Using Skyline to analyze the SPRG2013-2014 Targeted Proteomics Assay (TPA) standard ·...

Proteomics Standards Research Group (sPRG)

Using Skyline to analyze the SPRG2013-2014 Targeted Proteomics Assay (TPA)

standard

Brett S. Phinney, Ph D.

UC Davis Genome Center


What is the ABRF?

ABRF members are from over 300 international core laboratories in

academia, government, and industry, in a broad spectrum of biomolecular

technologies

The ABRF promotes the research, technology, communication and

education

ABRF has Research Groups and Committees, Affiliates and Chapters,

annual conferences with educational courses, a quarterly journal, a

Newsletter, Research Group publications, and Listserves

The ABRF is unique for providing benchmarking studies by its Research

Groups and has efficient mechanisms for networking and sharing


Proteomics Standards Research

Group (sPRG) 2013-2014

Christopher Colangelo (Chair) Yale University

Craig Dufresne Thermo Fisher Scientific

Alexander R. Ivanov (Ad hoc) Northeastern University

Toni Koller Stony Brook University

Brett Phinney (EB Liaison) University of California, Davis

Kristie Rose Vanderbilt University

Paul Rudnick National Institute of Standards and Technology

Brian Searle Proteome Software

Scott A. Shaffer University of Massachusetts Medical School

Brendan Maclean University of Washington SOM

David Hawke UT M.D. Anderson


• Design a comprehensive standard mixture of heavy-labeled tryptic peptides that can be used as internal standards for proteomics applications. • Widely applicable across human, mouse, and rat proteomes. • Proteotypic in nature. • Utility towards both discovery and targeted proteomics applications.

Study Rationale / Design


• Heavy/light peptide ratios will allow for normalization and evaluation for both inter- and intra- sample comparisons. • System suitability of LC-MS/MS performance including benchmarking and reproducibility of LC peptide elution profiles, evaluation of MS sensitivity and dynamic range. • Provide additional control for longitudinal proteomics studies and can be used in interlaboratory studies. • Peptides can be converted into absolute quantitation standards via cleanup, purification, and amino acid analysis.

Study Rationale / Benefits


• 1000 stable isotope labeled tryptic peptides conserved across Homo sapiens, Mus musculus and Rattus norvegicus.

• Peptide mixture will be spiked at fixed concentration to a HEK digest at 3 different dilutions (3 orders of magnitude).

• Participants will analyze the samples using LC-MS instrument platform of their choice, report ratios, return the questionnaire and raw data to the sPRG (via study anonymizer or possibly Panorama).

• Spectral library will be constructed, data searched, and peptide ratios determined for all data sets.

• Comparison of ratios, normalization of data, and peptide retention/elution order will be compared across data sets.

Study Design


Peptides selected via databases at Yale, NIST, Stony Brook

YPED; 115 LC-MS/MS HEK experiments • MASCOT SCORE > identity • BLAST search common to human, mouse, and rat • K or R tryptic cleavage (C-terminal)

6750 peptides, <1% to 67% observed in the samples • Removed peptides with upstream proline to K, R • Removed peptides with upstream KK, RR • Removed peptides with M • Removed peptides w/ miscleavages

1500 peptide candidates to send to JPT Peptide Technologies • JPT selected 1000 peptides for synthesis

Peptide Selection


0

50

100

150

200

250

300

350

400

1

43

85

12

7

16

9

21

1

25

3

29

5

33

7

37

9

42

1

46

3

50

5

54

7

58

9

63

1

67

3

71

5

75

7

79

9

84

1

88

3

92

5

96

7

spec

tral

co

un

ts

Charge State (NIST) Dynamic Range (YPED) observed spectral counts

0

100

200

300

400

500

1 23

105

446

1

Peptides Map to 552 proteins

peptide number

# peptides mapping to each protein

Peptide Selection and

Other Criteria


0

200

400

600

800

1000

1200

0

10

20

30

40

50

60

70

80

90

100

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Nu

mb

er

of

Pe

pti

de

s >

Id

en

tity

(1%

FD

R)

Perc

en

tag

e o

f A

BR

F P

ep

tid

es I

den

tifi

ed

Sample

% Coverage (> Homology) Peptide score > Identy score

Peptide Identification in

ABRF Labeled Peptide Standard Mix


Using Skyline for AUC analysis

• 1000 peptides + 5 ug HEK lysate

– We decided on AUC analysis as we did not have enough time to develop a scheduled MRM method

– Still possible for someone out there to develop a scheduled MRM method if you want

• Might be a good test for RT prediction algorithms


Using Skyline for AUC analysis

• 1 pmol peptide mix spiked into 5 ug HEK 293 lysate

• Set MS1 resolving power at 70K

– Q-Exactive Data


Example Peptide GGSASVWSER (523.2505, 2+)


Example Peptide GGSASVWSER (523.2505, 2+)

Isotope Dot Product idot = 0.81

Isotope Dot Product idot = 0.96

Retention Time from

Library


Lets look at raw data

RT: 0.00 - 120.00

0 10 20 30 40 50 60 70 80 90 100 110 120

Time (min)

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

25.5128.28

23.5720.89

20.36

18.17

31.8516.45

35.41

38.4615.64

71.59

39.81

60.40

44.7749.37 51.09

41.2695.2860.02 63.47

95.07 95.4553.4014.78

55.2763.68 91.06

66.21

13.94

13.20

69.95

107.83

12.46 110.91112.7784.15 102.0486.0575.49 77.60

12.06 79.707.793.56

NL:5.92E8

Base Peak MS qe2_20120108_6SPRGheavy


XIC 523.25-523.27 m/z

RT: 15.50 - 33.54

16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33

Time (min)

0

5

10

15

20

25

30

35

40

45

50

55

60

65

70

75

80

85

90

95

Re

lative

Ab

un

da

nce

19.52

19.47

19.55

26.1823.22

18.74

26.22

26.3326.0721.78

21.70

26.3631.4521.6416.32

26.0321.61 24.0631.31 31.5518.5116.25 20.56 32.2325.9321.56 26.44 29.3422.09 32.8528.26 31.2525.01 29.4017.80 27.8617.73

NL: 1.70E8

Base Peak m/z= 523.25-523.27 F: FTMS + p NSI Full ms [400.00-1600.00] MS qe2_20120108_6SPRGheavy


MS Spectra (522-527 m/z) Peak at 23 min

qe2_20120108_6SPRGheavy #6572 RT: 23.20 AV: 1 NL: 1.87E8T: FTMS + p NSI Full ms [400.00-1600.00]

522.5 523.0 523.5 524.0 524.5 525.0 525.5 526.0 526.5 527.0

m/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

522.75525z=2

523.25641z=2

524.79266z=2

523.75702z=2

525.29376z=2522.30762

z=? 524.25757z=2 525.79553

z=2522.91669z=?

526.29596z=2

526.96307z=?

523.25641 A+1 isotope idotp = 0.96


MS Spectra (522-527 m/z) Peak at 26 minutes

qe2_20120108_6SPRGheavy #7683 RT: 26.16 AV: 1 NL: 1.11E8T: FTMS + p NSI Full ms [400.00-1600.00]

522.5 523.0 523.5 524.0 524.5 525.0 525.5 526.0 526.5 527.0

m/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

523.25079z=2

523.75232z=2

525.29523z=?

526.79504z=2

524.25360z=2 525.79706

z=2524.79248

z=?522.76361

z=? 526.29883z=2

522.25598z=?

526.91724z=?

525.71985z=?

523.25079 Correct isotope idotp = 0.81


Precautions

• Skyline’s Isotope Dot Product (idotp) is actually better for the wrong analyte

– This can get a bit tricky when you are analyzing 1000 peptides

– Probably best to not rely on just idotp

– We have found that getting the RT’s correctly in the library really is critical


Making a library that has proper RT’s

• This has been very finicky in our hands

• Some workflows work, some don’t

– Different combinations of software all seem to affect this

• Peak List generator (mzML…mzXML….)

• Searching software (x!tandem, Mascot...)

• Meta file has to have the same name as the raw data file (except for the extension)


Generating libraries with correct RT’s (our workflow)

• Generate mzML using proteome discoverer

• Combine all peptides into an artificial protein

• Search with x!tandem (turning on the skyline option!)

• Rename the x!tandem output files to extension *.xtan.xml

• Make Library in skyline


Correct library

File Name

RT


HEK lysate Non-Co-eluting labeled

peptides??


More non co-eluting pairs


And more


Lets look at this one


MS spectra very

complex

qe2_20120108_10SPRGheavyHEK #13810 RT: 37.50 AV: 1 NL: 3.97E7T: FTMS + p NSI Full ms [400.00-1600.00]

400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600

m/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

521.31274z=2

555.31287z=2

461.28021z=2

608.81256z=2

659.97528z=3

709.84430z=2

756.42395z=2

864.37903z=2 921.55170

z=1 989.45789z=2

1109.61719z=1

794.92877z=2

1216.61597z=1

1296.76257z=?

1527.34009z=?

1410.73596z=?


Very complex


705 706 707 708 709 710 711 712 713 714 715 716 717 718

m/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

709.84430z=2

710.34546z=2

714.35980z=2

712.33911z=?

714.86041z=2

710.84686z=2

712.84064z=?

711.33990z=2

706.39160z=2

711.84100z=2706.89484

z=2715.36218

z=2713.34332z=?705.34259

z=?709.34143

z=? 715.86237z=2

713.88672z=?

707.39459z=2705.84265

z=?708.35510

z=? 717.34222z=?

716.36603z=2

Heavy Peptide

Light Peptide?


Light HEK peptide?


704.5 705.0 705.5 706.0 706.5 707.0 707.5 708.0 708.5

m/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

706.39160z=2

706.89484z=2

705.34259z=?

704.83911z=?

707.39459z=2

705.84265z=?

708.35510z=?

707.67383z=?

m/z =704.83911 Thero m/z =704.8408 Delta = 2.39 ppm


MS/MS spectra

• None Acquired for light peptide

• Is this the correct light peptide?

– Probably not

• Next steps

– Need to acquire with an isolation list or a MRM or Data Independent method??? Maybe use a much longer gradient or LC-LC-MS/MS


Skyline analysis

• Can take a lot of analysis time to go over 1000 peptides in skyline in full scan mode

• Need to consider adding additional data metrics to reduce need to manually check each integration

• Figuring out if the light peptide of a heavy/light pair is real or interference is a real problem. Especially the more complex your data is.

• sPRG sample is probably too complex to analyze by Full scan in one 2 hour LC-MS/MS run


•Progress has been made for the 1st year in development of an innovative proteomics standard.

•Study sample is designed to represent proteins spanning three orders of magnitude in concentration.

•1,000 isotope labeled peptides were synthesized and analyzed. Greater than 99% of the peptides were identified during the validation runs.

•In peptide dilution experiments, the majority of peptides behaved as expected with a linear response; however, a few peptides had a non-linear response.

•The peptides were spiked into a HEK digest and showed minimal variation in both retention time and peak area.

•The sPRG is hopeful that the designed formulation will become a valuable resource in various mass spectrometry-based proteomic applications, including quantitative and differential protein profiling, interlaboratory studies, as well as general LC-MS/MS benchmarking.

Conclusions


How to Get a sample

• E-mail

– [email protected]

• Samples are limited so be certain you can analyze the sample if you ask for one.

mailto:[email protected]


Acknowledgments

• SPRG

Current Membership

Dr. Christopher Colangelo (Chair) - Yale University Dr. Craig P. Dufresne - Thermo Fisher Scientific Dr. Alexander R. Ivanov - Northeastern University Dr. Antonius Koller - Stony Brook University Brendan MacLean - University of Washington Dr. Kristie L. Rose - Vanderbilt University Medical Center Dr. Paul A Rudnick - NIST Brian C. Searle - Proteome Software Inc. Dr. Scott A. Shaffer - University of Massachusetts Medical School Dr. Brett S Phinney (EB Liaison) - Proteomics Core UC Davis Genome Center Dr. David Hawke – MD Anderson Caner Center

http://www.abrf.org/index.cfm/dir.wpDetail?entID=4360















Acknowledgments part 2

• JPT Peptide

• ABRF (funding) www.abrf.org

• Dr. Rich Eigenheer UC Davis (UC Davis Data)

Using Skyline to analyze the SPRG2013-2014 Targeted Proteomics Assay (TPA) standard ·...

Documents

Transcript of Using Skyline to analyze the SPRG2013-2014 Targeted Proteomics Assay (TPA) standard ·...