Table 4.

Comparison with Variant Calling methods on all datasets.

Dataset	Method	Recall (%)	FP/TP ratio	Precision (%)	# of False negatives	Mapped reads (%)
HIV 100x	LoFreq	97.33	0.004	99.60	444	89.51
	Vphaser	98.90	0.007	99.26	183	89.51
	ShoRAH	55.21	0	100	7746	98.04
	MultiRes	99.69	0.011	98.88	51	97.89
HIV 400x	LoFreq	84.83	0	99.99	2522	99.55
	Vphaser	95.92	0.292	77.37	678	99.55
	ShoRAH	55.21	0	100	7746	99.95
	MultiRes	95.57	0.007	99.33	736	97.34
HCV1P	LoFreq	98.30	1.282	43.82	31	99.99
	Vphaser	93.51	1.628	38.05	118	99.99
	ShoRAH	91.92	0	100	147	99.99
	MultiRes	98.24	0.597	62.64	32	97.32
HCV2P	LoFreq	97.10	1.046	48.87	60	100
	Vphaser	95.65	1.492	40.13	90	100
	ShoRAH	83.73	0	100	337	99.95
	MultiRes	98.79	0.201	83.27	25	85.14
5-viral mix	LoFreq	99.06	0.085	92.15	101	98.59
	Vphaser	92.68	0.039	96.25	789	98.59
	ShoRAH	98.66	0.014	98.99	109	99.3
	MultiRes	99.39	0.077	92.82	66	96.29

The Recall, false positive to true positive ratios (FP/TP), Precision, number of false negatives, and % of mapped reads by methods LoFreq, VPhaser-2, ShoRAH, and MultiRes are computed for listed datasets. All reads from a sample were aligned using bwa-mem tool for LoFreq and VPhaser-2 under default settings. ShoRAH uses its own aligner for read alignment and variant calling, while k-mers detected by MultiRes were aligned using bwa-mem for MultiRes. Outputs from LoFreq (version 2.1.2), VPhaser-2 (last downloaded version October 2015), and ShoRAH (last downloaded version from November 2013) are compared against known variants for simulated datasets. For 5-viral mix, the consensus reference provided by [35] was used to determine ground truth variants. MultiRes variants are determined by aligning 35-mers to a reference sequence and bases occurring at more than 0.01 frequency as variants. Bold for each dataset indicates the best method for the performance measures.