Bioinformatics for LC-MS/MS-Based Proteomics - LC-MS/MS in Proteomics: Methods and Applications - page 67

Table 4.3
(Continued)

Machine-learning algorithms
Gist ( 26 ) svm.sdsc.edu

Percolator
( 27 , 28 )

http://noble.gs.washington.edu/proj/percolator/

Mascot
Percolator
( 28 )

http://www.sanger.ac.uk/Software/analysis/
MascotPercolator /

data. The original method was developed by Mann and Wilm ( 9 ) .
Because it is very rare to obtain complete sequence coverage in
an MS/MS spectrum, only a partial interpretation is performed
identifying three or more amino acids in a sequence. This short
stretch of sequence is combined with the fragment ion mass val-
ues which enclose it, the peptide mass, and the enzyme specificity
in a database search ( Table 4.3 ) . This can be enough informa-
tion to unambiguously identify a protein. A sequence tag acts
as a filter on the database. A more powerful approach to using
sequence tags is an error-tolerant tag. Relaxing the specificity by
removing the peptide molecular mass constraint allows the tag
to float within the candidate sequence so that a match is pos-
sible even if there is a difference in the calculated mass to one
side or the other side of the tag. This enables a sequence tag to
match a peptide when there is an unsuspected modification or a
variation in the primary amino acid sequence. The sequence tag
approach boasts rapid search times as it is essentially a filter and
can be error tolerant allowing for the matching of unknown mod-
ifications or single-nucleotide polymorphisms (SNPs) but requires
correct interpretation of the MS/MS spectrum, although ambi-
guity is acceptable.

Depending upon the complexity of a digested proteome and the
accuracy of the mass measurements of the resulting peptides,
many proteins can be identified by a single peptide which has a
unique mass termed an accurate mass tag (AM) ( 20 ) . This idea
was then extended to utilize the liquid chromatography reten-
tion time in a later paper ( 21 ) and termed an accurate mass and
time tag (AMT). The accurate mass and time tag idea is fairly
new, although the basis can be traced back to the original protein
identification search engines ( 22 ) . Because of the complexity of
protease-treated proteomes, only mass spectrometers with suffi-
ciently high mass accuracy, such as FT-ICR, Orbitrap, and very
accurate QTOF instruments are able to produce data suitable for
AM and AMT tag analysis. There are now a number of search

4.3.AccurateMass
andTimeTags