Comprehensive comparison of in silico MS/MS fragmentation tools of the CASMI contest: database boosting is needed to achieve 93% accuracy

Blaženović, Ivana; Kind, Tobias; Torbašinović, Hrvoje; Obrenović, Slobodan; Mehta, Sajjan S.; Tsugawa, Hiroshi; Wermuth, Tobias; Schauer, Nicolas; Jahn, Martina; Biedendieck, Rebekka; Jahn, Dieter; Fiehn, Oliver

doi:10.1186/s13321-017-0219-x

Table 2 Results for the challenge (validation) data of the CASMI 2016 contest

From: Comprehensive comparison of in silico MS/MS fragmentation tools of the CASMI contest: database boosting is needed to achieve 93% accuracy

#	Tools	Top hits	Top 5	Top 10	Top 20
1	CFM-ID + ID_sorted + DB + MS/MS Voting/consensus	181	194	201	204
2	CFM-ID + ID_sorted + MAGMa(+) + DB + MS/MS Voting/consensus	180	195	200	205
3	CFM-ID + ID_sorted + MAGMa(+) + MetFrag + DB + MS/MS Voting/consensus	180	194	200	204
4	CFM-ID + DB + MS/MS	180	193	199	201
5	MAGMa(+) + ID_sorted + DB + MS/MS Voting/consensus	180	193	197	201
6	CFM-ID + MAGMa(+) + DB + MS/MS Voting/consensus	180	192	195	202
7	MetFrag + MAGMa(+) + DB + MS/MS Voting/consensus	180	188	194	198
8	MAGMa(+) + DB + MS/MS	180	188	192	198
9	MetFrag + CFM-ID + MAGMa(+) + DB + MS/MS Voting/consensus	179	190	196	201
10	MetFrag + CFM-ID + DB + MS/MS Voting/consensus	178	192	199	203
11	CFM-ID + ID_sorted + MAGMa(+) + MetFrag + MS-FINDER + DB + MS/MS Voting/consensus	175	191	200	203
12	MetFrag + MS-FINDER + CFM-ID + DB + MS/MS Voting/consensus	175	189	194	200
13	MetFrag + MS-FINDER + CFM-ID + MAGMa(+) + DB + MS/MS Voting/consensus	175	189	194	200
14	MS-FINDER + ID_sorted + DB + MS/MS Voting/consensus	175	189	194	199
15	MS-FINDER + CFM-ID + MAGMa(+) + DB + MS/MS Voting/consensus	175	188	196	201
16	MetFrag + MS-FINDER + MAGMa(+) + DB + MS/MS Voting/consensus	175	186	191	197
17	MS-FINDER + MAGMa(+) + DB + MS/MS Voting/consensus	175	185	190	195
18	ID_SORTED + DB + MS/MS	174	195	198	204
19	MetFrag + ID_sorted + DB + MS/MS Voting/consensus	174	194	199	203
20	MS-FINDER + CFM-ID + DB + MS/MS Voting/consensus	174	189	195	201
21	MetFrag + DB + MS/MS	174	189	192	197
22	MetFrag + MS-FINDER + DB + MS/MS Voting/consensus	174	187	190	197
23	MS-FINDER + DB + MS/MS	174	184	185	191
24	MetFrag + CFM-ID + MAGMa(+) + DB Voting/consensus	151	184	192	198
25	CFM-ID + DB	151	183	191	197
26	MetFrag + MS-FINDER + CFM-ID + DB Voting/consensus	151	180	191	198
27	MS-FINDER + CFM-ID + MAGMa(+) + DB Voting/consensus	151	179	191	198
28	CFM-ID + MAGMa(+) + DB Voting/consensus	150	184	189	199
29	MetFrag + MAGMa(+) + DB Voting/consensus	150	181	189	194
30	MetFrag + MS-FINDER + MAGMa(+) + DB Voting/consensus	150	178	186	193
31	MS-FINDER + MAGMa(+) + DB Voting/consensus	150	174	183	191
32	MetFrag + CFM-ID + DB Voting/consensus	149	186	196	201
33	MAGMa(+) + DB	149	180	185	193
34	MS-FINDER + CFM-ID + DB Voting/consensus	149	179	189	199
35	MS-FINDER + DB	148	173	178	186
36	MetFrag + DB	147	185	188	194
37	MetFrag + MS-FINDER + DB Voting/consensus	147	178	184	193
38	ID_SORTED + DB	134	188	194	202
39	Randomize + DB + MS/MS	123	184	189	197
40	Randomize + DB	119	176	180	189
41	ID_SORTED	106	169	177	186
42	MetFrag in silico	53	92	111	137
43	MetFrag + MS-FINDER + CFM-ID in silico Voting/consensus	51	95	129	151
44	MetFrag + CFM-ID in silico Voting/consensus	47	102	129	153
45	MetFrag + MS-FINDER + CFM-ID + MAGMa(+) in silico Voting/consensus	46	97	128	152
46	MetFrag + CFM-ID + MAGMa(+) in silico Voting/consensus	42	104	126	150
47	CFM-ID + MAGMa(+) in silico Voting/consensus	39	94	123	148
48	MetFrag + MAGMa(+) in silico Voting/consensus	39	90	111	128
49	MetFrag + MS-FINDER + MAGMa(+) in silico Voting/consensus	38	79	117	138
50	MS-FINDER + CFM-ID + MAGMa(+) in silico Voting/consensus	34	97	127	147
51	MetFrag + MS-FINDER in silico Voting/consensus	33	76	103	125
52	MS-FINDER + MAGMa(+) in silico Voting/consensus	32	69	93	119
53	MS-FINDER + CFM-ID in silico Voting/consensus	30	76	110	139
54	CFM-ID in silico (dot product)	29	76	104	122
55	MAGMa(+) in silico	28	72	98	117
56	MS-FINDER in silico	23	57	79	93
57	Randomize	20	27	28	121

‘MetFragCL, CFM-ID, MAGMa+ and MS-FINDER’ designate results obtained by the in silico fragmentation software tools. ‘DB’ designates priority ranking by presence in chemical and biochemical databases. ‘MS/MS’ designates presence in MS/MS libraries based on >400 dot-product similarity. 208 MS/MS spectra of the CASMI 2016 training data were used

Back to article page

ISSN: 1758-2946

Contact us

Submission enquiries: journalsubmissions@springernature.com

Journal of Cheminformatics

Contact us