Table 1 

Performance Comparison on the simulated mRNASeq data 

Algorithms 
Condition 
IsoformEx 
RSEM [21] (v1.1.8) with default parameters 
Cufflinks [20] (v0.9.3) (minisoformfraction 0) 


The number of estimated transcripts (n) 
v_{est }≥ 0 
55416 
55441 
55441 


v_{est }> 0.01 
35064 
25967 
25839 



v_{est }≥ 0 
8.41 × 10^{6} 
1.02 × 10^{5} 
9.23 × 10^{6} 



v_{est }> 0.01 
1.32 × 10^{5} 
2.15 × 10^{5} 
1.96 × 10^{5} 



r = Corr(v_{true }, v_{est }) 
v_{est }≥ 0 
0.921 
0.839 
0.913 


v_{est }> 0.01 
0.920 
0.838 
0.912 



Estimation error and correlation coefficient between estimated expression levels (v_{est}) and known true expression levels (v_{true}) for our simulated dataset when all estimated transcripts (v_{est }≥ 0) were considered or some expressed transcripts (v_{est }> 0.01) were only considered. p_{true} (i) denotes the ith element of the proportion vector of true expression values (p_{true} = v_{true}/∑ v_{true}), and p_{est} (i) denotes the ith element of the proportion vector of estimated expression values (p_{est} = v_{est}/∑ v_{est}). The error was defined as the mean value of absolute difference between the true proportion vector and the proportion vector of the estimated values. 

Kim et al. BMC Bioinformatics 2011 12:305 doi:10.1186/1471210512305 