Rejuve

help edit space_dashboard
Draft — This content has not been approved for publication.

Entities: Rejuve Biotech (rejuve.bio) and Rejuve.AI (rejuve.ai)

Founded by: Jasmine Smith and Dr. Ben Goertzel

Ecosystem relationships: SingularityNET (parent ecosystem, spun off 2019), TrueAGI (AGI framework collaboration), Genescient Inc. (Methuselah fly omics data)

Status: Active. BioAtom Space knowledge graph operational with 330M+ atoms from FlyBase. Neural-symbolic AI platform in use for longevity hypothesis generation. Rejuve.AI app incorporates BayesExpert model and Transformer VAE. Biomarker study with 100+ longevity athletes conducted at Infinite Games 2026.

Rejuve is a SingularityNET ecosystem project applying neural-symbolic AI to longevity research and drug discovery. The initiative spans two entities: Rejuve Biotech (translational medicine research platform) and Rejuve.AI (decentralized longevity data network and consumer app). Both leverage Hyperon's knowledge representation and reasoning capabilities to accelerate the discovery of aging interventions.

BioAtom Space

The core technical contribution is the BioAtom Space — a large-scale knowledge graph connecting genes, proteins, diseases, phenotypes, mutations, and biological pathways into a unified, searchable map. FlyBase (the Drosophila genomics database) has been imported into OpenCog Hyperon's AtomSpace, creating a graph of 330 million+ atoms. This enables AI algorithms to perform graph traversal, pattern matching, and hypothesis generation directly over biological data.

A key dataset is omics data from the Methuselah flies — Drosophila experimentally evolved to live approximately 5× as long as control populations — accessed through a partnership with Genescient Inc. Cross-organism omics analysis identifies conserved longevity pathways that may translate to human aging interventions.

Neural-Symbolic AI Platform

Rejuve Bio's research platform combines deep neural networks (including LLMs and a Modular Transformer Variational Autoencoder) with symbolic and evolutionary AI implemented in Hyperon. The neural-symbolic approach produces candidate links between biological pathways and potential treatments, while AI evaluates interventions against known hallmarks of aging to prioritize the most promising candidates. The platform generates testable hypotheses that are validated with living systems.

The custom-atomspace-builder (maintained under the rejuve-bio GitHub organization) is a FastAPI service for transforming tabular biomedical data (CSV, JSON, Excel, XML) into knowledge graph formats compatible with AtomSpace, Neo4j, HugeGraph, and MORK. The Distributed AtomSpace (DAS) includes a Rejuve proof-of-concept demonstrating distributed knowledge graph queries over biological data.

Rejuve.AI Network

The consumer-facing Rejuve.AI app collects and analyzes human longevity data through a decentralized network. It includes a hand-crafted BayesExpert model and a Modular Transformer VAE that uses self-supervision to model human development across the lifespan. In early 2026, Rejuve.AI conducted a biomarker study enrolling 100+ longevity athletes at the Infinite Games event to measure how different therapies, training regimens, and lifestyles impact biological aging.

Implementation Anchors

  • custom-atomspace-builder — FastAPI service for tabular data → knowledge graph conversion
  • das — Distributed AtomSpace with Rejuve proof-of-concept

Key References

Related cards: Bioinformatics · DAS · AtomSpace



Discussion