Last Updated
10 March 2021

US HUPO: Deep plasma proteomics with trapped ion mobility and the Proteograph Product Suite

1,000 800 600 400 200 0
Proteome
Dynamic indicator of status
High utility but low accessibility of content
Proprietary Engineered Nanoparticles
1B Data Independent Acquisition (DIA) 2B Precision of DIA Method (Peptide CVs) DDA-Based Spectral Library
3B
Protein Abundance (High to Low)
Plasma
Proteins
Nanoparticles
LC/MS Analysis
Protein Coronas
Digestion < 200ng
Figure 3. Dynamic range of identified proteins matched with normalized protein intensities from the previously published data2. A) In DDA mode, the dynamic range coverage is improved by increasing the gradient length, up to one order of magnitude compared to the shortest gradient. B) In DIA mode, the dynamic range coverage is similar using both experimental library-based and library-free approaches. A slight improvement was observed when increasing the gradient to 20 min or longer.
Genome
Static indicator of risk
High accessibility of content but low utility
20 min
45 min
Data Analysis
Tryptic Peptides
0.20
0.36 0.75
0.41
5.43
3.36
0.83
Utility of Information
10M+ human exomes 1M+ genomes
< 0.2%
Core Technology
genetic variants catalogued
150
100
50
29.8
0
-2
-4 -6
-8
-4.5 -4.3
8 min
-4.7 -4.7 -4.7
MaxQuant
E.coli:Plasma Plasma:E.coli
Plasma:Plasma
ProteographTM multi-nanoparticle proteins coronas enable deep plasma
proteomics studies at scale with unmatched sensitivity in combination with
trapped ion mobility
Shadi Ferdosi1, Tristan Brown1*, Patrick A. Everley1, Michael Figa1, Matthew McLean1, Eltaher M. Elgierari1, Xiaoyan Zhao1, Veder J.Garcia1, Tianyu Wang1, Matthew E.K. Chang2, Kateryna Riedesel1, Jessica Chu1, Martin Goldberg1, Mark R. Flory2, Asim Siddiqui1, Juan Cruz Cuevas1, Nagarjuna Nagaraj3, Tharan Srikumar3, Michael Krawitzky3, Christopher Adams3, John E. Blume1, Daniel Hornburg1 and Omid C. Farokhzad1
1Seer, Inc., Redwood City, CA 94065, USA; 2Oregon Health & Science University, Portland, OR 97239-3098, USA; 3Bruker Daltonik GmbH, Fahrenheitstr. 4, 28359 Bremen, Germany
Proteograph Product Suite Delivers Unbiased, Deep and Rapid Proteomics at Scale
Blood plasma is a rich, readily available source of proteins that is commonly used in clinical profiling studies. However, plasma proteomics is inherently
constrained by the large dynamic concentration range and complexity of the proteome. The ability to overcome these constraints while interrogating deeply and broadly into the plasma proteome has only been partially addressed by laborious, unscalable and low throughput workflows. To fully enable high- throughput plasma proteomics, we have developed a quantitative profiling solution, Proteograph Product Suite1 that consists of a panel of 5 nanoparticles (NPs) with distinct physicochemical properties. This panel of NPs is used in parallel to enable high-performance plasma proteomics combining depth and breadth with precise and reproducible quantification.
Here we explore the synergy of the Proteograph assay with the Bruker timsTOF Pro mass spectrometer (MS), including; LC gradients making use of both data- dependent and data-independent acquisition (i.e., DDA and DIA) workflow coupled with our five nanoparticle Proteograph Assay using a control plasma sample.
Proteograph proteome profiling with the timsTOF Pro mass spectrometry was then evaluated in respect to depth of coverage and analysis throughput. We also investigated the linearity of response employing a multi-level proteome spike-in experiment.
Unbiased, Deep and Rapid Method for Plasma Proteomic Analysis at Scale with the Proteograph Product Suite and the Bruker timsTOF Pro Mass Spectrometer
Deep and Rapid LC/MS Method for Plasma Proteomics
~695M
of genetic variants fully characterized
1309 1331
Proteograph Product Suite
Sample is ready to be analyzed on most LC/MS instruments
Figure 4. Evaluation of accuracy and precision. In a spike-in E.coli:Control Plasma experiment, a 2, 10 and 50-fold range of experimental ratio was analyzed with a good accuracy across a wide dilution range in a 90 min DDA run.
Seer, Inc., Redwood City, CA – *tbrown@seer.bio
Seer, Proteograph and the Seer logo are trademarks of Seer. All other trademarks are the property of their respective owners.
1A Data Dependent Acquisition (DDA) 2A −
Precision of DDA Method (Peptide CVs)
3A
Protein Abundance (High to Low)
Ratio
10x 2x 50x
10x 2x 50x 10x 2x 50x
10x 2x 50x

-4.9
90 min
-4.7




1731 100 50
-3.9 -4.3
8 min 20 min
-4.5
45 min
1056 592
-6 0 -8

8 min
853
FASTA –Based Spectral Library

150
-2
-4
6.1
90 min 8 min 20 min 45 min 90 min
1294 1323
20 min
across different gradients. A) In DDA mode, protein intensities. A) In DDA mode, the median CV of ≤ 6.2 was identification is improved by increasing the length of observed across different gradient lengths. B) In DIA the gradient. B) In DIA mode, protein identification mode, significantly better CV distribution was observed was similar using 20 min and 45 min gradients and than in the library-free approach. The 20 min gradient significantly higher than the 8 min gradient. A library- had the lowest median CV when compared to the other free approach resulted in similar identifications when gradient lengths.
compared to a DDA-based spectral library approach.
Quantitation Accuracy with Optimized DDA Method
991 945
8 min
15.5 21.2
9.1 22.9 9.9
45 min 8 min
Figure 1. Comparison of protein identification Figure 2. Replicate CV for median-normalized peptide
E.coli:E.coli
6.11 3.3
0.88
0.82
Conclusions
 For TIMS-TOF DDA analysis, protein coverage improves with increasing gradient length, with a median peptide CV ≤ 6.2
 In DDA the median dynamic range of a 90 min method improved by an order of magnitude when compared to the 8-min analysis
 TIMS-TOF Pro DIA data vs. DDA showed improved protein coverage when comparing same gradient lengths
 DIA data analysis using a library-free approach (MaxQuant ref) potentially omits the need for generation of experimental spectral libraries by maintaining the protein identification information and improving the quantification precision. It also highlights the utility of this approach for multi-species proteomics such as microbiome where an unknown proteome may be present in the samples without prior knowledge
 Proteome spike-in experiments showed a good quantitation accuracy across 2–50 fold changes
 Proteograph Product Suite in combination with the timsTOF Pro provides a high- performance combination for rapid deep, precise, and accurate proteome profiling for biomarker discovery and biomedical research
References
1. Blume et al. Nat. Comm. (2020)
2. Keshishian et al. Nature Protocols (2017)
0.38
6.0
6.2
5.5
20 min
45 min
20 min 45 min
Log2 (Ratio)
Cataloged variants (mm)
Relative Log 10 Intensity Relative Log 10 Intensity