Biomedical Engineering Reference
In-Depth Information
Table 1. Results of GG, WW and WPSB
Three procedures
GG WW WPSB
Summary
Number of peaks class
258
60
474
Number peaks in one class
(minimum)
2
1
2
(1st Quintile)
2
15
4
(Median)
5
20
6
(Mean)
7.63
17
7
(3rd Quintile)
12
20
10
(Maximum)
20
20
20
Range of the bin
(minimun)
3.6
0
1.89
(1st Quintile)
9.53
37.75
6.89
(Median)
12.85
122.5
9.51
(Mean)
13.87
200
11.34
(3rd Quintile)
17.37
257
14.35
(Maximum)
30.89
1382
32.33
5. Discussion and Future Study
The new procedure, WPSB, provides a potential framework for the feature
extraction method in general. Within this framework, every step has some
room to improve. For example, in the future, dierent types of wavelets may
be adopted for dierent types of MS data; more exible semi-parametric
functions may be considered to t the baseline; and normalization part
requires more understanding on the biological knowledge to come up a
better schema for spectra comparisons.
It is important to ascertain whether or not the peaks being found by the
algorithm correspond to real phenomena in the spectra. So for searching
a criteria for evaluating a procedure, we need to relay on the biological
knowledge. Statistical results alone may not be adequate to demonstrate
it as a reasonable method or not. Therefore, the idea of evaluating the
methods itself is still an open research topic.
The new generation of mass spectrometers produces an astonishing
amount of high-quality data in a brief period of time, leading to inevitable
data analysis bottlenecks. Automated data analysis algorithms are required
for rapid and repeatable processing of proteomic MS data. Toward this
end a mathematical algorithm is presented in [17] that automatically lo-
Search WWH ::
Custom Search