pVAC-Seq: A genome-guided in silico approach to identifying tumor neoantigens

Table 1 Summary of predicted epitope candidates through pVAC-Seq pipeline

	MEL21			MEL38			MEL218	MEL69
	LN	Skin	Skin	Axilla	Breast	AbWall	LN	Skin / Limb	Skin / Scalp
	(2011)	(2012)	(2013)	(2012)	(2013)	(2013)	(2005)	(2013)	(2013)
Total SNVs	702	838	1099	359	402	385	695	256	282
Missense SNVs	443	515	598	219	247	238	437	141	162
21-mer FASTA entries (WT & MT)	856	1,004	1,002	424	482	462	850	272	314
Raw NETMHC output (9-mers)	11,152*2 (HLA-A02:01, HLA-A01:01)	13,072*2 (HLA-A02:01, HLA-A01:01)	13,044*2 (HLA-A02:01, HLA-A01:01)	5,512*3 (HLA-A02:01, HLA-A31:01, HLA-B07:02)	6,270*3 (HLA-A02:01, HLA-A31:01, HLA-B07:02)	6,010 *3 (HLA-A02:01, HLA-A31:01, HLA-B07:02)	11,050*3 (HLA-A02:01, HLA-A03:01, HLA-B44:02	3,542*2 (HLA-A02:01, HLA-A11:01)	4,088*2 (HLA-A02:01, HLA-A11:01)
Parsed NetMHC output (compared WT with MT)	3,796*2 (HLA-A02:01, HLA-A01:01)	4,465*2 (HLA-A02:01, HLA-A01:01)	4,458*2 (HLA-A02:01, HLA-A01:01)	1,871*3 (HLA-A02:01, HLA-A31:01, HLA-B07:02)	2,131*3 (HLA-A02:01, HLA-A31:01, HLA-B07:02)	2,042*3 (HLA-A02:01, HLA-A31:01, HLA-B07:02)	3,770*3 (HLA-A02:01, HLA-A03:01, HLA-B44:02	1,217*2 (HLA-A02:01, HLA-A11:01)	1,395*2 (HLA-A02:01, HLA-A11:01)
Filter 1: Binding based	110	121	144	103	112	111	161	50	65
HLA-A02:01 candidates only	79	96	111	52	48	46	93	25	34
Filter 2: Manually reviewed HLA-A02:01 candidates (Exome plus RNA-Seq)	11	11	12	14	16	16	24	6	12
Filter 3: Experimentally tested	16			14			18	12
Filter 4: Vaccine tested	7			7			7	10
Immunogenicity	3			3			3	4

The table illustrates the number of raw candidates predicted by NetMHC, and the parsing and filtering strategies applied thereafter to the final list of neoantigen candidates. These candidates were then communicated to our vaccine design collaborators who evaluated this list by patient-specific immunological assays (Filters 3 & 4) [20]

ISSN: 1756-994X