Prioritize drug candidates before Phase II
Your drug shrinks tumors in mice — but will it work in humans? DNAI's Sim-to-Real engine strips species-specific noise from preclinical data and simulates human-scale treatment outcomes in silico.
The Hybrid Engine
Two complementary model paths — data type determines routing
Data-type routing — human clinical data uses Path A; preclinical PDX data uses Path B via DSN.
Standard Alignment Silently Destroys Biological Signals
When you align mouse and human latent spaces naively, the adversarial discriminator forces proliferation-linked channels to collapse. The model reports no error — but the simulation is silently wrong.
"This is structural, not a tuning problem."
It's a fundamental failure mode of domain adaptation applied to dynamics pipelines. The adversarial discriminator cannot distinguish proliferation signals from species-specific signals — so it destroys both.
Prioritizing Drug Candidates Early
Your new compound shrinks tumors in mice. Simulate human-scale outcomes in silico to generate hypotheses for your Go/No-Go decision — before investing in Phase II.
Drug Calibration (DSN Pipeline)
Virtual Trial Simulation
How DNAI supports R&D
From preclinical translation to trial design — in silico
Sim-to-Real Translation
Simulate human-scale outcomes from mouse data. The DSN strips species-specific noise to reveal conserved tumor biology.
Virtual trials
Simulate thousands of patients to explore optimal enrollment criteria before recruiting a single patient.
Biomarker discovery
Discover which molecular subtypes show highest simulated sensitivity. Inform your companion diagnostic strategy early.
R&D applications
Trial enrichment
Simulate which patient subgroups show highest response, informing enrollment criteria and reducing required sample sizes.
Mechanism-Based Enrichment
PATHWAYUse DNAI's pathway-level analysis to find patients with the specific pathway dysregulation your drug targets. Generate mechanism-linked hypotheses — identify which molecular features associate with simulated response.
Combination screening
Simulate combination therapies to explore potentially synergistic drug pairs and optimal sequencing strategies in silico.
Resistance simulation
Model clonal evolution to explore potential resistance mechanisms and inform adaptive treatment protocol design.
Hybrid Control Arms
Generate physics-constrained synthetic patient trajectories for control arms. Modeled potential to reduce control-arm enrollment — enabling more patients to receive experimental treatments.
Dose-Response Optimization
IN SILICODNAI's differentiable engine enables gradient-based exploration of dosing schedules that balance simulated efficacy and safety constraints. Optimize dose-response curves before clinical testing.
Virtual Tumor Burden Endpoints
v3.1DNAI simulates longitudinal tumor volume trajectories, enabling estimation of tumor burden changes over time. Volume-to-response classification (CR/PR/SD/PD) serves as an approximation of clinical imaging endpoints.
Platform Capabilities
How each component supports drug development and research
| Feature | Pharma Value (Drug Dev) | Research Value |
|---|---|---|
| DSN (Sim-to-Real) | ESSENTIAL Translate mouse data to human-scale simulations | Background — ensures physics engine uses conserved biology |
| Imputation | ESSENTIAL Use partial preclinical data | Background — handles missing modalities |
| Neural ODE | VIRTUAL TRIAL Simulate patient cohorts for trial design | PROGNOSIS Simulate patient trajectories |
| Safety Layer | QC Flag unreliable PDX models | ABSTENTION Flag when model cannot reliably simulate |
See the Biopharma Platform
Explore the tools for virtual trials, cohort simulation, and mechanism discovery
Identify patients most likely to benefit — before enrollment
Simulation suggests up to 10× enrichment in responder prevalence for select cancer types, based on retrospective analysis of observational data.
Methodology & Limitations
- •Based on retrospective analysis of TCGA observational data (N=9,393)
- •CATE estimates from S-learner model with propensity weighting — not randomized trials
- •For hypothesis generation and trial design planning, not efficacy claims
- •All enrichment results require prospective validation in controlled trials
How Simulation-Based Enrichment Works
Enrichment Potential by Cancer Type
Retrospective simulation on TCGA data (N=9,393). Enrichment = HR improvement when selecting CATE-predicted top responders.
| Cancer | N | All-Comers HR | Enriched HR | Enrichment Potential |
|---|---|---|---|---|
| LGG | 516 | 0.83 | 0.42 | Strong |
| HNSC | 521 | 0.80 | 0.52 | Strong |
| BRCA | 1091 | 0.72 | 0.61 | Moderate |
| ESCA | 184 | 0.60 | 0.25 | Strong |
| KIRC | 534 | 0.83 | 0.76 | Moderate |
| SARC | 255 | 1.31 | 1.31 | Insufficient |
Known Limitations
Intended use
DNAI is intended solely as an in silico research tool for hypothesis generation in drug development. It is not validated for regulatory submissions, clinical decision-making, or patient selection. All simulations should be interpreted alongside standard preclinical and clinical evidence.
Partner with us
Discuss how DNAI simulation tools can support your drug development programs.
Contact Us