Skip to main content
. 2017 Aug 29;9(10):2862–2878. doi: 10.1093/gbe/evx170

Table 1.

Summary of In Vitro TE Validation Experiments for Ref− and Ref+ Insertion Loci

Type Set Informative Loci
All Loci
N TP FP FN N TP FP FN
Ref− All Ref 80 0.90 0.04 0.06 111 0.87 0.07 0.06
Hypothesis-driven 48 0.93 0.03 0.04 71 0.88 0.05 0.07
Random 32 0.82 0.05 0.06 40 0.80 0.13 0.07
Ref+ All Ref+ 31 0.74 0.23 0.04 40 0.70 0.26 0.03
Pindel+Break Dancer 17 0.76 0.23 0.02 20 0.71 0.28 0.01
Pindel 8 0.79 0.14 0.07 10 0.70 0.24 0.06
BreakDancer 6 0.67 0.31 0.02 10 0.71 0.27 0.14

Note.—Results are shown for loci that were phylogenetically informative and all loci, that is, those lacking amplicons in more than one sample (All). The number of tested loci (N) and frequency of amplicon size differences that matched the computational prediction (true positives, TP), and false positively (FP) or false negatively (FN) predicted insertions are shown. For Ref− loci, random loci (Random), and loci predicted to support a specific phylogenetic hypothesis (Hypothesis-driven) were selected. For Ref+ markers, all loci were randomly selected.