Evaluating the effect of database inflation in proteogenomic search on sensitive and reliable peptide identification

Table 3 Number of peptides with charge 2+ at 1% FDR identified from search against real proteogenomic databases using X!Tandem

Database (target + decoy)		TD	BP	MB	SepTD	SepBP	SepMB
6FTT_y + 6FTD_y	Total	3,626	4,281	3,807	3,942	4,515	3,870
	Known	3,603	4,106	3,781	3,942	4,443	3,870
	Novel	23	175	26	0	72	0
6FTT_h + 6FTD_h	Total	4,177	5,620	4,813	6,347	6,018	6,188
	Known	4,115	5,316	4,765	6,336	5,950	6,180
	Novel	62	304	48	11	68	8
SGT_h + SGD_h	Total	8,034	11,059	9,152	8,957	11,150	9,552
	Known	7,966	11,016	9,087	8,940	11,136	9,531
	Novel	68	43	65	17	14	21

6FTT_y (or 6FTT_h): proteogenomic database constructed by 6-frame translation of yeast (or human) genome. 6FTD_y (or 6FTD_h): decoy database for 6FTT_y (or 6FTT_h). SGT_h: proteogenomic database constructed by splicing information obtained from human RNA sequencing data. SGD_h: decoy database for SGT_h. TD: target-decoy strategy. BP: target-decoy strategy using a refined score calculated by the self-boosted Percolator. MB: mixture model-based method. SepTD, SepBP, and SepMB denote separate filtering of known and novel peptides using TD, BP, and MB, respectively

ISSN: 1471-2164