Skip to main content

Table 2 Quality validation of running LTRharvest on the D. melanogaster genome sequences (release 3)

From: LTRharvest, an efficient and flexible software for de novo detection of LTR retrotransposons

Chr

Predictions

References

TP

hTP

FP

FN

2L

91

66

64

1

26

1

2R

89

54

49

3

37

2

3L

96

59

51

7

38

1

3R

96

67

60

6

30

1

4

11

4

4

0

7

0

X

122

54

51

3

68

0

Total

505

304

279

20

206

5

  1. The used options are listed in Table 1 in column 'value D. mel.'. Column 'Predictions' gives the number of predicted and column 'References' the number of annotated 'full length' LTR retrotransposons for each chromosome. A prediction is classified as true positive (TP), if the maximum allowed distance between 5' and 3' coordinates of the prediction and the reference is at most 20 nucleotides. If only one of the two predicted boundary coordinates lies within the allowed distance of 20 nucleotides, the prediction is categorised as a half true positive (hTP). All other predictions are labelled false positive (FP). LTR retrotransposons that are missing in the prediction are labelled false negative (FN).