Faster and more accurate pathogenic combination predictions with VarCoPP2.0

BMC Bioinformatics

Table 2 Classification report for the stratified LOGO cross validation and the independent validation set per class

		pre	rec	spe	f1	geo	Support
LOGO	Neutral	1.00	0.95	0.93	0.98	0.94	150,500
	Disease-causing	0.04	0.93	0.95	0.07	0.94	301
	avg/total	1.00	0.95	0.93	0.97	0.94	150,801
VAL	Neutral	1.00	0.95	0.98	0.98	0.97	10,000
	Disease-causing	0.10	0.98	0.95	0.19	0.97	53
	avg/total	1.00	0.95	0.98	0.97	0.97	10,053

The LOGO rows indicate the results obtained during the stratified LOGO cross validation, while the VAL rows indicate the results obtained for the independent validation set. The bottom row presents a weighted (relative to the number of positive and negative instances) average of the metrics. Pre:precision, rec:recall, spe:specificity, geo:geometric mean, support: number of class instances (see “Evaluation Metrics” section for the definition of these metrics)

ISSN: 1471-2105