Table 1. ‘2316 ambiguous CircBase circRNAs’ on HeLa and Hs68 samples.

Dataset	HeLa						Hs68
Tools	RNaseR-	RNaseR+	Not depleted	Percent (%)	Enriched	Percent (%)	RNaseR-	RNaseR+	Not depleted	Percent (%)	Enriched	Percent (%)
CF	110	168	79	71.82	24	21.82	102	407	85	83.33	59	57.84
CE	110	167	79	71.82	24	21.82	103	407	86	83.50	60	58.25
CIRI	148	217	111	75.00	27	18.24	126	390	112	88.89	81	64.29
DCC	96	137	64	66.67	17	17.71	91	330	81	89.01	53	58.24
FC	82	91	41	50.00	11	13.41	52	227	40	76.92	32	61.54
KNIFE	170	199	111	65.29	25	14.71	131	395	109	83.21	73	55.73
MS	88	122	61	69.32	11	12.50	70	249	61	87.14	42	60.00
NCLS	34	37	19	55.88	3	8.82	23	100	16	69.57	11	47.83
PF	186	206	114	61.29	26	13.98	141	449	118	83.69	84	59.57
SG	178	213	105	58.99	23	12.92	137	366	80	58.39	56	40.88
UB	55	63	20	36.36	3	5.45	29	64	5	17.24	4	13.79

Note: Candidates with ≥ 2 supporting back-spliced junction reads were used in the analysis, and the number of supporting reads was normalized with sequencing depth before fold change calculation. After RNase R treatment, detected candidates with ≥ onefold enrichment was defined as ‘Not depleted’, while candidates with ≥ fivefold enrichment was regarded as ‘Enriched’. CF: circRNA_finder; CE: CIRCexplorer; FC: find_circ; MS: MapSplice; SG: Segemehl; NCLS: NCLScan; PF: PTESFinder; UB: UROBORUS.