AI Biotech Picks and Shovels
Conor G馃憖dwin@goodyy22
+0.2%
since Oct 15
The idea
I believe AI will reshape drug discovery, but the surest way to play it is owning the picks and shovels rather than guessing which drugmaker or model wins. So this basket spreads across three layers: the labs generating new biological data, the equipment and supplies every experiment depends on, and the firms running clinical trials. Steady large suppliers carry the most weight, smaller data-focused names get lighter positions, and a small cash buffer sits on the side since this is a multi-year thesis, not a near-term trade.
What it owns
- Every experiment, AI-guided or not, runs on their instruments and consumables, so more research means more sales. 路 Thermo Fisher ScientificIt supplies the lab equipment, chemicals and services that nearly every experiment depends on, so it gets paid across the full spectrum of research and drug manufacturing., DanaherIt sells lab instruments and bioprocessing equipment, earning revenue both when drugs are discovered and when they are manufactured, plus its partnership with NVIDIA brings AI computing directly into labs., Agilent TechnologiesIt makes the analytical instruments labs use to test samples during drug research and is expanding into AI-driven pathology, bringing it closer to the intersection of AI and biotech., Waters CorporationIt leads in mass spectrometry and liquid chromatography, the machines that test drugs from early research through manufacturing, and its new AI Link product aligns with lab automation trends., Mettler-ToledoIt makes the precision scales and lab basics present on nearly every lab bench, benefiting from increased lab activity without depending on any single technology., Bio-TechneIt sells proteins, antibodies and reagents that researchers consume in experiments, so more lab work means more repeat orders.35%
- Clinical trials are the slow step AI can't skip, so more drug candidates means more trials for these companies to run. 路 IQVIA HoldingsIt runs clinical trials and owns one of the largest healthcare datasets, so every AI-found drug that reaches human testing becomes a potential customer., Medpace HoldingsIt runs clinical trials primarily for smaller biotech companies, where many AI-discovered drugs are likely to start their human testing., Veeva SystemsIt provides the software drugmakers use to run trials and manage regulatory work, so more drugs in development means more software seats., ICON plcIt runs large global clinical trials for major pharma and biotech companies, serving as a toll booth every new drug candidate must pass through., Charles River LaboratoriesIt performs the preclinical safety and animal testing that drugs require before reaching human trials, so AI-found drug candidates still flow through its labs.31%
- AI models are only as good as the biological data they learn from, and these companies make the machines that generate it. 路 IlluminaIt makes most of the world's DNA sequencers, so if AI drug discovery requires vast amounts of new genetic data, much of it gets generated on its machines., QiagenIt makes the kits that extract DNA and RNA from samples before sequencing, so it gets paid on nearly every genetic test regardless of which sequencer is used., BrukerIt sells instruments that measure proteins and map molecular locations in tissue, providing the next layer of biological detail after DNA sequencing., RevvityIt sells drug screening systems, reagents and software that labs use to generate the experimental data AI models learn from., Twist BioscienceIt prints custom DNA, which is how AI-designed genes and proteins actually get built and tested in laboratories., 10x GenomicsIt makes the tools that read individual cells one at a time, generating the detailed biological data that AI models need for drug discovery.25%
- Cash9%