The biology behind LNP delivery optimization
Gendelivr's platform is grounded in the physics of lipid self-assembly, the mechanics of endosomal escape, and the cell biology of hepatic gene editing. Here's what's under the hood.
Three foundational topics
LNP Biology
How ionizable lipids assemble into nanoparticles, how pKa governs endosomal escape, and how particle size affects biodistribution in vivo.
Read articleCRISPR-Cas9 Delivery
From guide RNA design to Cas9 ribonucleoprotein delivery: why LNP encapsulation of mRNA-based CRISPR components outperforms viral vectors for in-vivo hepatic editing.
Read articleHepatic Targeting
Why the liver is the dominant site of LNP uptake, the role of ApoE-mediated endocytosis, and how we tune formulations for hepatocyte-selective transfection.
Read articleFrom first principles to bench protocol
Gendelivr's Formulation Engine is grounded in coarse-grained molecular dynamics parameterized against published lipid force fields (MARTINI-based). We don't use black-box neural networks trained on proprietary data — we use physics that generalizes.
This means our predictions degrade gracefully outside the training distribution — a critical property for novel ionizable lipid classes your chemistry team may be synthesizing for the first time.
We do not claim to predict in vivo editing efficiency — that depends on guide RNA activity, chromatin accessibility, and Cas9 protein folding, which fall outside the LNP formulation problem. What we predict is formulation-level performance: pKa, encapsulation efficiency, particle size, and the probability of hepatic uptake and endosomal escape. Those predictions are where we have mechanistic grounding and where computational screening creates real leverage.
Explore the PlatformCGMD Simulation
Coarse-grained molecular dynamics at physiological and endosomal pH to predict protonation states and membrane perturbation.
Thermodynamic pKa Model
QSPR-based apparent pKa prediction for ionizable lipid headgroups across a library of 1,200+ structures.
Pharmacokinetic Pre-filter
Particle size and surface charge predictions feed into a hepatic biodistribution pre-filter to eliminate formulations unlikely to reach hepatocytes.
Questions about our modeling approach?
We publish our modeling methods openly and welcome scientific discussion. Reach our research team directly.