<img alt="" src="https://secure.soil5hear.com/223550.png" style="display:none;">
Contact Us Select Models
Select Models
iStock-1310232039

Champions Oncology data platform

Response and multi-omic data from the same tumor, not stitched together

Matched multi-omic, clinical and drug-response data from 1,500+ patient-derived tumor models, delivered as licensed data or as the answer to the question your program is asking.

1,500+ characterized models65% from pretreated patientsMatched by donor

Overview

Most oncology data is either molecular or functional, and the link between them is guessed

Public datasets such as TCGA, DepMap and CPTAC were built for different questions. They profile primary, mostly untreated tumors or immortalized cell lines, and few carry drug response measured on the same sample. When the response and the molecular profile come from different tumors, every attribution between them rests on an assumption you cannot test.

We profile and dose the same patient-derived models. Exome and RNA sequencing, whole-cell and cell-surface proteomics, IHC, clinical history and in vivo response are matched by donor, so a candidate marker and the response it explains come from one tumor. Because the models are renewable, you can return to that tumor and test the next hypothesis in it.

Most models come from late-stage, pretreated patients, so the bank carries the resistance biology that public data largely misses. You can license the data into your own environment, or have our scientists run the analysis and deliver the result as a defined package, starting with target discovery.

data_tree_720

At a glance

Models
1,500+ characterized
Pretreated
65% of models
Late stage
80% of models
Since 2007
400+ biopharma partners
View study package

What you can ask

Four questions that need response and biology from the same tumor

Each of these can be approached with unmatched data. Each becomes easier to defend when the molecular profile and the functional result were measured on one tumor, from a patient with a recorded treatment history.

Target discovery

Is the target present, localized at the surface and accessible in tumors that have already seen therapy? Matched surface and whole-cell proteomics answer it directly.    

Ranked surface candidates

Biomarkers

What separates responders from non-responders? Baseline profiles are matched to response on the same models, so the marker and the response it explains share a tumor.

Candidate response markers

Resistance

Why does a response stop? Profiling through treatment and into acquired resistance shows the mechanism as it develops, and points to combinations worth testing.

Combination hypotheses

Benchmarking

How does your compound compare to the standard of care? We run it against a concurrent reference arm on the models you select, not against historical controls.

Head-to-head response data

Ways to work

Take the data into your stack, or take the answer

The same models, molecular layers and analytics sit underneath every route. What changes is who runs the analysis and what you receive at the end.
pharma_stock

New

Target discovery packages

Name the indication you care about. We return a target report for it, with each candidate's prevalence, RNA, protein and membrane expression, a normal-tissue safety screen, validation scores, the competitive landscape and the models available to validate it.

Every surface call is labeled as measured by direct membrane proteomics or inferred from other data, and every report is checked by a named scientific reviewer before it is issued.

Best suited to

  • Teams building a target pipeline
  • ADC and radioligand programs
  • Programs without in-house omics teams
Starts with
Indication
Delivered as
Target report
Ask about target packages
biotech_stock

Analysis service

Answers to your program's question

We design the study around your question, select the models, run your compound in vivo where needed and analyze the result against the matched molecular data already held on those models.

You receive a report with the figures, tables and methods behind each conclusion. The data underneath is available on request, so any finding can be traced back and checked.

Best suited to

  • Response and resistance mechanism
  • Biomarker and stratification hypotheses
  • Benchmarking against standard of care
Starts with
Your question
Delivered as
Study report
Scope an analysis
tumor_bank_stock

Data licensing

Matched data in your environment

Curated cohorts are licensed by indication or data layer and delivered into your own systems, ready to enrich internal and public datasets with deeply characterized, donor-matched tumors.

Licenses are scoped by cohort and term. Layers can be generated on demand for models you select that have not yet been profiled, so the cohort fits your question.

Best suited to

  • Computational and data-science teams
  • Training and validating your own models
  • Enriching public reference data
Starts with
Your cohort
Delivered as
Licensed data
Discuss a license

How it grows

Each study adds to the data already held on the same tumor

Depth accumulates on models you have already validated, so each result builds on the last rather than starting a new characterization study.
  1. 01

    Select the models

    Choose by indication, prior therapy, stage, biomarker or target expression, against data that already exists before anything is dosed.

    Exit criterion — models chosen against existing molecular and clinical data.

  2. 02

    Run the perturbation

    Your compound, combination or schedule runs in vivo, with a concurrent reference arm wherever a comparator matters to the decision.

    Exit criterion — response measured beside a same-model reference arm.

  3. 03

    Match the response

    The response joins the molecular baseline already held on the same models, with no bridging across separate cohorts.

    Exit criterion — response linked to baseline data on each model.

  4. 04

    Go deeper

    Add profiling under treatment, sampling into resistance or surface proteomics on the models that matter most to the program.

    Exit criterion — new layers generated on models already validated.

  5. 05

    Test it again

    Take the hypothesis the data produced back into the same biological system, ex vivo or in vivo, and test it there.

    Exit criterion — hypothesis tested in the tumor it came from.

Clinical relevance

Models that reflect the patients your trial will enroll

From model sourcing to final study report, every step runs under one roof.

Our models come from patients with recorded treatment and response histories: 80% late stage, 65% pretreated and 55% metastatic. That matters because resistance biology is shaped by prior therapy, and models built from untreated primary tumors cannot show it.

In matched patient cases, the models predicted clinical outcome with a positive predictive value of 85% and a negative predictive value of 91% (Izumchenko et al., Annals of Oncology, 2017). Ask us for the paper, cohort and endpoints behind those figures.

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

Frequently Asked Questions

We profile it. Molecular layers are generated on demand, so model selection is driven by fit to your question rather than by which models already carry data. Tell us the models and layers you need and we will come back with a timeline and a price.

Not without your explicit agreement. Ownership and use of data generated in your study are set out in the study agreement before work starts, so raise any constraints early and we will scope the work to them.

Around 90% of models come from patients in the US and Western Europe, sourced through our clinical network with documented consent and traceable to the source tumor. Prior therapies and responses stay linked to each model in the order the patient received them.

It depends on the question. Surface target work can start from data that already exists, response studies are paced by the in vivo arm, and resistance studies take longest because they are longitudinal. We scope each one against the models and layers involved.

PDX samples are aligned to a combined mouse-human genome before variant calling and expression quantification, so stromal reads are removed rather than left to inflate the signal. Whole-cell proteomics includes stromal protein, and we share our species-assignment approach under CDA.

Related pages

Go deeper into the platform

Each page below covers one part of the platform in the detail a technical review needs.
iStock-1310232039

Start here

Send the question and the indication

Tell us what you are trying to find out, the indication or mutation, and whether you have a compound in hand. If you want data for your own models, say which layers and roughly how many models. We will tell you whether these models can answer it.

  • A data scientist replies to every request personally

  • Model and layer availability included in the reply

  • CDA handled before any program detail is shared