Table 2.

Neo-Fusion Results for Simulated and Real Spliced Peptide Searches

	HLA-1	HLA-II	T rypsina
Search Parametersb
Peptide Length Limitation	8–16	8–25	7–50
Precursor Mass Tolerance (PPM)	6.6	5.2	4.5
Product Mass Tolerance (PPM)	15.2	23.6	21.0
Benchmark Results Using Simulated PSPs			(Fraction #5 of 9)
Total Simulated PSPs	4450	440	3608
Simulated PSPs per Replicate	222	22	180
Average Number of Simulated Cis-PSPs Found ± SD	55.5 ± 10.4	4.6 ±2.4	76.0 ±8.1
Average Percent of Simulated Cis-PSPs Found ± SD	25.0% ±4.7	20.7% ± 10.7	42.2% ±4.5
Average Simulated Cis-PSP FDR ± SD	5.3% ±2.6	0.0% ± 0.0c	3.5% ± 1.8
Average Number of Simulated Cis- and Trans-PSPs Found ± SD	48.0 ±8.4	5.3 ±2.4	89.9 ±8.7
Average Percent of Simulated Cis- and Trans-PSPs Found ± SD	21.6% ±3.8	24.3% ± 10.9	50.0% ± 4.8
Average Simulated Cis- and Trans-PSPs FDR ± SD	31.7% ±6.7	23.1% ±20.5c	30.3% ± 4.4
Real Spliced Peptide Sequences Discovered			(All 9 fractions)
Real Cis-Spliced (“1%” FDR)	62	13	487
Real Cis-Spliced Without Ambiguityd (“ 1 %” FDR)	47	9	340
Real Cis- and Trans-Spliccd (“1%” FDR)	156	81	2490
Real Cis- and Trans-Spliced Without Ambiguityd (“1%” FDR)	29	10	496

^aThe trypsin sample was offine fractionated, while the HLA samples were not. A single trypsin fraction (#5 of 9) was used for the benchmark simulation, but all fractions were used for the discovery of real spliced peptides.

^bAll files were searched using b- and y-ions from HCD fragmentation, fixed carbamidomethylation of cysteine, variable oxidation of methionine, a maximum of 4096 isoforms, and a maximum of 2 modifications per peptide. Trypsin was allotted two missed cleavages of lysine and arginine. Mass tolerances were selected based on MetaMorpheus calibration suggestions.

^cThe low FDRs and high variance reported for the HLA-II sample are due to the scarcity of nonspliced identifications; no richer HLA-II datafiles are available to our knowledge.

^dAmbiguity is defined as two or more PSP sequences sharing the highest score for a single spectrum.