Skip to main content

Table 1 Relative change in area under CID (RCA) and relative change in sum of squared error (RCE) when running Qtip and Bowtie 2 on Mason-simulated Illumina-like samples of various lengths

From: A tandem simulation framework for predicting mapping quality

  

End-to-end

Local

  

RCA

RCE

RCA

RCE

 

Read length

Mean

SD

Mean

SD

Mean

SD

Mean

SD

Unpaired

50

−9.03

0.26

−24.61

0.19

−2.49

0.47

−15.30

0.57

 

100

−7.43

1.82

−18.53

2.61

−10.26

1.52

−27.87

1.59

 

150

−9.11

1.15

−16.22

0.62

−14.77

2.15

−29.27

1.03

 

250

−16.82

2.01

−19.97

0.44

−22.51

1.77

−28.85

0.62

 

500

−33.77

0.60

−27.26

0.37

−37.75

1.32

−31.65

1.01

Paired

50

−13.11

0.44

−19.60

0.44

−18.60

0.28

−33.63

0.24

 

100

−15.80

1.29

−21.79

0.33

−36.84

0.69

−45.42

0.53

 

150

−22.39

1.74

−25.94

0.28

−46.87

0.54

−52.76

0.56

 

250

−37.65

0.97

−33.48

0.22

−58.39

0.91

−58.75

1.38

 

500

−54.59

0.58

−44.91

0.34

−68.54

0.95

−70.37

3.45

  1. Relative change is expressed as a percentage. Each sample consists of 4 million reads/pairs. Samples are either unpaired or paired-end, and Bowtie 2 is run in either end-to-end or local alignment mode as indicated. Results are means and standard deviations over ten random trials, repeated starting from the input modeling step
  2. CID cumulative incorrect difference
  3. RCA relative change in area under CID
  4. RCE relative change in sum of squared errors
  5. SD standard deviation