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

KRAS-Mutant Atlas

Benchmark your KRAS program against daraxonrasib, model by model

A reference atlas of response, resistance and multi-omic profiling across 50+ KRAS-mutant patient-derived tumor models, each run under daraxonrasib, ready to compare your compound against.

50+ KRAS-mutant modelsNSCLC, colorectal, pancreaticDaraxonrasib reference arm

Overview

Activity alone no longer answers the question a KRAS program will be asked

Daraxonrasib was approved in August 2026 for previously treated metastatic pancreatic cancer, the first approval for a RAS(ON) inhibitor. KRAS-directed programs will now be asked how they compare, where they add benefit and whether they work where daraxonrasib has stopped working. Shrinkage against vehicle cannot answer those questions.

The atlas is a reference dataset built on that comparison. Panel models are profiled at baseline, under treatment and at endpoint, all under daraxonrasib. Your compound runs through the same profiling at the same timepoints, on the models you choose, so the readout is differentiation rather than activity.

The panel spans non-small cell lung, colorectal and pancreatic cancer, and includes models induced for resistance to daraxonrasib. Optional comparator arms, such as a second KRAS inhibitor or a chemotherapy standard, widen the frame where your development plan needs them.

kras-drug-development

KRAS-Mutant Atlas at a glance

Models
50+ KRAS-mutant
Indications
NSCLC, CRC, pancreatic
Reference arm
Daraxonrasib
Profiling
Baseline to endpoint
View study package

Study depths

Three depths of study, each building on the last

Start with a head-to-head response benchmark and add molecular depth where the result raises a question. The same analytics run at every depth, so results stay comparable as you go deeper.

In vivo response benchmark

Your compound and a parallel daraxonrasib arm run on the models you select. Tumor growth inhibition and regression, read by clinically correlated modified RECIST, show where your asset beats, matches or trails the reference, by indication and mutation.

A subset of models induced for daraxonrasib resistance shows whether your compound holds where the reference has already failed.

Supports

  • Differentiation go or no-go

  • Choosing which indication first

  • Positioning after daraxonrasib

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.

Deliverables

What a benchmark study puts in your hands

Every study delivers the result, the evidence behind it and the methods used. It does not include clinical response prediction; findings are framed as hypotheses for your team to test.

  • Head-to-head response data

    Tumor growth inhibition and regression for your compound and daraxonrasib on identical models and timepoints.

  • Resistance-model performance results

    How your compound performs in models already resistant to daraxonrasib, reported model by model.

  • Matched baseline molecular profiles

    Exome, transcriptome and proteomic data for every model in your study, each linked to its response.

  • Candidate response markers

    Features that separate responders from non-responders, with the contrast and model subset behind each one.

  • Ranked combination hypotheses

    Mechanism-backed combinations ranked by the resistance biology they target, ready for in vivo testing.

  • Methods and source data

    Figures, tables and methods behind each conclusion, with the underlying data available on request.

kras-battleground

Why now

A resistance reference you would otherwise have to build yourself

Clinical resistance to daraxonrasib will take years to characterize in patient samples. The atlas induces and profiles that resistance in vivo now, in models whose molecular baseline is already known, so your team can study it before the clinical picture is complete.

Building an equivalent reference in-house would mean sourcing KRAS-mutant models, dosing the standard of care and profiling every layer at every timepoint. Here that work is already underway, and your compound joins it rather than repeating it.

FAQ

What KRAS teams ask before a benchmark

Questions that usually come up before a benchmark study against daraxonrasib is scoped.

Daraxonrasib response has been measured across 50+ panel models. Deeper layers, such as on-treatment and longitudinal profiling, vary by model and indication, so tell us the models you are interested in and we will confirm what each already carries.

Yes. Optional arms can include a second KRAS inhibitor or a chemotherapy standard, run on the same models at the same timepoints so the comparison stays like for like across every arm in the study.

Not without your explicit agreement. Ownership and use are set in the study agreement before work starts, and data generated under your compound does not reach other users of the atlas unless you agree that it should.

It is the longest of the three depths, because it is longitudinal by design. Models are dosed and sampled through treatment and into acquired resistance, so the schedule is set by the biology. We scope it once the models are chosen.

Related pages

Built on the same platform

The atlas draws on the matched dataset and on in vivo capabilities covered in more detail on these pages.
kras

Plan a study

Send the compound class and indications

Tell us the compound class or mechanism, the indications and mutations you are prioritizing, and which depth of study you have in mind. If the compound is confidential, describe the class and we will put a CDA in place before details are shared.

  • A scientist replies to every request personally

  • Available panel models noted in the reply

  • CDA handled before any compound detail is shared