The Science Behind InSilicon™
A Clear View of the Approach
InSilicon™ is built on established data science and biomedical methods, with an emphasis on transparency, validation, and clear limits. This page outlines the methodologies, how model performance is evaluated, and how results are communicated for review.
Scientific Standards & Principles
Our models are built on peer-reviewed methods, validated datasets, and standards derived from government-funded research programs.
Transparent model assumptions and clearly documented limitations
Alignment with emerging regulatory guidance on in silico evidence
Reproducible workflows suitable for audit and external review
Ethical science practices that support the reduction, refinement, and replacement (3Rs) of animal testing
Planning Meets Execution
AI and ML approaches
Uses established modeling approaches selected for the data, the question, and interpretability requirements.
Data sources
Draws on preclinical datasets and relevant public sources, curated for consistency, traceability, and appropriate use.
Historical benchmarking
Benchmarks model outputs against historical in vivo outcomes using predefined endpoints and comparison rules.
Performance metrics
Tracks standard performance metrics such as sensitivity and specificity, reported with applicability notes and stated uncertainty.
