Researchers at the Wyss Institute, Boston Children’s Hospital and Harvard Medical School described FIND — Functional Interaction–based Nanoswitch Discovery — a DNA-nanoswitch method for finding molecules that change protein–protein interactions, in a Wyss news post and a Science paper published 5 November 2026 (Hansen, Lipstein et al., DOI 10.1126/science.adt7185). The work is a laboratory screening platform tested in vitro and in cellular assays. It is not a clinical trial and not an approved medicine.
FIND builds each experiment on a single DNA molecule so many tests can run in one tube, Wyss faculty member Wesley Wong said: “Our idea was to build each experiment onto a single DNA molecule, so that many experiments could run in parallel in one tube.” Co-author Ryan Hansen said the method can “directly search for molecules that have the effect we want on a disease-related protein-protein interaction.”
What FIND does in the lab
DNA nanoswitches link a target protein, a partner protein and a barcoded library member. Looped versus linear shapes are separated by gel electrophoresis, then barcodes are sequenced to identify which library members changed the interaction. Demonstrations in the paper and Wyss summary include a nanobody that stabilised the ACE2–SARS-CoV-2 spike trimer interaction with activity in a cellular infection assay; PPAR-gamma agonists drawn from a DNA-encoded library of 400 million compounds that increased binding to a coactivator peptide and increased lipid accumulation in a cellular assay; and recovery of a known beta-2 adrenergic receptor activator as a validation control.
The technology is licensed to Merge Bio Therapeutics, whose cofounders include Wong, Hansen and Lipstein. Competing interests noted with the Science paper include patents and that company link.
Gel electrophoresis separation of looped versus linear nanoswitches is a classical molecular-biology readout married to modern DNA-encoded libraries. That combination is why the authors can claim both physical interaction-state information and an identity barcode from the same construct. Validation by recovering a known beta-2 agonist is the negative-control style check that screening papers use to show the assay is not only finding noise.
How this sits with other AI and screening drug-discovery posts
TSN has covered other discovery-platform stories including Abaucin antibiotic work, UCLA’s MAESTRO drug-combination screen, and PARP-inhibitor research. Those pieces also sit upstream of the clinic. FIND’s distinctive claim is physical DNA nanoswitches that encode interaction state and library identity in one molecular construct, then read hits by shape and sequence rather than by a purely computational proposal.
What a one-tube PPI nanoswitch screen changes for early discovery
On the published experiments, FIND is a way to hunt functional modulators of protein–protein interactions inside very large DNA-encoded libraries without claiming a patient-ready drug. The ACE2–spike nanobody and PPAR-gamma cellular readouts show biology past a pure binding assay, still short of animal or human efficacy. Merge Bio’s licence is a commercialisation path for the platform, not evidence that any FIND hit is in clinical development.
A 400-million-member DNA-encoded library is only useful if the detection method can keep functional signal attached to a recoverable barcode; FIND’s loop-versus-linear separation is the authors’ answer to that bookkeeping problem. Parallelism in one tube is the throughput argument Wong emphasises, while Hansen stresses searching for the desired effect on a disease-related interaction rather than for binders alone.
Hits still need medicinal chemistry, selectivity work and later trials before treatment claims. The Science package’s value is the screening architecture and the three demonstration classes — viral interaction, nuclear-receptor agonist library and a known GPCR activator recovered as a control — attributed to the authors’ assays. Competing interests around patents and Merge Bio are disclosed with the paper and belong in any fair reading of the commercial path.
Sources
- Finding new drug leads faster with DNA nanotechnology (Wyss Institute, 5 November 2026)
- Hansen, Lipstein et al., Discovering molecular modulators with nanoswitch screening, Science (5 November 2026)
Related stories
- AI Screened 6,680 Molecules and Helped Find a New Antibiotic, Abaucin
- UCLA’s MAESTRO AI Ranked 3,592 Compounds to Pair With a Cancer Drug
- NYU Links ESCO2 Loss to PARP Inhibitor Sensitivity
Follow @tsnmediahub on X for breaking AI news as it happens.
Follow on X
