Formulation Engine
Screens 10,000 ionizable lipid and excipient combinations in silico. Our physics-informed model predicts pKa, encapsulation efficiency, and particle size before any bench synthesis.
Explore moduleGendelivr's computational screen ranks lipid nanoparticle candidates by hepatocyte transfection efficiency — compressing 18-month delivery optimization to 5.
18 months of bench optimization, condensed to 5. Here's how.
Each module addresses a specific bottleneck in the LNP development pipeline — from initial design space exploration to final IND documentation.
Screens 10,000 ionizable lipid and excipient combinations in silico. Our physics-informed model predicts pKa, encapsulation efficiency, and particle size before any bench synthesis.
Explore moduleRanks candidates by predicted hepatocyte transfection efficiency vs. cytotoxicity. Top-ranked formulations are exported as prioritized bench-synthesis queues — typically 20–30 candidates from 10,000 screened.
Explore moduleStructures validated formulation data into IND Module 3 format. Generates CMC sections, analytical characterization summaries, and stability assessment documentation.
Explore moduleOur ranking engine outputs a scored table of candidate formulations, each annotated with predicted transfection efficiency, cytotoxicity index, and particle stability score. Amber rows indicate top-tier candidates flagged for bench validation.
All data is export-ready: CSV for informatics pipelines, PDF for CMC documentation, JSON for ELN integration.
| Rank | Formulation ID | Ionizable Lipid | Transfection % | CTX Index | Score |
|---|---|---|---|---|---|
| 01 | GDV-F-0042 | IL-C8-linDMA | 87.3% | 0.08 | 9.82 |
| 02 | GDV-F-0117 | IL-C12-oxDMA | 84.1% | 0.11 | 9.56 |
| 03 | GDV-F-0089 | MC3-analog-4 | 82.7% | 0.13 | 9.34 |
| 04 | GDV-F-0201 | IL-C10-esterDMA | 78.4% | 0.17 | 8.91 |
| 05 | GDV-F-0334 | piperazine-C12 | 76.9% | 0.19 | 8.73 |
| ... | 9,995 additional formulations ranked | ||||
We were staring at a 16-month bench screen before IND. After running Gendelivr's computational campaign against our target tissue profile, we narrowed to 18 candidates in four days. The pKa and encapsulation predictions were consistently within measurement error of our bench results.
Senior Scientist, Gene Therapy R&D — CRISPR therapeutics company
The hepatic targeting model is notably well-calibrated. The ApoE corona prediction correctly ranked our MC3-analog formulations in order of in vivo liver uptake. That kind of mechanistic accuracy is what separates this from generic ML screening tools.
Head of Delivery Platform — gene therapy program within a pharma
Request access to the Gendelivr platform and upload your first target tissue profile. Results within 72 hours.
Boston, MA · [email protected] · +1 (617) 555-0962