Biomimetic membrane
The bilayer is built from the same kinds of components human membranes already use, aiming for high compatibility and low immunogenicity. Its surface is left open for the navigation and fusion parts.
Shenzhen · AI-native biotech
Getting genetic medicines
into the cytosol
SILLOGIC · PNDDS, a biomimetic vesicle delivery platform
A nucleic-acid or gene medicine only works if it gets inside the cell. We use AI to build one biomimetic vesicle out of mechanisms that already work on their own. By design it fuses with the cell membrane, puts its cargo in the cytosol, and reaches an organ without depending on liver uptake.
- mRNA and siRNA
- Large genes
- Intracellular proteins
Why we bet the company on delivery
The past decade proved that mRNA and siRNA can be drugs. What is left is getting them to the cell type that matters and then across into the cytosol, and from where we sit that is where a lot of programs stop.
01
Molecule
Sequence design and chemical modification for siRNA and mRNA are in decent shape by now. When a program stops today, the molecule is rarely the reason.
02
Mechanism
Every mechanism we use has been validated by someone else in published work. No one had put them in the same carrier. Doing that assembly is what the AI is for.
03
Carrier
How far a program can go depends on the carrier: which organ it reaches, and whether it gets inside the cell.
Where today’s carriers stop
-
L-01
Trapped in the endosomeEndosomal entrapment
Once a carrier is taken up, most of it is held in endosomes and degraded. In the classic measurement, only a single-digit percentage reached the cytosol (Gilleron et al., Nat Biotechnol, 2013), and that order of magnitude has not changed since.Gilleron et al., Nat Biotechnol, 2013
-
L-02
Beyond the liverExtra-hepatic targeting
Mainstream lipid carriers accumulate in the liver on their own, which suits liver disease. For the brain, the pancreas and the lung, systemic dosing still has no route that reproduces reliably.
-
L-03
Repeat dosingImmune tolerance
The usual fix is a polymer coat on the surface, which provokes immune responses and speeds up clearance. Chronic and neurological indications are dosed for years at a stretch, so that chemistry is out.
-
L-04
The payload ceilingCargo capacity
Viral vectors have a hard capacity limit that rules out a whole class of large-gene replacement therapies. For some inherited diseases the causal gene has been known for years and there is still no treatment.
PNDDS · Precision Navigation Drug Delivery System
PNDDS is a biomimetic nanovesicle. Our AI system does the designing. The capabilities below sit on one membrane, and active membrane fusion is what decides whether the drug reaches the cytosol.
Active navigation
Recognition elements on the surface find the target cell through molecular interaction. The organ and the cell type a carrier goes to are settled at the design stage.
Active membrane fusion
The carrier fuses with the cell membrane and releases its cargo into the cytosol rather than going through the endosome. Large genes and intracellular proteins that will not fit a viral vector fit here.
Immune compatibility
We leave out the polymer shell in common use, and with it the antibody response and the faster clearance it brings. The design is for repeat dosing; the long-term immune data is not in yet.
An AI system runs the discovery loop
How a carrier gets designed is decided by this AI system. A round goes like this: it proposes designs from the indication and the target cell, most are cut before anything is synthesized, the rest are made and measured at the bench, and what comes back sets up the next round. Nothing counts until it has been through the wet lab.
The more programs we run, the sharper the next round of predictions gets.
The areas we work on first
An area has to have a real clinical gap, and delivery has to be what is holding those programs up.
F-01
Extra-hepatic solid organsSolid organ delivery
If delivery extends from the liver into solid organs such as the pancreas and the lung, molecules that already exist gain new indications.
F-02
Central nervous systemAcross the blood-brain barrier
There are a great many neurodegeneration programs in development, and most of them cannot get across the blood-brain barrier.
F-03
Inherited retinal diseaseLocal administration
The eye can be dosed locally and the dose is easy to control. Even so, most of these patients still have no approved treatment.
F-04
Large-gene replacementBeyond viral capacity
For some diseases the causal gene is well characterized but too long for a viral vector. Those need a route that does not use one.
We are the platform. We don’t compete with partners for indications.
Technology licensing
Delivery technology licensed by target or by indication. Development after that is yours to run.
Co-development
We share the funding and the risk on one program, and the proceeds on agreed terms. You bring the molecule; we bring the carrier and process.
Feasibility study
Send us the molecule. We run one round of delivery testing in the cell type or organ you name and hand back a data package you can take into your own review.
About SILLOGIC
SILLOGIC is an AI-native biotechnology company based in Shenzhen. We work on delivery, and nothing else.
AI and structural biology only became genuinely combinable after 2024, and that is the point the company started from. Whatever the models design, we build and measure ourselves. Nothing counts until it has been.
- Mechanisms with papers behind them
- The only data we count comes off a bench
- We take on one extra-hepatic organ at a time
If your molecule is stuck at delivery,
write and tell us which cell it needs to get into.
- br.liu@sillogic.ai
- Location
- Shenzhen, China