Table 2. Scaffold and correctness statistics for NA12878 assemblies scaffolded with different Hi-C libraries.

“True links” is an idealized case where the Hi-C links have been filtered in advance. The NG50 of human reference GRCh38 is 145 Mbp. The ratio between NG50 and NGA50 represents how many erroneous joins affect large scaffolds in the assembly. The bigger the difference between these values, the more aggressive the scaffolding was at the expense of accuracy. Longest chunk represents the longest error-free portion of the scaffolds. We observed that the 3D-DNA mis-assembly detection was overly aggressive in some cases, and so we ran some assemblies both with and without this feature. For the Illumina assembly as an input, 3D-DNA w correction did not finish within two weeks and is omitted. An evaluation of a previously published [20] 3D-DNA assembly from short-read contigs is included in S3 Table but did not exceed SALSA2’s NGA50.

Dataset	Method	NG50(Mbp)	NGA50(Mbp)	Longest Chunk (Mbp)	Orientation Errors	Ordering Errors	Chimeric Errors
Arima-HiC	SALSA2 true links	83.31	79.48	172.19	78	101	0
	SALSA2 w graph	112.08	71.54	164.46	102	106	90
	SALSA2 wo graph	118.42	58.81	155.68	148	112	135
	3D-DNA	90.15	22.44	89.46	182	133	115
	SALSA1	19.09	14.81	73.14	99	176	96
Mitotic Hi-C	SALSA2 w graph	61.97	24.18	145.53	81	54	19
	SALSA1	27.88	15.62	85.71	142	78	120
	3D-DNA w correction	0.199	0.627	2.24	9775	10563	6159
	3D-DNA wo correction	85.56	17.18	70.18	250	215	164
Chicago	SALSA2 w graph	5.80	4.54	34.60	46	60	98
	SALSA1	5.21	3.94	34.60	83	21	187
	3D-DNA w correction	3.63	2.74	18.62	63	69	324
	3D-DNA wo correction	9.61	4.76	44.48	67	63	137
Illumina Assembly	SALSA2 w graph	96.78	7.99	43.56	1830	2299	635
	SALSA2 wo graph	119.57	4.16	26.22	2225	2353	738
	3D-DNA w correction	NA	NA	NA	NA	NA	NA
	3D-DNA wo correction	176.09	1.00	13.12	5935	3433	2119