InSilicoTrials Platform

InSilicoTrials is an end-to-end in silico platform designed to democratize computational simulation across the full drug development lifecycle — from preclinical research through clinical trials to post-approval decisions. By integrating mechanistic models, artificial intelligence, digital twins, synthetic data generation, and operations forecasting into a single unified environment, the platform enables pharmaceutical and biotech organizations to make earlier, safer, and more cost-effective development decisions. InSilicoTrials operates at the intersection of scientific innovation and regulatory compliance, with contributions to emerging good practice frameworks including Toward Good Simulation Practice, co-authored with the U.S. FDA.

The platform is built for drug developers, clinical teams, and research organizations seeking to reduce reliance on in vivo studies and early human exposure, consistent with FDA initiatives encouraging alternatives to traditional animal testing. It supports a continuous optimisation loop across preclinical mAb programmes, biosimilarity confirmation, oncology ADC development, rare disease trials, multiple sclerosis strategy, and in-licensing business development decisions — with documented time savings ranging from 6 months to 5 years and cost savings of $5M to $150M depending on the use case.

Scientific Platform Capabilities

• Enables full clinical trial simulation, including synthetic patient generation and scenario comparisons across dosing, sample size, inclusion/exclusion criteria, adaptive designs, and statistical methods
• Includes a Model Library, Data Library, Data Integrator, Workflow Builder, Clinical Trial Simulator, and Output Dashboard
• Supports a broad range of model types: safety, efficacy, biomarkers, disease progression, quantitative systems pharmacology (QSP), multi-omics, in vitro/in vivo correlation, and AI/ML models
• Provides a unified data integration framework powering multi-modal data at global scale

Operations Platform Capabilities

• Enables clinical trial simulation with the same core toolset as the Scientific Platform, including synthetic patient generation and scenario comparisons for dosing, sample size, I/E criteria, adaptive designs, and statistical methods
• Includes Model Library, Data Library, Data Integrator, Workflow Builder, Clinical Trial Simulator, and Output Dashboard
• Supports safety, efficacy, biomarker, disease progression, QSP, multi-omics, in vitro/in vivo, and AI/ML models for operational forecasting and planning

IRIS: Multi-Agentic Simulation Workflow Orchestration

• AI-powered orchestration layer built on a multi-agent architecture that connects internal tools, internal data, external information sources, and the simulation platform
• Searches external sources such as clinicaltrials.gov to suggest and inform clinical trial designs
• Interprets data, generates recommendations, and analyses simulation results autonomously
• Independently launches new simulations based on retrieved information and user-defined preferences

Flagship Use Cases and Demonstrated Outcomes

• In Silico Antibody Optimisation (Preclinical mAb): Using AI-assisted TMDD modelling, clearance and affinity scenario simulation, and virtual patient modelling, the platform identified an optimised antibody variant with improved affinity, reduced clearance, and stronger target suppression — saving 12–30 months and $15–30M while reducing reliance on in vivo studies
• ALS Synthetic Control Arm (Phase II, Rare Disease): ML-based disease-progression modelling and causal inference were used to generate synthetic control patients for an ALS trial with ultra-small patient populations. The control arm was augmented with 60 synthetic patients, increasing statistical power and reducing placebo burden — saving 1–2 years and $10–40M
• Oncology Phase III ADC Digital Twin: Multi-scale QSP digital twins linking PK, tumour dynamics, and platelet counts were used to compare fixed versus adaptive dosing strategies for an antibody-drug conjugate. The optimal weekly regimen was identified with superior efficacy and acceptable safety trade-offs — saving 6–18 months and up to $50M
• MS 5-Year QSP Digital Twin (Label Extension): QSP modelling of immune response and lymphocyte dynamics, combined with long-horizon simulation of relapse rates across virtual MS populations, provided evidence to support label extension without new multi-year clinical trials — saving 3–5 years and $50–150M

Awards and Recognition

• ARPA-H CATALYST — CARDIOVERSE digital-twin programme participant
• FDA GenAI Precision Challenge Top 5 Winner
• EU Virtual Human Twin initiative participant
• London AI Summit Winner
• Grind AI Startup of the Year Winner

InSilicoTrials maintains alignment with FDA and EMA regulatory standards and has co-contributed to the development of emerging good simulation practice frameworks. The platform is supported by a broad global network of scientific and industry partners, including collaborations with Premier Research, Axoltis Pharma, and the Research Center Pharmaceutical Engineering (RCPE), and is positioned as a single-source repository for pharmaceutical development from discovery through to dosage form.