Documentation
Epistra Accelerate
AI-Powered Experimental Condition Optimization Software for Life Sciences
Technical Overview
Epistra Accelerate is an experimental condition optimization software based on Bayesian optimization. It uses a Gaussian Process (GP) as its surrogate model to predict experimental outcomes and quantify uncertainty, enabling the discovery of optimal conditions with significantly fewer trials than conventional methods in high-dimensional and high-cost parameter spaces typical of life science experiments. The optimization algorithms have been designed on benchmark problems that recapitulate real-world life science experimental tasks such as cell culture media optimization.
Key Features
- Proprietary optimization algorithms specialized for life sciences
- Multi-objective optimization across quality, cost, and productivity
- Visualization and interpretation tools for experimental results
Input / Output
Input
- Historical experimental data (table of conditions and results)
- Optimization settings (objective variables, search ranges, etc.)
Output
- Table of recommended experimental conditions for the next round
Performance Benchmarks
Validated results from industry collaborations:
| Application | Metric | Result |
|---|---|---|
| Novel assay system construction | R&D timeline | 44% reduction |
| Cell preservation reagent development | Reagent performance | 133% improvement |
| Biopharmaceutical media development | Raw material cost | 70% reduction |
Version History
| Version | Summary |
|---|---|
| 2.0.7 | Enhanced constraint handling, design space estimation, and cost estimation features |
| 2.0.0 | Major refactoring and solver performance update |
| 1.5.1 | Initial release (shipped with Shimadzu CellTune v1) |
Citation
If you use Epistra Accelerate in your research, please cite as follows:
@misc{epistra_accelerate,
title = {Epistra Accelerate},
author = {Epistra Inc.},
year = {2024},
url = {https://epistra.jp/epistra-accelerate/docs},
note = {version 2.0.7, Accessed: YYYY-MM-DD}
}Related Publications
- Kanda GN, Tsuzuki T, et al. (2022) "Robotic search for optimal cell culture in regenerative medicine." eLife 11:e77007
- Kawabata T, Tsuzuki T, Tatsukawa T, Matsui K, Kawakami E. (2025) "Black-box optimization in immunology and beyond: A practical guide to algorithms and future directions." Allergology International