Latest Results

Terrier v0.4 was released and this was trained on RepBase31.04 (released 04-23-2026) which as substantially more sequences than the previous version (RepBase29.10) used to train the version of Terrier for the publication in Briefings in Bioinformatics (2025).

This has altered the performance of the model. See the tables below for the performance of both versions of Terrier on the test data.

Rice

Order

Superfamily

Version

Threshold

Proportion Classified

Accuracy

Proportion Classified

Accuracy

v0.3

0.7

94.7%

94.4%

90.7%

98.5%

v0.4

0.7

97.3%

94.5%

92.0%

98.6%

v0.3

0.9

90.7%

97.1%

89.3%

98.5%

v0.4

0.9

93.3%

98.6%

90.7%

100.0%

Fruit Fly

Order

Superfamily

Version

Threshold

Proportion Classified

Accuracy

Proportion Classified

Accuracy

v0.3

0.7

80.8%

87.9%

63.6%

91.0%

v0.4

0.7

81.6%

88.6%

66.0%

91.8%

v0.3

0.9

65.4%

94.5%

53.7%

96.6%

v0.4

0.9

67.3%

93.1%

53.7%

96.4%

Human

Order

Superfamily

Version

Threshold

Proportion Classified

Accuracy

Proportion Classified

Accuracy

v0.3

0.7

85.3%

89.5%

75.9%

92.3%

v0.4

0.7

84.2%

89.6%

73.0%

93.3%

v0.3

0.9

78.0%

94.4%

71.3%

95.5%

v0.4

0.9

76.6%

94.9%

68.2%

96.7%

Mouse

Order

Superfamily

Version

Threshold

Proportion Classified

Accuracy

Proportion Classified

Accuracy

v0.3

0.7

84.6%

92.1%

75.9%

95.8%

v0.4

0.7

84.2%

92.2%

74.8%

95.5%

v0.3

0.9

78.5%

95.6%

72.2%

97.5%

v0.4

0.9

76.6%

96.7%

70.2%

97.9%