs 2_3; 2 vs 1_3; 3 vs 1_2 } within control fish. 3‘Two-sided’ analysis tests if the distribution in Infected samples is equal to the distribution of Control samples (i.e.H0, null hypothesis of the KS test). P-values< 5% *; <1% **; <0.1% *** indicate that the distributions are significantly different. 4‘One-sided’ analysis tests if the distribution in Infected samples is shorter than in Control samples (H0). P-values<5% *; <1% **; <0.1% *** indicate that the distributions in control fish are significantly shorter than in infected fish. 5NS: non significant, i.e. H0 = “equal distribution” is not rejected.