AI-Designed Medicine
Trading Aloha@tradingaloha
−10.1%
since Sep 30
The idea
I believe AI is going to move from hype to real approved medicines over the next several years, so I built this around companies actually doing that work, deliberately skipping big pharma. The core is data and compute backbone plus lab tools that get paid regardless of which drug wins, alongside pure-play AI discovery platforms. I added a small sleeve of pre-revenue moonshots, sized at just a few percent each, since a single buyout in this space can move the whole take. I'm holding cash on the side and thinking in years, not months.
What it owns
- Genomic data, patient databases, and computing power that AI drug design models train on, sold by suppliers regardless of which discovery programs succeed. · NVIDIAIt sells the computing power that trains AI models to design drugs, the deepest infrastructure layer in the take., IlluminaIts sequencers generate the genomic data that AI models learn from, earning revenue today regardless of which discovery programs ultimately succeed., IQVIA HoldingsIt holds patient and trial data at the scale AI drug programs need to prove their molecules in the clinic, earning fees as more AI-designed drugs enter trials., 10x GenomicsIt sells tools that read biology at the single-cell level, the resolution AI models increasingly train on for drug discovery., Twist BioscienceIt manufactures synthetic DNA that turns AI-designed molecules into testable compounds, earning payment on every experiment as discovery programs validate their ideas.29%
- The instruments and bioprocessing equipment that every drug discovery operation runs through, providing steady revenue as AI-directed R&D spending flows through the lab. · Thermo Fisher ScientificIt supplies the instruments and consumables nearly every drug discovery lab uses, providing steady revenue as AI-directed R&D keeps flowing through its equipment., DanaherIt manufactures the bioprocessing tools needed to produce biologic drugs once AI designs them, positioned to grow as AI-discovered molecules advance toward manufacturing., Agilent TechnologiesIt sells the analytical instruments labs use to verify what AI predicted, receiving steady payment as lab demand for drug discovery persists.24%
- Companies selling AI-powered drug design software and services with established revenue streams, the closest to proven business models in the space. · SchrodingerIt sells physics-based simulation software the industry uses to design molecules, a pure-play on AI producing approved drugs., Tempus AIIt converts patient data into AI models for drug discovery and testing, a platform that earns as drugmakers shift from hype toward actual approvals., AbCellera BiologicsIt runs an AI antibody discovery engine that partners pay to use, earning revenue through deals and royalties as AI-found antibodies reach approval., CertaraIt simulates drugs and trials on computers before real testing begins, software that earns revenue today as computational modeling spreads through the industry.23%
- Early-stage AI-driven drug discovery pipelines held as potential acquisition targets and breakthrough opportunities. · Recursion PharmaceuticalsIt builds AI-driven drug pipelines from the ground up, a potential acquisition target as its computational candidates advance toward trials., Relay TherapeuticsIt models protein movement to find drugs competitors miss, a moonshot pipeline that remains an attractive acquisition target as it produces clinical candidates., AbsciIt uses generative AI to design new antibodies from scratch, a small pre-revenue pipeline where successful trials could attract premium acquisition offers., Structure TherapeuticsIt designs oral drugs computationally against targets that normally need injections, a pre-revenue moonshot with large acquisition potential.14%
- Cash10%