PeakErazor

1
PeakErazor
Pre-processing of mass lists prior to peptide
mass searching
Karin Hjernø, Peter Roepstorff and Peter
Højrup Department of Biochemistry and Molecular
Biology, University of Southern Denmark, Odense
M, Denmark.
Problem Very advanced peptide mass search
programs are available today, making
identification of proteins by peptide mass search
(PMS) using tryptic MALDI-TOF MS data more and
more trustworthy. However, the procedure still
faces a number of practical problems Usually
two-point calibration (either external or
internal) of the mass spectra are used and in
most cases the obtained accuracy is sufficient
for identification of the protein in question.
to be excluded from the peptide mass list before
using the list as input in a protein
identification In addition, identification of the
project specific contaminants and determination
of the exact mass of unknown contaminants can be
tricky.
However, in many spectra the peaks used for
calibration can be missing, of poor quality or
the isotope envelope may overlap with other
peaks. Another problem is peptide contamination.
Most people working with proteins separated by
2D-gel come across peaks that appear in many
spectra and does not seem to be specific for the
protein of interest. Including these peaks in the
peptide mass search will obviously make the
search less precise and consequently they have
Solution To address these problems we have
created a small, user-friendly Windows-based
program, PeakErazor, using a simple concept By
comparing all mass values from a tryptic digest
against a list of known contaminants,
contaminating peaks can easily be removed. At the
same time you can perform a multipoint
calibration leading to greater precision and the
program helps you detect new contaminants as you
go on. These features leaves you with a much
larger precision in the actual PMS search. Four
of the features of the program are illustrated
here. Below is a small discussion of the
practical problems of calibration and the future
improvements of the program. PeakErazor is freely
available at http//welcome.to/GPMAW.
FILTERING AND PRECISION CHECK
FILTERING AND PRECISION CHECK
Contamination peak automatically erazed from the
mass list
Graph illustrating the mass precision of the
contaminants identified
RED Rejected peaks in all analysed spectra from
the same project
One of the best ways to beat your enemies is to
know them! The Peak-Erazor program contains an
Erazor-list containing already known contaminants
( the enemies!). Therefore it can immediately
identify peaks that are non-specific for the
protein of interest.
When contaminating peaks are removed from your
mass spectrum, you will have larger score in the
actual PMS search, as the contaminations will not
give rise to spurious hits, and the statistics
on the correct hit will be better reducing the
risk of false positive hits..
BLUE Accepted peaks in all analysed spectra
from a specific project
2

Calibration In a linear time-of-flight (TOF)
instrument the mass-to-charge (m/z) value is
proportional to the square of the ion flight time
(t2). However, linear TOF instruments are not
sufficient to get a high enough mass accuracy
when working with peptides. Therefore,
time-delayed ion extraction and reflectrons are
built into the instrument to give a very high
mass resolution, a low signal-to-noise ratio and
high accuracy on the mass determination.
Unfortunately, these improvements disturb the
simple correlation between m/z and t (1). Instead
the correlation is depending on different
parameters like initial ion velocity, specific
conditions and instruments used. In order to take
these effects into consideration, one could
do
Where to find PeakErazor! The program will be
avaliable for free at http//welcome.to/gpmaw/ P
lease contact hjernoe_at_bmb.sdu.dk for further
information
multipoint calibration with calibration points
distributed uniformly over the whole mass range.
This can be done using polymers as an external
calibration (2). However, a slight difference in
the ion flight time is found depending on the
location on the target, which means that external
calibration is less accurate than internal
calibration. Mixing the sample with the polymer
is not the way to solve this problem due to ions
suppression effects, which will result in polymer
suppressing signals from the peptides and vica
versa. Instead internal correction of the sample
position-dependent errors can be used (2). Other
groups have developed different more or less
complicated calibration algorithms (e.g. using
polynomial fitting procedure or
quadratic correction methods (all resulting in
more accurate mass values (3,4). At the moment,
a simple linear fit is used for multipoint
calibration in PeakErazor. In most cases this
enhance the overall precision of the mass values
by a factor of 1.5-2x. In the future the goal is
to incorporate a higher-order calibration
function based on the known peaks already found
in the spectrum (trypsin, keratin etc.) or by
feeding the program with knowledge from previous
protein identifications and in this way build up
a useful higher-order function which can be used
for a more accurate calibration than the
two-point calibration used in most laboratories
today.
References 1. Hack, C.A and Benner, W.H., Rapid
Commun. Mass Spectrom. 2002161304 2. Gobom, J.,
Mueller, M., Egelhofer, V., Theiss, D., Lehrach,
H. and Nordhoff, E., Anal. Chem. 2002743915 3.
Christian, N.P., Arnold, R.J. and Reilly, J.P.,
Anal. Chem. 2000723327 4. Whittal, R.M.,
Schriemer, D.C. And Li, L., Anal. Chem.
1997692734