fluids under pathological conditions and thus not detected in healthy ones. In the
first 10 years (decade 2001–2010), thousands of papers claimed biomarker discov-
ery, and millions of euros (dollars) were spent at this new field. However, no more
than 100 biomarkers were validated and really used for diagnostic and prognostic
purposes.^13 Protein profiling based on chromosome to chromosome, combined with
shotgun proteomics, can actually yield in the identification of up to 4000–5000
proteins,^14 but it is only the first step in a long way that yields in validated disease
biomarker(s) at its end. According to Mitchell,^15 “variability in sample processing,
problems with the instrumentation (both separation technology and mass spectro-
metry (MS) systems) and problems with data analysis all contributed to the diffi-
culties.” Further problems are also poorly defined sampling and small number of
samples that were analyzed. The main reason for this bottleneck was the lack of
methods for reproducible high-throughput sample preparation, complicated and
time-consuming protein analysis by LC-MS/MS, and incomplete data analysis. In
order to solve these problems, methods for isolation of organelles and fast frac-
tionation were developed. After these steps, further high-throughput enrichment
(or isolation) and identification of low-abundance proteins as potential biomarker
candidates can be performed. Further step is the application of quantitative prote-
omics by use of isotope tags (e.g. iTraq) and newly also label-free techniques.^16
Again, it has to be emphasized that the proteomic profile in a biological sample is a
function of the status in the cell or organ in a particular moment or condition. In
order to discover the real status, especially under pathological conditions, a much
larger number of samples have to be analyzed, and there is absolute necessity for
use of high-throughput methods for both sample preparation and further proteomic
analysis (mass spectrometry and data processing, see Mitchell 2010 ). This discus-
sion resulted in a reconsideration of some strategies, especially for biomarker
discovery, but the scheme developed by Koomen at al.^17 that defines the role and
contribution of proteomics in personalized patient care is still relevant (see Fig. 2 ).
This scheme was recently followed in both studies that utilized personalized omics
(or proteomics) techniques in order to create integrated patient profile on the way to
further personalized medical treatment.^18
(^13) Poste ( 2011 ), pp. 156–157.
(^14) Muraoka et al. ( 2013 ), pp. 208–213.
(^15) Mitchell ( 2010 ), pp. 665–670.
(^16) Lawson et al. ( 2006 ), pp. 2747–2758; Breen et al. ( 2012 ), pp. S89–S100; Cao et al. ( 2013 ),
pp. 8112–8120.
(^17) Koomen et al. ( 2008 ), pp. 1780–1794.
(^18) Chan et al. ( 2012 ), p. A16530; Pieborn et al. ( 2014 ), pp. 2846–2855.
The Role of Proteomics in Personalized Medicine 183