DTreePred: an online viewer based on machine learning for pathogenicity prediction of genomic variants

BMC Bioinformatics

Table 2 Confusion matrices of the classification performance of the predictors on 200 recent variants (pathogenic and neutral) deposited in ClinVar

		Expected class
		Pathogenic	Neutral
DTreePred (Precision and Recall*: 97.06% and 99.00%)
Predicted class	Pathogenic	99	3
Predicted class	Neutral	1	97
NDamage (Precision and Recall*: 79.09% and 87.00%)
Predicted class	Pathogenic	87	23
Predicted class	Neutral	13	77
ML (Precision and Recall*: 90.48% and 95.00%)
Predicted class	Pathogenic	95	10
Predicted class	Neutral	5	90

*Precision: True positive (pathogenic)/(true positive (pathogenic) + false positive (pathogenic))
**Recall: True positive (pathogenic)/(true positive (pathogenic) + false negative (neutral))

ISSN: 1471-2105