Table 11.

Switch error percentage on datasets of NA12878

Chrom.	HapCHAT	HapCol	WhatsHap	WhatsHap	HapCUT2	ReFHap	ProbHap	FastHare
			cov. 15x	cov. 20x
1	1.929	-	1.926	1.924	1.920	-	-	2.191
2	0.038	-	0.050	0.035	0.030	-	-	0.374
3	0.044	-	0.045	0.039	0.031	-	-	0.381
4	2.042	-	2.052	2.048	2.033	-	-	2.237
5	1.829	-	1.828	1.824	1.825	-	-	1.998
6	1.991	-	1.990	1.991	1.983	-	-	2.205
7	0.659	-	0.669	0.666	0.660	-	-	0.924
8	1.743	-	1.746	1.748	1.749	-	-	1.992
9	1.966	-	1.965	1.966	1.940	2.140	-	2.187
10	0.949	-	0.949	0.948	0.939	1.171	-	1.232
11	2.092	-	2.101	2.101	2.081	2.282	-	2.325
12	0.041	-	0.055	0.048	0.043	0.319	-	0.405
13	0.051	-	0.036	0.049	0.029	0.285	-	0.349
14	0.034	-	0.042	0.039	0.032	0.347	-	0.421
15	0.055	0.331	0.069	0.065	0.043	0.358	-	0.427
16	0.022	0.289	0.022	0.029	0.027	0.322	-	0.420
17	0.055	0.277	0.071	0.067	0.047	0.337	-	0.426
18	1.895	-	1.879	1.876	1.889	2.072	-	2.122
19	2.629	-	2.642	2.644	2.616	2.807	-	2.914
20	0.043	0.277	0.046	0.043	0.043	0.412	-	0.451
21	0.033	-	0.044	0.041	0.030	0.364	-	0.408
22	2.102	2.323	2.106	2.114	2.068	2.378	-	2.452

Each row corresponds to a chromosome. The dataset consists of all reads aligned to the chromosome. We report the results obtained by running the tools with maximum coverage 30 × for HapCHAT, 25 × for HapCol, 15 × and 20 × for WhatsHap. No maximum coverage was set for HapCUT2, ReFHap, ProbHap, and FastHare. The best result (lowest value) for each dataset is boldfaced