mass spectrometry-based proteomics

1

Mass Spectrometry-based Proteomics

Xuehua Shen

(Adapted from slides with textbook)

2

Outline

• Motivation of proteomics

• Mass spectrometry-based proteomics

• Instrumentation of mass spectrometry

• De novo sequencing algorithm

• Database search

• Algorithms of real software (e.g., sequence tags)

3

Motivation

• Proteins are working units of the cells

– The number of found genes is much less than the number of expressed proteins

– Directly related with cell processes and diseases

>1,000,000 distinct protein forms

~30,000 human genes

DNA

Alternative

splicing

mRNA Protein

Post-translational

Modification

>100,000 RNA messages

SNP

4

Tools for Proteomics

• Edman degradation reaction

• NMR (Nuclear Magnetic Resonance)

• X-ray crystallography

• Protein array

• Mass Spectrometry

5

Mass Spectrometry-based Proteomics

• Primary sequence (sequencing, identification)

• Post-translational modification (PTM) (characterization)

• Quantitative proteomics (quantification)

• Protein-protein interaction

6

7

Components of Mass Spectrometer

• Ion source (ESI and MALDI)

• Mass analyzer (ion traps, TOF, Quadrupole, FT, etc.)

– Mass-to-charge ratio (m/z)

• Ion detector

8

Peptide and Intact Protein

• Peptide: a fragment of protein

• Some enzymes, e.g. trypsin, break protein into peptides.

• Some technology put intact protein into the mass spectrometer

9

Peptide Fragmentation

• Peptides tend to fragment along the backbone.

• Fragments can also loose neutral chemical groups like NH3 and H2O.

H...-HN-CH-CO . . . NH-CH-CO-NH-CH-CO-…OH

Ri-1 Ri Ri+1

H+

N-Terminus C-Terminus

Collision Induced Dissociation

10

Ideal Mass Spectrum

11

Real Mass Spectrum

12

N- and C-terminal Peptides

N-term

inal

pep

tides

C-te

rmin

al p

eptid

es

13

Terminal peptides and ion types

Peptide

Mass (D) 57 + 97 + 147 + 114 = 415

Peptide

Mass (D) 57 + 97 + 147 + 114 – 18 = 397

without

14

N- and C-terminal Peptides

N-term

inal

pep

tides

C-te

rmin

al p

eptid

es

415

486

301

154

57

71

185

332

429

15

N- and C-terminal Peptides

N-term

inal

pep

tides

C-te

rmin

al p

eptid

es

415

486

301

154

57

71

185

332

429

16

N- and C-terminal Peptides

415

486

301

154

57

71

185

332

429

17

N- and C-terminal Peptides

415

486

301

154

57

71

185

332

429

Problem:

Reconstruct peptide from the set of masses of fragment

18

Mass Spectra

G V D L K

mass0

57 Da = ‘G’ 99 Da = ‘V’LK D V G

• The peaks in the mass spectrum:

– Prefix

– Fragments with neutral losses (-H2O, -NH3)

– Noise and missing peaks.

and Suffix Fragments.

D

H2O

19

Protein Identification with MS/MS

G V D L K

mass0

Inte

nsity

mass0

MS/MSPeptide Identification:

20

Protein Identification by Tandem Mass Spectrometry

SSeeqquueennccee

S#: 1708 RT: 54.47 AV: 1 NL: 5.27E6T: + c d Full ms2 638.00 [ 165.00 - 1925.00]

200 400 600 800 1000 1200 1400 1600 1800 2000

m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e Ab

unda

nce

850.3

687.3

588.1

851.4425.0

949.4

326.0524.9

589.2

1048.6397.1226.9

1049.6489.1

629.0

MS/MS instrumentMS/MS instrument

De Novo interpretation•SherengaDatabase search•Sequest

21

De Novo vs. Database Search

S#: 1708 RT: 54.47 AV: 1 NL: 5.27E6T: + c d Full ms2 638.00 [ 165.00 - 1925.00]

200 400 600 800 1000 1200 1400 1600 1800 2000m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Re

lative

Ab

un

da

nce

850.3

687.3

588.1

851.4425.0

949.4

326.0524.9

589.2

1048.6397.1226.9

1049.6489.1

629.0

WR

A

C

VG

E

K

DW

LP

T

L T

WR

A

C

VG

E

K

DW

LP

T

L T

De Novo

AVGELTK

Database Search

Database ofknown peptides

MDERHILNM, KLQWVCSDL, PTYWASDL, ENQIKRSACVM, TLACHGGEM, NGALPQWRT,

HLLERTKMNVV, GGPASSDA, GGLITGMQSD, MQPLMNWE,

ALKIIMNVRT, AVGELTK, HEWAILF, GHNLWAMNAC,

GVFGSVLRA, EKLNKAATYIN..

Database ofknown peptides

MDERHILNM, KLQWVCSDL, PTYWASDL, ENQIKRSACVM, TLACHGGEM, NGALPQWRT,

HLLERTKMNVV, GGPASSDA, GGLITGMQSD, MQPLMNWE,

ALKIIMNVRT, AVGELTK, HEWAILF, GHNLWAMNAC,

GVFGSVLRA, EKLNKAATYIN..

Mass, Score

22

Current Status

• It is still a open problem of protein sequencing no matter whether using de novo sequencing or database search methods

• Following algorithms only deal with simplified (or ideal) spectrums

• Some algorithms combine de novo sequencing and database search

23

Pros and Cons of de novo Sequencing

• Advantage:– Gets the sequences that are not necessarily in the database.

– An additional similarity search step using these sequences may identify the related proteins in the database.

• Disadvantage:– Requires higher quality data.

– Often contains errors.

24

Outline

• Motivation of proteomics

• Mass spectrometry-based proteomics

• Instrumentation of mass spectrometry

• De novo sequencing

• Database search

• Algorithms of real software (e.g., sequence tags)

25

De novo Peptide Sequencing

S#: 1708 RT: 54.47 AV: 1 NL: 5.27E6T: + c d Full ms2 638.00 [ 165.00 - 1925.00]

200 400 600 800 1000 1200 1400 1600 1800 2000

m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Rel

ativ

e A

bund

ance

850.3

687.3

588.1

851.4425.0

949.4

326.0524.9

589.2

1048.6397.1226.9

1049.6489.1

629.0

SequenceSequence

26

Peptide Sequencing Problem

Goal: Find a peptide with maximal match between an experimental and theoretical spectrum.

Input:

– S: experimental spectrum

– Δ: set of possible ion types

– m: parent mass

Output:

– P: peptide with mass m, whose theoretical spectrum matches the experimental S spectrum the best

27

Procedure of De Novo Sequencing

• Build spectrum graph

– How to create vertices (from masses)

– How to create edges (from mass differences)

• Find best path or rank paths of spectrum graph

– How to find candidate paths

– How to score paths

28

S E Q U E N C Eb

Mass/Charge (M/Z)Mass/Charge (M/Z)

From Sequence to Spectrum

29

a

Mass/Charge (M/Z)Mass/Charge (M/Z)

S E Q U E N C E

From Sequence to Spectrum(cont.)

30

S E Q U E N C E

Mass/Charge (M/Z)Mass/Charge (M/Z)

a is an ion type shift in b

From Sequence to Spectrum(cont.)

31

y

Mass/Charge (M/Z)Mass/Charge (M/Z)

E C N E U Q E S

From Sequence to Spectrum (cont.)

32

Mass/Charge (M/Z)Mass/Charge (M/Z)

Inte

nsit

yIn

tens

ity

From Sequence to Spectrum (cont.)

33

Mass/Charge (M/Z)Mass/Charge (M/Z)

Inte

nsit

yIn

tens

ity

From Sequence to Spectrum (cont.)

34

noise

Mass/Charge (M/Z)Mass/Charge (M/Z)

From Sequence to Spectrum (cont.)

35

MS/MS Spectrum

Mass/Charge (M/z)Mass/Charge (M/z)

Inte

nsit

yIn

tens

ity

36

Some Mass Differences between Peaks Correspond to Amino Acids

ss

ssss

ee

eeee

ee

ee

ee

ee

ee

qq

qq

qquu

uu

uu

nn

nn

nn

ee

cc

cc

cc

37

Now decoding from spectrum to sequence…?

Build spectrum graph

38

Peptide Fragmentation

HO NH3+

| |

R1 O R2 O R3 O R4

| || | || | || |H -- N --- C --- C --- N --- C --- C --- N --- C --- C --- N --- C -- COOH | | | | | | | H H H H H H H

y3

b2

y2 y1

b3a2 a3

b2-H2O

y3 -H2O

b3- NH3

y2 - NH3

• Different ion types (b, y, b-NH3, b-H2O)• Fragment at one site (internal ions)

39

Example of Ion Type

• Δ={δ1, δ2,…, δk}

• Ion types

{b, b-NH3, b-H2O}

correspond to

Δ={0, 17, 18} *Note: In reality the δ value of ion type b is -1 but we will “hide” it for the sake of simplicity

40

Why Peptide Sequencing hard

• Two ladders of overlapping masses, could not tell whether it is b ion or y ion

• Incomplete fragmentation

• Chemical noise

• Mass accuracy of the instrument is not good enough (Q=K, G+V=156.090, R=156.101)

• Q: Is sequencing shorter or longer peptide harder?

41

Vertices of Spectrum Graph

• Vertices are generated by reverse shifts corresponding to

ion types Δ={δ1, δ2,…, δk}

• Every mass s in an MS/MS spectrum generates k vertices

V(s) = {s+δ1, s+δ2, …, s+δk}

corresponding to potential N-terminal peptides

• Vertices of the spectrum graph:

{initial vertex}V(s1) V(s2) ... V(sm) {terminal vertex}

42

Reverse Shifts

Shift in H2O+NH3

Shift in H2O

43

Edges of Spectrum Graph

• Two vertices with mass difference corresponding to

an amino acid A:

– Connect with an edge labeled by A (Directed Graph)

• Gap edges for di- and tri-peptides

– Potential sequence tag method (covered later)

44

Best Path of Spectrum Graph

• How to find candidate paths

• There are many paths, how to find the correct one?

• We need scoring to evaluate paths

45

Find Candidate Paths

• Heuristics: find a path with maximum number

of edges

• Longest path problem in DAG

• DFS (Depth First Search)

46

Path Score

• p(P,S) = probability that peptide P produces spectrum S= {s1,s2,…sq}

• Scoring = computing probabilities

47

Finding Optimal Paths in the Spectrum Graph

• For a given MS/MS spectrum S, find a peptide P’ maximizing p(P,S) over all possible peptides P:

• Peptides = paths in the spectrum graph

• P’ = the optimal path in the spectrum graph

• Some software rank paths

p(P,S)p(P',S) Pmax

48

Ratio Test Scoring for Partial Peptides

• Incorporates premiums for observed ions and penalties for missing ions.

• Example: for k=4, assume that for a partial peptide P’ we only see ions δ1,δ2,δ4.

The score is calculated as:

RRRR q

q

q

q

q

q

q

q 4321

)1(

)1(

49

Why Not Sequence De Novo?

• De novo sequencing is still not very accurate!

• Less than 30% of the peptides sequenced were completely correct!

Algorithm Amino Acid Accuracy

Whole Peptide Accuracy

Lutefisk (Taylor and Johnson, 1997). 0.566 0.189

SHERENGA (Dancik et. al., 1999). 0.690 0.289

Peaks (Ma et al., 2003). 0.673 0.246

PepNovo (Frank and Pevzner, 2005). 0.727 0.296

50

De Novo vs. Database Search

S#: 1708 RT: 54.47 AV: 1 NL: 5.27E6T: + c d Full ms2 638.00 [ 165.00 - 1925.00]

200 400 600 800 1000 1200 1400 1600 1800 2000m/z

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

100

Re

lative

Ab

un

da

nce

850.3

687.3

588.1

851.4425.0

949.4

326.0524.9

589.2

1048.6397.1226.9

1049.6489.1

629.0

WR

A

C

VG

E

K

DW

LP

T

L T

WR

A

C

VG

E

K

DW

LP

T

L T

De Novo

AVGELTK

Database Search

Database ofknown peptides

MDERHILNM, KLQWVCSDL, PTYWASDL, ENQIKRSACVM, TLACHGGEM, NGALPQWRT,

HLLERTKMNVV, GGPASSDA, GGLITGMQSD, MQPLMNWE,

ALKIIMNVRT, AVGELTK, HEWAILF, GHNLWAMNAC,

GVFGSVLRA, EKLNKAATYIN..

Database ofknown peptides

MDERHILNM, KLQWVCSDL, PTYWASDL, ENQIKRSACVM, TLACHGGEM, NGALPQWRT,

HLLERTKMNVV, GGPASSDA, GGLITGMQSD, MQPLMNWE,

ALKIIMNVRT, AVGELTK, HEWAILF, GHNLWAMNAC,

GVFGSVLRA, EKLNKAATYIN..

51

Outline

• Motivation of proteomics

• Mass spectrometry-based proteomics

• Instrumentation: Mass Spectrometry

• De novo sequencing algorithm

• Database search

• Algorithms of real software (e.g., sequence tags)

52

Peptide Identification Problem

Goal: Find a peptide from the database with maximal match between an experimental and theoretical spectrum.

Input:

– S: experimental spectrum

– database of peptides

– Δ: set of possible ion types

– m: parent mass

Output:

– A peptide of mass m from the database whose theoretical spectrum matches the experimental S spectrum the best

53

Match between Spectra and the Shared Peak Count

• The match between two spectra is the number of masses

(peaks) they share (Shared Peak Count or SPC)

• In practice mass-spectrometrists use the weighted SPC

that reflects intensities of the peaks

• Match between experimental and theoretical spectra is

defined similarly

54

MS/MS Database Search

Database search in mass-spectrometry has been successful in identification of already known proteins.

Experimental spectrum can be compared with theoretical spectra of database peptides to find the best fit.

SEQUEST (Yates et al., 1995)

But reliable algorithms for identification of peptides is a much more difficult problem.

Q: Why can a peptide be not identical to a sequence in the database

55

Deficiency of the Shared Peaks Count

Shared peaks count (SPC): intuitive measure of spectral similarity.

Problem: SPC diminishes very quickly as the number of mutations increases.

Only a small portion of correlations between the spectra of mutated peptides is captured by SPC.

56

SPC Diminishes Quickly

S(PRTEIN) = {98, 133, 246, 254, 355, 375, 476, 484, 597, 632}

S(PRTEYN) = {98, 133, 254, 296, 355, 425, 484, 526, 647, 682}

S(PGTEYN) = {98, 133, 155, 256, 296, 385, 425, 526, 548, 583}

no mutations

SPC=10

1 mutation

SPC=5

2 mutations

SPC=2

57

Post-Translational ModificationsProteins are involved in cellular signaling and

metabolic regulation.

They are subject to a large number of biological modifications.

Almost all protein sequences are post-translationally modified and 200 types of modifications of amino acid residues are known.

58

Examples of Post-Translational Modification

Post-translational modifications increase the number of “letters” in amino acid alphabet and lead to a combinatorial explosion in both database search and de novo approaches.

59

Search for Modified Peptides: Virtual Database Approach

Yates et al.,1995: an exhaustive search in a virtual database of all modified peptides.

Exhaustive search leads to a large combinatorial problem, even for a small set of modifications types.

Problem (Yates et al.,1995). Extend the virtual database approach to a large set of modifications.

60

Modified Peptide Identification Problem

Goal: Find a modified peptide from the database with maximal match between an experimental and theoretical spectrum.

Input:

– S: experimental spectrum

– database of peptides

– Δ: set of possible ion types

– m: parent mass

– Parameter k (# of mutations/modifications)

Output:

– A peptide of mass m that is at most k mutations/modifications apart from a database peptide and whose theoretical spectrum matches the experimental S spectrum the best

61

Spectrum Alignment

• See 8.14 and 8.15 in the text book for one algorithm

• Complicated for real spectrums

62

Outline

• Motivation of proteomics

• Mass spectrometry-based proteomics

• Instrumentation: Mass Spectrometry

• De novo sequencing algorithm

• Database search

• Algorithms of real software (e.g., sequence tags)

63

Combining de novo and Database Search in Mass-Spectrometry

• So far de novo and database search were presented as two separate techniques

• Database search is rather slow: many labs generate more than 100,000 spectra per day. SEQUEST takes approximately 1 minute to compare a single spectrum against SWISS-PROT (54Mb) on a desktop.

• It will take SEQUEST more than 2 months to analyze the MS/MS data produced in a single day.

• Q: Can slow database search be combined with fast de novo analysis?

64

What Can be Done with De Novo?

• Given an MS/MS spectrum:

– Can de novo predict the entire peptide sequence?

– Can de novo predict a set of partial sequences, that with high probability, contains at least one correct tag?

A Covering Set of Tags

- No! (accuracy is less than 30%).

- Yes!

65

Peptide Sequence Tags

• A Peptide Sequence Tag is short substring of a peptide.

Example: G V D L KG V D V D L

D L K

Tags:

66

Filtration with Peptide Sequence Tags

• Peptide sequence tags can be used as filters in database searches.

• The Filtration: Consider only database peptides that contain the tag (in its correct relative mass location).

• First suggested by Mann and Wilm (1994).

• Similar concepts also used by:– GutenTag - Tabb et. al. 2003.

– MultiTag - Sunayev et. al. 2003.

– OpenSea - Searle et. al. 2004.

67

Why Filter Database Candidates?

• Effective filtration can greatly speed-up the process, enabling expensive searches involving post-translational modifications.

• Goal: generate a small set of covering tags and use them to filter the database peptides.

68

Summary

• Protein sequencing

• Mass spectrum

• De novo search and database search

• Difficulty of protein sequencing

69

The End

70

Quality Measure of Mass Spectrometer

• Sensitivity

• Mass accuracy

• Resolution

• Dynamic range

71

Exhaustive Search for Modified Peptides

• YFDSTDYNMAK

• 25=32 possibilities, with 2 types of modifications!

Phosphorylation?

Oxidation?

• For each peptide, generate all modifications.

• Score each modification.

72

Peptide Identification Problem: Challenge

Very similar peptides may have very different spectra!

Goal: Define a notion of spectral similarity that correlates well with the sequence similarity.

If peptides are a few mutations/modifications apart, the spectral similarity between their spectra should be high.

73

Why Filtration ?

Scoring

Protein Query

Sequence Alignment – Smith Waterman Algorithm

Sequence matches

Protein Sequences

Filtration

Filtered Sequences

Sequence Alignment – BLAST

Database

actgcgctagctacggatagctgatccagatcgatgccataggtagctgatccatgctagcttagacataaagcttgaatcgatcgggtaacccatagctagctcgatcgacttagacttcgattcgatcgaattcgatctgatctgaatatattaggtccgatgctagctgtggtagtgatgtaaga

• BLAST filters out very few correct matches and is almost as accurate as Smith – Waterman algorithm.

Database

actgcgctagctacggatagctgatccagatcgatgccataggtagctgatccatgctagcttagacataaagcttgaatcgatcgggtaacccatagctagctcgatcgacttagacttcgattcgatcgaattcgatctgatctgaatatattaggtccgatgctagctgtggtagtgatgtaaga

74

Filtration and MS/MS

Scoring

MS/MS spectrum

Peptide Sequencing – SEQUEST / Mascot

Sequence matches

Peptide Sequences

Filtration

Peptide Sequences

Database

MDERHILNMKLQWVCSDLPTYWASDLENQIKRSACVMTLACHGGEMNGALPQWRTHLLERTYKMNVVGGPASSDALITGMQSDPILLVCATRGHEWAILFGHNLWACVNMLETAIKLEGVF

GSVLRAEKLNKAAPETYIN..

Database

MDERHILNMKLQWVCSDLPTYWASDLENQIKRSACVMTLACHGGEMNGALPQWRTHLLERTYKMNVVGGPASSDALITGMQSDPILLVCATRGHEWAILFGHNLWACVNMLETAIKLEGVF

GSVLRAEKLNKAAPETYIN..

75

Filtration in MS/MS Sequencing

• Filtration in MS/MS is more difficult than in BLAST.

• Early approaches using Peptide Sequence Tags were not able to substitute the complete database search.

• Current filtration approaches are mostly used to generate additional identifications rather than replace the database search.

• Can we design a filtration based search that can replace the database search, and is orders of magnitude faster?

76

Asking the Old Question Again: Why Not Sequence De Novo?

• De novo sequencing is still not very accurate!

Algorithm Amino Acid Accuracy

Whole Peptide Accuracy

Lutefisk (Taylor and Johnson, 1997). 0.566 0.189

SHERENGA (Dancik et. al., 1999). 0.690 0.289

Peaks (Ma et al., 2003). 0.673 0.246

PepNovo (Frank and Pevzner, 2005). 0.727 0.296