Exhaustive prediction of disease susceptibility to coding base changes in the human genome

BMC Bioinformatics

Table 1 SVM performance during testing (labeled with the keyword 'Max') and comparison with SIFT predictions.

	HGMD		dbSNP
	TP	FN	TN	FP	Sensitivity	Specificity	Accuracy	F-measure
	*Testing set created using random mutations not used for training*
SVM Complete ; Max	16527	3567	64378	12454	82.2	83.8	83.5	67.4
SVM Conservation Only ; Max	14467	5627	49178	27659	72.0	64.0	65.7	46.5
	*Testing set created to allow comparison with SIFT*
SIFT	159	245	13995	5281	39.4	72.6	71.9	5.4
SVM Complete	287	117	12049	7227	71.0	62.5	62.7	7.2
SVM Conservation Only	230	174	10216	9060	56.9	53.0	53.1	4.7

Abbreviations: TP = True Positives; FP False Positives; TN = True Negatives; FN = False Negatives.

ISSN: 1471-2105