E-PharmNova combines open-weight structural and generative models, GPU-scale screening, and a Phase I clinical research unit to turn targets of aging into clinical candidates in months, not years.
Explore the platformPartner with usNinety percent of drug programs fail in the clinic. We start from molecules with known human safety, let AI find the ones that act on aging biology, and test them where we already run trials.
Large pharma is spending billions on AI laboratories. We take a different route: open models with commercial licences, cloud and on-premise NVIDIA GPUs, outsourced wet labs, and a rare asset that most AI-first biotechs lack — our own early-phase clinical infrastructure and, with informed consent, the longitudinal clinical and omics data it generates.
Aging itself is not an indication. Every program is framed around a concrete condition or a validated biomarker endpoint in a defined population, under ethics-committee and regulator approval.
Four GPU-backed stages, orchestrated end to end in containers so every hit can be traced back to its data and model versions.
LLM agents mine literature and public omics (Open Targets, GTEx, CELLxGENE, GenAge, DrugAge) to rank targets and senescence signatures.
Boltz-2 and OpenFold3 structures, GNINA / Vina / DiffDock docking, and CMap–LINCS signature reversal across approved and Phase I-tested libraries.
ADMET-AI and RDKit filters, physics-based refinement (OpenMM, OpenFE) and medicinal-chemistry review reduce hundreds of hits to a shortlist.
Senescence assays and C. elegans lifespan studies at contract labs, then biomarker-driven pilots in our partner Phase I unit.
| Program | Target ID | In silico screen | In vitro | In vivo | Clinical pilot |
|---|---|---|---|---|---|
| NOVA-001 · Senolytic repurposingApproved & Phase I-tested molecules vs. senescence signatures | |||||
| NOVA-002 · Senomorphic (SASP) modulatorsRepurposing against SASP signalling in stromal cells | |||||
| Target discovery · aging signaturesOngoing signature and target ranking feeding future programs | |||||
Daily inference — structure prediction, docking, generative chemistry — runs on an on-premise NVIDIA workstation. Peak screening campaigns burst to rented H100 / B200 capacity. Workflows run in Docker under Nextflow, with results and audit trails in an ELN so the data is IND-ready from day one.
Patient-derived data is processed in a segregated, de-identified environment compliant with Panama's Ley 81 de 2019 and, for US data, HIPAA.
Boltz-2OpenFold3GNINAAutoDock VinaDiffDockREINVENT 4RFdiffusionProteinMPNNOpenMMOpenFEADMET-AIRDKitCUDADockerNextflowPythonPhysician-entrepreneur. Founder of clinical research organizations and a network of medical centers, and of a Phase I clinical research unit in Panama that E-PharmNova uses for early clinical validation.
Leads target selection, clinical strategy and partnerships.
Builds and runs the Discovery AI pipeline: data ingestion, GPU screening workflows, containerized and orchestrated with Docker and Nextflow.
Technical contact for platform, infrastructure and compute.
Trial design, statistical analysis plans and data management for the clinical pilots that follow computational and wet-lab validation.
Compound selection, freedom-to-operate review and regulatory dossiers for candidates moving toward the clinic.
Remote-first. Latin America and Europe time zones preferred. Write to us with a short note and links to your work.
We co-develop repurposing programs, run early-phase studies for partners, and license validated hits.