Computational Methods for Biomarker Discovery in Proteomics and Glycomics

1
Computational Methods for Biomarker Discovery in
Proteomics and Glycomics

• Vijetha Vemulapalli
• School of Informatics
• Indiana University
• Capstone Advisor Dr. Haixu Tang

2
What are Biomarkers?

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Substances present in increased or decreased
  amounts in body fluids or tissues that indicate
  exposure, disease or susceptibility to disease.

3
Some Uses of Biomarkers

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Biomarkers are increasingly being used for the
  following purposes
• Prognosis / Diagnosis of disease
• Monitoring response to medication
• With high sensitivity and throughput, proteomics
  and glycomics is capable of identifying many
  potential biomarkers simultaneously.

4
More on Biomarkers

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• A lot of times biomarkers have not been
  identified clearly. But based on the signature
  pattern of glycans and proteins, samples can be
  classified as healthy and diseased.

5
What is Proteomics?

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Proteomics Proteomics is the study of proteins
  and proteomes using high-throughput technology.

http//parasol.tamu.edu/groups/amatogroup/foldings
erver/images/proteinL.gif http//biology.clc.uc.ed
u/graphics/bio104/cell.jpg
6
What is Glycomics?

• Problem Definition
• Background
• LC-MS Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Glycoproteins Proteins with attached
  polysaccharides .

• Glycans Polysaccharide chain attached to a
  protein

• Glycome The entire set of glycans that are
  present in a cell or a bodily fluid at a certain
  point of time under certain conditions.

• Glycomics Study of structure and function of
  oligosaccharides in a cell or organism.

http//www.glyfdis.org/images/bg_image.jpg
7
High Throughput Technologies to Identify
Biomarkers

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

http//phy.asu.edu/phy598-bio/D420Notes2006_file
s/image002.jpg
8
Why the Focus on Proteomics and Glycomics?

• Information content

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Transcriptome
Static
Genome
Transcriptome
Dynamic
Proteome
Glycome
9
Biomarker Discovery using Proteomics
10
Liquid Chromatography / Mass Spectrometry (LC/MS)

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Why LC/MS for analysis of proteomes?
• LC spreads complexity of the sample over time.
• MS identifies ions based on their mass/charge
  value.
• Software exists currently to identify proteins
  in a sample using data from a LC-MS experiment.

11
Liquid Chromatography (LC)

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Liquid Chromatography is a technique that
  separates ions or molecules dissolved in a
  solvent based on size of the ion/molecule,
  adsorption, ion-exchange or other similar
  characteristics.

http//wwwlb.aub.edu.lb/webcrsl/high_p3.jpg
12
What is Mass Spectrometer?

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Mass Spectrometry (MS) is an instrument that
  identifies ions based on their mass-to-charge
  ratio.

Source http//www.chemguide.co.uk/analysis/masspe
c/howitworks.html http//www.bmms.uu.se/ltq-ft.h
tm
13
Visualization of LC/MS Data 2D Map

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

14
How Do We Find Biomarkers From LC-MS Data?

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Identified Proteins and Peptides
15
How Do We Find Biomarkers From LC-MS Data?
Continued

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Sample 1
Quantification 1
Analyze to find Biomarkers
Sample 2
Quantification 2
Sample 3
Quantification 3
Sample N
Quantification N
16
MSView

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

MSView
Relative Quantification
Visualization
Components
Visual comparison /Analysis
Further analysis for Biomarker Discovery
Purpose
17
Extracted Ion Chromatogram (XIC)

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Chromatogram created by plotting the intensity
  of the signal observed at a chosen m/z value in a
  series of mass spectra recorded as a function of
  retention time.

Source http//www.lcpackings.com/applications/Pro
bot/images/dual_fract04B.png
18
Visualization XIC

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

19
Relative Quantification using Peptide
Identification Results

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Identification of peptides
Data from LC-MS experiment
Extracted Ion Chromatogram of peptide
MS View
Peak selection
Area calculation
20
Quantification Peak Selection Algorithm

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

21
Quantification Sample Results

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

22
Biomarker Discovery using Glycomics
23
How does Capillary Electrophoresis (CE) work?

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

http//faculty.washington.edu/dovichi/UBUBTpage/re
search/Methods/CEintro/ceintro.GIFimgrefurlhttp
//faculty.washington.edu/dovichi/UBUBTpage/researc
h/Methods/CEintro/CE_LIF.htmlh531w684sz25hl
enstart3um1tbnid_JDf4X3dJn170Mtbnh108tb
nw139prev/images3Fq3Dcapillary2Belectrophore
sis26svnum3D1026um3D126hl3Den
24
What does the data look like?

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

25
Biomarker Discovery using Glycomics Overview

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Data from different samples
CE Analyze
Analysis of quantification for identifying
Biomarkers
26
Direct Comparison Dynamic Time Warping (DTW)

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• DTW algorithm aligns two time series having
  similar curves but are skewed differently over
  time.

Source http//db-www.aist-nara.ac.jp/theme/bioinf
o_kenji-h_dtw.png
27
Direct Comparison DTW continued

• Sakoe-Chuba Band is used to reduce time space
  complexity.
• Parameters used in DTW
• - Band width - Peak extention penalty
• - Difference in peak intensities.
• - Difference in peak direction

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Stan Aslvador and Philip Chan. FastDTWToward
Accurate Dynamic Time Warping in Linear Time and
Space, KDD Workshop on Mining Temporal and
Sequential Data, 2004
28
Method Dynamic Time Warping

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

29
Method continued

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Corresponding peaks
30
Results

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

Corresponding peaks
31
Summary

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

• Proteomics - MSView
• Glycomics - CE Analyze

Quantification results for Biomarker Discovery
CE Data
32
Acknowledgements
Dr. Haixu Tang - My advisor Dr. Randy
J.Arnold Dr. Yehia Mechref Dr. Milos
Novotny Dr. David E.Clemmer Dr. Sun Kim Dr.
Jeong-Hyeon Choi Dr. Stephen J. Valentine Yin Wu
Manolo D.Plasencia School of
Informatics Funding NIH/NCRR MetaCyt
Initiative _at_ Indiana University

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References

33
References
1 Higgs, R.E., Knierman, M.D., Gelfanova, V.,
Butle,r J.P. and Hale, J.E. (2005) Comprehensive
label-free method for the relative quantification
of proteins from biological samples. J. Proteome
Res., 4, 1442-1450. 2 Linsen, L., Locherbach,
J., Berth, M., Becher, D. and Bernhardy, J.
(2006) Visual Analysis of Gel-Free Proteome Data.
IEEE Transactions on Visualization and Computer
Graphics,12, 497-508. 3 Prakash, A., Mallick,
P., Whiteaker, J., Zhang, H., Paulovich, A.,
Flory, M., Lee, H., Aebersold, R., and
Schwikowski, B. (2006) Signal maps for mass
spectrometry-based comparative proteomics. Mol.
Cell. Proteomics 5, 423 432 4 Leptos, K. C.,
Sarracino, D. A., Jaffe, J. D., Krastins, B., and
Church, G. M. (2006) MapQuant open-source
software for large-scale protein quantification.
Proteomics 6, 1770 1782 5 Aebersold, R., and
Mann, M. (2003) Mass spectrometry-based
proteomics. Nature 422, 198 207

• Problem Definition
• Background
• LC-MS
• Method
• Results
• CE
• Method
• Results
• Acknowledgements
• References