Achieving high-sensitivity for clinical applications using augmented exome sequencing

Patwardhan, Anil; Harris, Jason; Leng, Nan; Bartha, Gabor; Church, Deanna M.; Luo, Shujun; Haudenschild, Christian; Pratt, Mark; Zook, Justin; Salit, Marc; Tirch, Jeanie; Morra, Massimo; Chervitz, Stephen; Li, Ming; Clark, Michael; Garcia, Sarah; Chandratillake, Gemma; Kirk, Scott; Ashley, Euan; Snyder, Michael; Altman, Russ; Bustamante, Carlos; Butte, Atul J.; West, John; Chen, Richard

doi:10.1186/s13073-015-0197-4

Table 1 Accuracy across target regions. Errors, Sensitivity, and FDR for the ACE, WGS, SSCR, SS, NX, and NG platforms based on evaluation of observed variant calls using data normalized to 100× mean coverage (conventional WES and ACE) or 31× WGS. Calculations are based on position and genotype matching to the GIBv2.18 high-confident call-set within the MIG (left), a target region common to all ACE and WES platforms (middle, Common Target File), and a target region aggregated across all ACE and WES specific target files that contain moderate-impact and high-impact loci (right, Union Target File)

From: Achieving high-sensitivity for clinical applications using augmented exome sequencing

		MIG					Common Target File					Union Target File
		TP	FP	FN	%Sens	%FDR^a	TP	FP	FN	%Sens	%FDR^a	TP	FP	FN	%Sens	%FDR^a
		TP	FP	FN	95%CI	95%CI	TP	FP	FN	95%CI	95%CI	TP	FP	FN	95%CI	95 % CI
ACE	SNV	5362	5	62	98.9	0.1	7133	12	90	98.8	0.2	7486	6	191	97.5	0.1
	SNV	5362	5	62	98.5-99.1	<0.1-0.2	7133	12	90	98.5-99.0	0.1-0.3	7486	6	191	(97.1-97.8)	(<0.1-0.2)
	InDel	34	1	2	94.4	2.9	83	0	0	100	<0.1	198	3	16	92.5	1.5
	InDel	34	1	2	81.3-99.3	0.1-14.9	83	0	0	95.7-100	<0.1-4.3	198	3	16	(88.1-95.7)	(0.3-4.3)
WGS^b	SNV	5309	2	115	97.9	<0.1	7076	6	147	98.0	0.1	7479	2	198	97.4	<0.1
	SNV	5309	2	115	97.5-98.2	<0.1-0.1	7076	6	147	97.6-98.3	<0.1-0.2	7479	2	198	(97–97.8)	(<0.1-0.1)
	InDel	33	1	3	91.7	2.9	78	0	5	94.0	<0.1	197	2	17	92.1	1.0
	InDel	33	1	3	77.5-98.2	0.1-15.3	78	0	5	86.5-98.0	<0.1-4.6	197	2	17	(87.6-95.3)	(0.1-3.6)
SSCR	SNV	5341	4	83	98.5	0.1	7107	11	116	98.4	0.2	7443	4	234	97.0	0.1
	SNV	5341	4	83	98.1-98.8	<0.1-0.2	7107	11	116	98.1-98.7	0.1-0.3	7443	4	234	(96.5-97.3)	(<0.1-0.1)
	InDel	34	2	2	94.4	5.6	82	0	1	98.8	<0.1	194	4	20	90.7	2
	InDel	34	2	2	81.3-99.3	0.7-18.7	82	0	1	93.5-100	<0.1-4.4	194	4	20	(85.9-94.2)	(0.6-5.1)
SS	SNV	5355	2	69	98.7	<0.1	7126	5	97	98.7	0.1	7468	3	209	97.3	<0.1
	SNV	5355	2	69	98.4-99.0	<0.1-0.1	7126	5	97	98.4-98.9	<0.1-0.2	7468	3	209	(96.9-97.6)	(<0.1-0.1)
	InDel	33	2	3	91.7	5.7	82	0	1	98.8	<0.1	192	5	22	89.7	2.5
	InDel	33	2	3	77.5-98.2	0.7-19.2	82	0	1	93.5-100	<0.1-4.4	192	5	22	(84.8-93.4)	(0.8-5.8)
NX	SNV	5240	4	184	96.6	0.1	7020	8	203	97.2	0.1	6754	10	923	88.0	0.1
	SNV	5240	4	184	96.1-97.1	<0.1-0.2	7020	8	203	96.8-97.6	<0.1-0.2	6754	10	923	(87.2-88.7)	(0.1-0.3)
	InDel	33	2	3	91.7	5.7	77	2	6	92.8	2.5	160	6	54	74.8	3.6
	InDel	33	2	3	77.5-98.2	0.7-19.2	77	2	6	84.9-97.3	0.3-8.8	160	6	54	(68.4-80.4)	(1.3-7.7)
NG	SNV	5190	31	234	95.7	0.6	6900	39	323	95.5	0.6	7065	38	612	92.0	0.5
	SNV	5190	31	234	95.1-96.2	0.4-0.8	6900	39	323	95.0-96.0	0.4-0.8	7065	38	612	(91.4-92.6)	(0.4-0.7)
	InDel	31	4	5	86.1	11.4	74	2	9	89.2	2.6	168	10	46	78.5	5.6
	InDel	31	4	5	70.5-95.3	3.2-26.7	74	2	9	80.4-94.9	0.3-9.2	168	10	46	(72.4-83.8)	(2.7-10.1)

FDR false discovery rate, FN false negatives, FP false positives, MIG medically interpretable genome, SENS Sensitivity, TP true positives
^aFDR is used in lieu of specificity due to a large skew in the TN, FP class distribution
^bIn WGS data, there was no difference in error rates when using either VQSLOD scores or hard-thresholding cutoffs for InDels

Back to article page

ISSN: 1756-994X

Contact us

Submission enquiries: editorial@genomemedicine.com
General enquiries: info@biomedcentral.com

Genome Medicine

Contact us