Conventional drug testing treats the brain as a uniform system, ignoring its distinct cellular neighborhoods, spatial architectures, and localized disease signatures. DNAxon integrates single-cell and spatial transcriptomics to build the world's first spatial AI for neuropharmacology. By simulating how drugs interact with human brain tissue at the single-cell level, we forecast clinical efficacy, map off-target toxicity, and pinpoint responder subgroups years before a trial begins.
Developing CNS therapies is one of the most complex challenges in modern medicine, historically limited by the inability to track drug interactions within the brain's physical architecture. Our mission is to equip researchers with the missing layer of spatial intelligence. By illuminating the exact cellular coordinates of disease, we empower R&D teams to validate targets, predict safety, and advance clinical pipelines with unprecedented visibility and confidence.
What traditionally takes over a decade of trial and error can now be mapped and de-risked in a fraction of the time.
A foundation model trained on human brain genomic data from CELLxGENE, the Human Cell Atlas and ROSMAP — learning the shared and divergent genetic signatures of neurological disease across cell types, cohorts and disease states simultaneously.
Maps gene expression across brain regions and cellular neighbourhoods by integrating single-cell RNA-seq with spatial transcriptomics data from Atera, Visium, Xenium, GeoMX, CosMx and MERFISH — capturing how drug targets behave within the spatial architecture of intact human neural tissue. As volumetric spatial methods mature, DNAxon is designed to incorporate 3D resolution.
A drug prediction layer built exclusively for CNS assets — simulating drug interactions at the single-cell level to forecast blood-brain barrier penetrance, cell-type specific binding and off-target neurotoxicity across genetically distinct patient subgroups, translating spatial biology into drug development decisions before clinical trials begin.
Most platforms in this space address one or two of these dimensions. Few address the brain specifically. Fewer still combine spatial resolution with drug response prediction. The table below reflects the current landscape as understood from published sources and public company information.
| Company / Model | Brain-Specific | Spatial | Drug Discovery | Status | Data Type |
|---|---|---|---|---|---|
| Recursion | ✗ Multi-disease | ✗ No spatial | ✓ | Nasdaq: RXRX | Cell imaging |
| Isomorphic Labs | ✗ All targets | ✗ No spatial | ✓ | $2.1B Series B | Molecular |
| Aitia | Partial (neuro + oncology) | ✗ No spatial | ✓ | $65M raised | Multi-omic |
| BrainStorm Therapeutics | ✓ Brain | ✗ No spatial | ✓ | Early stage | scRNA + organoid |
| scGPT-spatial | ✗ All tissues | ✓ | ✗ | Academic preprint | Limited |
| Nicheformer | ✗ 73 organs | ✓ | ✗ | Nature Methods 2025 | Multi-tissue |
| BrainBeacon | ✓ Brain | ✓ | ✗ | bioRxiv preprint 2025 | Cross-species |
| CellType | ✗ All biology | ✗ No spatial | ✓ | YC W26 · Founded 2025 | scRNA only |
| DNAxon ★ | ✓ Brain-dedicated | ✓ Spatial scRNA | ✓ Drug response | Concept stage | ✓ Public datasets |
UK Biobank · Allen Brain Atlas · GTEx Brain · ADNI · ROSMAP · PsychENCODE · AMP-AD · PPMI · Human Cell Atlas · NDKP Portal
DNAxon is at the concept and pre-incorporation stage. We are seeking scientific collaborators, advisors and early partners who want to work on one of the hardest unsolved problems in medicine — predicting how drugs behave in the human brain at single-cell spatial resolution before a single patient is dosed.
Neuroscientists, bioinformaticians, clinicians, computational biologists, computer scientists, biostatisticians and mathematicians who want to contribute to building a spatially-resolved neurogenomics AI platform at the earliest stage.
Pharma R&D veterans, academic directors, and industry leaders who can guide scientific strategy and open pharma doors. Advisory equity available.
Research coordinators, patient advocates, science communicators and academic partners who want to support the translation of spatial neurogenomics research into a clinical reality.
Angels, impact investors and early-stage biotech funds who believe the next advance in neurodegeneration will come from spatially-resolved AI genomics — and want to be involved at the earliest stage.
The most important conversations happen at the earliest stage.
✉ [email protected]