Protein Valida-on (Sta-s-cal Inference) and Protein Quan-fica-on
Terminology Pep-de Spectrum Match Target / Decoy False discovery rate Shared pep-de Parsimony One hit wonders
SPECTRUM Rela-ve Abundance 200 400 600 800 1000 1200 m/z 3
Tandem mass spectrometry and pep/de sequence y-series: b-series: y 14 H 2 N-N--S--G--D--I--V--N--L--G--S--I--A--G--R-COOH b 1 y 13 y 12 y 11 y 10 y 9 y 8 y 7 y 6 y 5 y 4 y 3 y 2 y 1 b 2 b 3 b 4 b 5 b 6 b 7 b 8 b 9 b 10 b 11 b 12 b 13 b 14 Relative Abundance 200 400 600 800 1000 1200 m/z An MS/MS spectrum contains a mixture of b and y ions 4
PEPTIDE SPECTRUM MATCH Mass spectrum Pep/de Spectral Match Reference Protein Database from genomic annotation
PEPTIDE SPECTRUM MATCH Visual inspection Scoring function f(s,p) measures the quality of the match between spectrum S and peptide sequence P Look at fragment ions predicted mass expected intensity compare to next best presence of immonium ions etc. Nesvizhskii et al Nature Methods - 4, 787-797 (2007) 6
PEPTIDE SPECTRAL MATCH Nesvizhskii et al Nature Methods - 4, 787-797 (2007) 7
TARGET-DECOY SEARCH
TARGET-DECOY SEARCH INPUTS : PEAKLISTS and SEARCH db SEARCHGUI PEPTIDESHAKER 9
PROTEIN INFERENCE: FROM PEPTIDES TO PROTEINS Slide from Alexey Nesvizshkii talk at hwp://www.scivee.tv/node/12671 10
Nesvizhskii, A. I.; Aebersold, R. Mol. Cell. Proteom. 4.10, 1419-1440, 2005 Shared Pep-des
Shared Pep-des YMACCLLYR SIQFVDWCPTGFK Tubulin alpha 6 Tubulin alpha 4 Tubulin alpha 3 85% 85% 85%
AEPTIR 85% From Pep-des to Proteins IDVCIVLLQHK 65% Protein (4%) NTGDR 25% 0.15 * 0.35 * 0.75 = 0.04 Feng J, Naiman DQ, Cooper B. Anal Chem. 2007 May 15;79(10):3901-11.
One Hit Wonders IDVCIVLLQHK 95% Protein (5%)
One Hit Wonders Quan-fy as a score: If different pep-des agree: Good! If pep-des are one- hit- wonders: Bad! Protein Prophet, etc. Then can use FDR at protein level
Pep-de / Protein Quan-fica-on
Terminology Absolute Quan-fica-on Rela-ve Quan-fica-on Label- free Normaliza-on Labelled itraq Reporter ions
Pep-de / Protein Quan-fica-on Absolute Es-mate the molar amount of protein / pep-de in the biological sample PTMs Valida-on Rela-ve Fold change / sta-s-cally significant difference between 2 biological states Biological varia-on Biomarker studies
Label- free Area Under Curve MS1 Integrate XIC Spectral Coun-ng MS2 High abundant proteins Käll L, Vitek O (2011) Computa-onal Mass Spectrometry Based Proteomics. PLoS Comput Biol 7(12): e1002277. doi:10.1371/journal.pcbi.1002277
MS Quan-fica-on MS not inherently quan-ta-ve Physiochemical proper-es invoke different MS responses MS only samples a small percentage of total pep-des Bias and variability Population Biological Instrument Sample Handling
Normaliza-on Remove bias and variability between runs Global commonly used Median scale Total ion current (TIC) Local very recent development Proximity- based intensity normaliza-on (PIN)
Labeled Quan-fica-on Run samples simultaneously on in a single run Add label to samples Mix samples together Compute ra-os / sta-s-cally significant diffs.
Labeled Isobaric MS2, itraq Number of samples Synthe-c Pep-des MS1 Absolute (AQUA) Metabolic MS1, SILAC Not higher life forms Käll L, Vitek O (2011) Computa-onal Mass Spectrometry Based Proteomics. PLoS Comput Biol 7(12): e1002277. doi:10.1371/journal.pcbi.1002277
itraq 8- Plex Reagent Chemical Structure Isobaric Tag Total mass = 305 Reporter Group 113 119, 121 m/z N N ~~~~~~~ O N O Amine specific pep/de reac/ve group (NHS) N- hydroxysuccinimide O Balance Group (?) Mass 184, 186 192 m/z Isobaric Reporter Groups 113 119, 121 m/z N N N N N+ N+ N+ N+ 113 114 115 116 13 CH 3 13 CH 3 13 CH 3 N 15 N 15 N 15 N 15 N+ 15 N+ 15 N+ CH 2 15 N+ 117 118 119 121 Applied Biosystems has granted permission to use this slide.
itraq Experiment Obtain protein- containing sample, extract protein Normal Brain Tissue- 1 Normal Brain Tissue- 2 Huntington s Disease-1 Huntington s Disease-2 Reduce, alkylate Cysteines Reduce, alkylate Cysteines Reduce, alkylate Cysteines Reduce, alkylate Cysteines Proteoly-c Diges-on Trypsin Digest Trypsin Digest Trypsin Digest Trypsin Digest Label pep-des with itraq Reagents itraq TAG 114 itraq TAG 115 itraq TAG 116 itraq TAG 117 MIX 2D LC- MS/MS Tissue Images: Rosas HD et al, (2002) Neurology, 58, 695
itraq Experiment MS2 Spectrum Intensity, counts Intensity, counts +TOF Product (513.3): Experiment 2, 33.388 min from keyxx001_perez061_fxn13_queb2_1004.wiff a=3.56649482550429800e-004, t0=4.09124284567078580e+001 1200 1100 1000 900 800 700 600 500 400 114.1 117.1 145.1 y1 175.1 185.1 216.2 a1 255.1 b3 428.3 b1 y9(+2) y8 244.2 782.5 y4 513.3 400.2 a3 212.1 b2 315.2 Pep-de itq VAIVVGAPR MWPeptide mono = VAIVVGAPR 1024.62 Protein MW Match = 1024.6249 Platelet Protein ID = platelet membrane glycoprotein glycoprotein 11b 11b a4 y5 499.3 y6 a5 598.4 Max. 1224.0 counts. Reporter Ion Mass Tags from which quantitation is calculated Pep-de match is made from product ions, e.g., b- and y- ion series 300 y9 y2 y7 881.6 200 272.2 711.4 I 70.1 I I 383.3 b4 86.1 100 y3 527.4 199.1 230.2 343.2 326.2 412.3 440.3 0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 m/z, amu y9
112.72493 113.91559 115.10626 116.29692 117.48759 118.67826 itraq Results Reporter ion intensi-es reflect rela-ve pep-de amounts control disease 100 115.13 976.9 90 80 114.13 116.13 117.13 70 % Intensity 60 50 40 No change disease:control 30 20 10 113.11 0 Mass (m/z)
112.74415 113.88269 115.02123 116.15977 117.29831 118.43685 itraq Results What fold changes are significant? Do they represent biological relevance as opposed to experimental variability? 100 control disease 116.14 117.14 1.1E +4 90 80 70 Increase disease:control % Intensity 60 50 40 30 114.14 115.14 20 10 0 113.12 Mass (m/z) 100 3070.1 90 control 114.12 115.12 disease 80 70 Decrease disease:control % Intensity 60 50 40 30 116.12 117.12 20 10 113.11 0 112.52129 113.84865 115.17602 116.50338 117.83075 119.15811 Mass (m/z)
Labeled & Label- free From Käll L, Vitek O (2011) Computa-onal Mass Spectrometry Based Proteomics. PLoS Comput Biol 7(12): e1002277. doi:10.1371/ journal.pcbi.1002277