Table 3 The performance of precision, recall, and F-score for each software tool
From: InDel marker detection by integration of multiple softwares using machine learning techniques
Indels | ||||||
Tool | Precision(%) | Diff * | Recall (%) | Diff | F-score (%) | Diff |
BF-M | 99.32 | 0.00 | 65.20 | 0.00 | 78.72 | 0.00 |
SVM-M | 95.75 | −3.57 | 84.56 | 19.37 | 89.81 | 11.09 |
GATK | 99.49 | 0.17 | 25.50 | −39.70 | 40.59 | −38.13 |
Pindel | 94.69 | −4.63 | 41.36 | −23.84 | 57.57 | −21.15 |
Samtools | 97.46 | −1.86 | 65.71 | 0.51 | 78.49 | −0.23 |
SOAPIndel | 97.25 | −2.07 | 74.74 | 9.54 | 84.52 | 5.80 |
Varscan | 98.59 | −0.73 | 64.66 | −0.54 | 78.10 | −0.62 |
Deletions | ||||||
Tool | Precision(%) | Diff | Recall (%) | Diff | F-score (%) | Diff |
BF-M | 99.21 | 0.00 | 66.34 | 0.00 | 79.51 | 0.00 |
SVM-M | 95.85 | −3.37 | 84.84 | 18.50 | 90.01 | 10.50 |
GATK | 99.40 | 0.19 | 26.03 | −40.30 | 41.26 | −38.25 |
Pindel | 89.89 | −9.32 | 39.06 | −27.28 | 54.45 | −25.06 |
Samtools | 97.78 | −1.43 | 66.32 | −0.02 | 79.03 | −0.48 |
SOAPIndel | 96.86 | −2.35 | 75.43 | 9.09 | 84.81 | 5.30 |
Varscan | 98.55 | −0.66 | 65.17 | −1.17 | 78.46 | −1.05 |
Insertions | ||||||
Tool | Precision(%) | Diff | Recall (%) | Diff | F-score (%) | Diff |
SVM-M | 95.66 | −3.77 | 84.29 | 20.23 | 89.62 | 11.69 |
GATK | 99.59 | 0.16 | 24.96 | −39.10 | 39.92 | −38.00 |
Pindel | 99.45 | 0.01 | 43.66 | −20.40 | 60.68 | −17.24 |
Samtools | 97.14 | −2.29 | 65.09 | 1.03 | 77.95 | 0.03 |
SOAPIndel | 97.64 | −1.79 | 74.05 | 9.99 | 84.23 | 6.31 |
Varscan | 98.64 | −0.80 | 64.15 | 0.09 | 77.74 | −0.18 |