Simulated  data Model 3a CV 3 Model 4a CV 4 Model 5 CV 5 Model 6 CV 6
# of SV 3 2 3 32 32 3 3 3 3
log L   -720.31   -882.16   -570.07   -567.24  
# params   7   37   8   8  
AIC   1454.61           1150.47  
BIC   1477.7           1176.86  
RSS     815.75   1396.51   932.92   796.93
µ1 9.5 9.5 9.5 9.52 9.53 9.47 9.52 9.5 9.5
µ0 6 7.82 7.83 8.71 6.01 6.08 6.31 6.08 6.09
σ2 0.15 0.16 0.16 0.24 0.31 0.15 0.15 0.16 0.16
σ2 0.25 3.1 2.90 2.13 0.24 0.18 0.28 0.18 0.19
ρ1 0.2 0.27 0.25 0.85 0.63 0.25 0.24 0.25 0.25
ρ0 0 0.27 0.20 0.47 0.27 0.15 0.3 0.05 0.04
1 1 1 1 1 0.85 0.843 0.836 0.11 0.28 0.842 0.75 0.85 0.85
1 0 0 0 0 0.05 0.157 0.164 0.027 0.015 0.044 0.19 0.05 0.05
1 1 1 1 0 0.1     0.091 0.066 0.114 0.059 0.1 0.1
# of SV=Number of splicing variants; log L= log likelihood;
# of params=Number of parameters; AIC =Akaike information criterion; BIC=Bayesian information criterion; RSS=Residual Sum of Squares
Model 3a=Known splice variants; Model 4a=All possible splice variants; Model 5=Known and high mixture proportions splice variants; Model 6=Manually selected splice variants CV= Cross-validation analysis of corresponding model
Table 1: Results on the simulation data of 5 exons and three splicing variants.