<img alt="" src="https://secure.soil5hear.com/223550.png" style="display:none;">
Contact Us Select Models
Select Models
2204875890

Cell-surface proteomics

Know whether a target is on the surface before you build against it

Matched whole-cell and cell-surface proteomes from 600+ low-passage patient-derived tumor models, most from pretreated patients, measured from a separately enriched membrane fraction.

600+ surface-profiled modelsMatched to whole-cell dataPretreated patient tumors

Overview

Transcript levels often miss what a binding agent will actually find on the cell surface

For an antibody-drug conjugate, radioligand or other surface-directed modality, the question is not whether a gene is expressed. It is whether the protein is present, correctly localized and accessible. RNA cannot answer that, and whole-cell proteomics averages surface and internal pools together, hiding the signal for proteins that traffic between them.

We enrich and quantify the membrane fraction separately, then compare it with whole-cell abundance on the same tumor. Surface localization is measured rather than inferred, and each candidate arrives with its transcript, whole-cell protein and clinical context from one model.

Because most models come from pretreated patients, the data maps the targets tumors present after therapy, which is where many conjugates will be used in the clinic. It supports target discovery, target validation, biomarker work and model selection for surface-directed programs.

Cell-surface proteomics at a glance

Surface-profiled
600+ models
Whole-cell protein
1,000+ models
Acquisition
Ultra-deep DIA LC-MS
Source tumors
Low-passage PDX
iStock-1206633493

RNA vs protein

Most transcripts are a weak proxy for the protein you plan to target

Across our tumor models, most transcripts do not correlate strongly with protein abundance. In primary AML models, only about 17% of genes showed a positive correlation between RNA and protein. B7-H3 surface abundance tracks its RNA weakly across tumor types, and least in breast and ovarian models.

The relationship is target-specific. MET, EGFR and CEACAM5 show strong RNA to surface protein concordance in NSCLC models, while for B7-H3 the link is weak. Ranking candidates on RNA alone can promote targets that are barely present and miss ones that are.

Method

From tumor to a ranked surface target, checked at every step

Each step has a defined quality gate, so a surface call can be traced back to the fraction and the model it came from.
  1. 01

    Low-passage tumor

    Tissue is taken from low-passage patient-derived models, preserving the source tumor's biology and its clinical record.

    Exit criterion — tissue traced to an annotated source model.

  2. 02

    Membrane enrichment

    A proprietary enrichment protocol separates the membrane fraction from the rest of the cell before any measurement is made.

    Exit criterion — membrane and cytosolic fractions prepared separately.

  3. 03

    Deep acquisition

    Both fractions and the whole-cell lysate are acquired by ultra-deep data-independent acquisition mass spectrometry.

    Exit criterion — protein abundance quantified in each fraction.

  4. 04

    Localization QC

    Canonical membrane markers must enrich and intracellular markers deplete, confirming the fraction behaved as intended for that sample.

    Exit criterion — canonical membrane and intracellular markers separate.

  5. 05

    Matched ranking

    Surface candidates are ranked with whole-cell protein and transcript data from the same model shown alongside each call.

    Exit criterion — each candidate reported with matched context.

Validation

Surface calls checked against readouts your team already trusts

A surface proteomics value is only useful if it agrees with orthogonal measures. These examples come from our own data, and the analyses behind each are available to review under CDA.

CDH17

Cell-surface abundance of CDH17 predicts immunohistochemistry staining intensity in the same models, linking the measurement to a familiar pathology readout.

Agrees with IHC

TROP2

Surface enrichment of TROP2 correlates strongly with the in vivo activity of datopotamab deruxtecan, connecting measured target presence to conjugate response.

Tracks ADC activity

Marker check

Canonical membrane proteins such as ATP1A1 and CD81 enrich in the membrane fraction while GAPDH and ENO1 deplete, sample by sample.

Checked per sample

KRAS resistance

In KRAS-mutant models treated with daraxonrasib, surface data flagged resistance-associated proteins that whole-cell data detected weakly or not at all.

Seen only at the surface

Target discovery

Start with an indication, end with a target package

Name the indication your program cares about. We run the analysis across the matching models and return a target report: each candidate with its prevalence, RNA, protein and membrane expression, a normal-tissue safety screen, validation scores and the models available to validate it. Every surface call is labeled as measured or inferred.
Ask about target packages
01

You set the scope

Choose the indication. The analysis runs per indication, so prevalence and response association are never pooled across the bank.

02

We rank the candidates

Candidates are ranked on surface abundance, localization and specificity over normal tissue, with whole-cell and RNA data alongside.

03

You receive the evidence

Each report is checked by a named scientific reviewer and lists the models that express each target, ready for validation.

2174551157

Get started

Access the Champions Data Room

For partners ready to explore the depth of our data, Champions provides secure access to our Data Room. Here you can evaluate sample datasets, review enrichment metrics, and see how our cell surface proteomics integrate with broader multiomic profiles.

  • Discover how surface expression correlates with treatment response.

  • Explore new target opportunities validated in real human tumors.

  • Assess the quality and interpretability of our data for your own programs.

FAQ

What conjugate teams ask about surface data

Questions that come up when a team evaluates surface proteomics for target selection.

No. RNA-based approaches infer surface presence from expression and annotation. We measure protein in a separately enriched membrane fraction and compare it with whole-cell abundance on the same tumor, so localization is observed rather than predicted from a database.

Whole-cell measurements include protein from the surrounding mouse stroma as well as the tumor. Species assignment uses peptide-level evidence, and we share the search parameters and assignment approach under CDA so your team can judge it.

Yes. Fractionation and acquisition can be run on any model in the bank, so a study can include models selected for clinical or molecular fit rather than only those already profiled. Timelines depend on the number of models.

Yes. Surface and whole-cell proteomes can be licensed with matched RNA and exome data on the same models, delivered into your environment for target ranking or model training, with per-model clinical annotation and documented methods.

Related pages

Where surface data feeds in

Surface calls are most useful alongside the rest of the matched dataset and an in vivo readout.
2204875890

Get started

Let’s Build the Future of Cancer Research Together

Take this opportunity to align your platform, pipeline, or portfolio with the dataset built for modern drug discovery.