中文 English
Foundation Intelligence for Equine Biology

Every equine decision begins with biological truth.

We build the biological knowledge graph that the global horse industry runs on. From breeding to veterinary to conservation — decisions backed by data, not assumptions.

22.8GB Unified Database
1.92M+ SNP Frequency Records
1,662 Variant Records
100% Traceable
Scroll to explore

The equine industry operates on
fragmented biological knowledge.

Phenotype data trapped in PDFs. Genotype data scattered across labs. Breed knowledge passed down orally, never digitized. No unified representation of "what a horse is." Every decision starts from scratch.

0
Unified Database (GB)
4 DuckDBs + 772 Parquet files, 1,643 objects unified
0
SNP Frequency Records (10K)
16 Chinese breeds × 40 SNPs — globally exclusive data
0
Variant Records
Full variant catalog including 14 gold-standard pathogenic mutations
100%
Traceable
Every data point links back to source paper, sample ID, and genomic coordinate

The Problem

Every equine decision — breeding, veterinary, racing, conservation — starts from fragmented knowledge with no unified biological foundation.

Our Approach

EQAI builds structured data, biological embeddings (HorseVector), connected knowledge graphs, and equine-specific foundation models. Biological complexity becomes computable intelligence.

From raw observation to
decision-ready intelligence.

Every layer is traceable to the one below it. Nothing is invented. Everything is grounded. And everything is continuously updated.

07

Decision

Breeding choice · Treatment plan · Racing strategy · Conservation action

06

Prediction

Risk score · Mating recommendation · Variant classification

05

Foundation Intelligence

Reasoning over structured knowledge — the equine-specific intelligence layer

04

Knowledge

Curated, connected, validated — the living knowledge graph

03

Validation

Cross-referenced · Peer-reviewed · Literature-confirmed

02

Evidence

Genomes · Phenotypes · Variants · Published research

01

Raw Observation

Samples · Veterinary records · Field data · Research papers

The Key Principle: A prediction links to the foundation intelligence that produced it. Foundation intelligence links to the knowledge that informed it. Knowledge links to the evidence that supports it. Evidence links to the raw observation that generated it.

The biological intelligence
infrastructure layer.

Just as an operating system sits between hardware and applications — Foundation Intelligence sits between raw biological data and decision-making.

HorseVector

Biological embeddings that encode equine entities as computable vectors. The universal representation for genes, variants, phenotypes, and breeds.

Embeddings Representation
In Development

Knowledge Graph

Connected biological entities. Gene → Variant → Phenotype → Breed relationships, curated and cross-validated against published literature.

Integration Validation
In Development

Foundation Model

Equine-specific foundation models pre-trained on curated biological data. We aim to build the reference model for equine biology.

Reasoning Prediction
Training

Radical Traceability

Every output includes confidence level, validation status, and full evidence chain. Every data point links to source paper, sample ID, and genomic coordinate.

Traceability Trust
Active

Five Core Capabilities of Foundation Intelligence

R
Representation Unified encoding of biological entities → HorseVector embeddings
I
Integration Multi-source data fusion → Knowledge Graph construction pipeline
R
Reasoning Biological inference → Equine Foundation Model
T
Traceability Every output links back to evidence → Radical Traceability system
P
Prediction Decision-ready outputs → Decision Intelligence layer

From genome to decision.

Decision intelligence applications built on the foundation layer. Every output is traceable, validated, and designed to support — never replace — expert judgment.

🔬

JumpID™

Genetic identification & parentage

Breed-agnostic genetic identification using curated SNP panels. Parentage verification with statistical confidence, supporting breed registries and conservation programs.

  • Multi-breed SNP identification panels
  • Statistical parentage verification
  • Breed diversity monitoring
  • Conservation-ready reporting
⛓️

JumpChain™

Provenance-tracked lineage data

Immutable lineage records connecting genetic data to verified samples. Designed for stud books, breed registries, and research institutions requiring auditable provenance.

  • Immutable sample-to-result provenance
  • Stud book integration ready
  • Auditable evidence chains
  • FAIR-compliant data standards

JumpHealth™ reports include a mandatory disclaimer: "This report identifies genetic risk factors. It is not a diagnosis. Consult a licensed veterinarian for clinical decisions."

Built on verified evidence.

Every EQAI output is grounded in internationally recognized databases and standards. We adhere to FAIR, MIAME, and GA4GH principles.

NCBI Reference SNP Database

Online Mendelian Inheritance in Animals

National Center for Biotechnology Information

Equine Reference Genome

Findable, Accessible, Interoperable, Reusable

Global Alliance for Genomics and Health

Five-Layer Validation Framework

1

Internal

Coordinate consistency, allele frequency checks

2

Cross-Breed

Findings verified across independent populations

3

Literature

rsID confirmed in dbSNP, OMIA verified

4

Expert

Domain expert review of interpretations

5

Continuous

Ongoing re-validation as new data arrives

Open science. Closed hype.

Data, methods, and validation pipelines are open unless there is a compelling reason to restrict. Science advances through transparency.

Research

Cold Adaptation Study

57 candidate genes identified across 17 breeds. 14 Tier A targets validated through cross-breed analysis. A systematic investigation into the genetic basis of cold tolerance in equine populations.

✅ Validated 14 Tier A · 43 Tier B
Infrastructure

Knowledge Graph Architecture

A comprehensive equine biological knowledge graph connecting genes, variants, phenotypes, and breeds with full evidence provenance. Gene → Variant → Phenotype → Breed.

🔄 In Development 31,685 entities · 1.37M edges
Data

Breed Frequency Database

17 breeds genotyped, 9 in production. Allele frequencies for all tracked variants across global equine populations. All data publicly available via API.

✅ Production 17 breeds · Open data
Model

NCDS Gene Model (v3.4)

102-gene model structured across Core, Neutral, and Hybrid layers. A framework for understanding neuromuscular and metabolic gene networks in equine biology.

✅ Research Complete 102 genes · 3 layers

Navigate GenoMatrix™

Quick access to all platforms, documentation, services, and resources.

LIVE
🏠

Brand Homepage

GenoMatrix™ brand overview, core capabilities, product suite, applications & technology architecture

DEMO
📊

Dashboard v5

Gene search · Exercise classification · JumpHealth reports · Knowledge graph visualization · Full analytics

LIVE
🧬

API Documentation v2

16 standardized API endpoints · RESTful design · Full request/response examples · Online testing

LIVE
🔬

CRO Services

Professional genomics contract research — genetic disease screening / exercise genotyping / breed identification / whole-genome analysis

LIVE
🧪

Research

Research methodology / NCDS model / Data pipeline architecture / EquCab 3.0 reference genome / 982 tables / 22GB

LIVE
🛡️

Data Compliance

GDPR · PIPL · CCPA compliance across three jurisdictions · Data classification · Cross-border transfer · Security audits

LIVE
💡

Industry Solutions

Genetic disease screening / Performance prediction / Breed identification / Transaction notarization / Conservation / Breeding decisions

LIVE
🏢

About Us

GenoMatrix Technology (Shanghai) Co., Ltd. — Leading equine genomics technology provider with three global centers

LIVE
🤝

Partners

Academic institutions, industry associations, breeding farms, laboratory partners — building the equine genomics ecosystem

LIVE

FAQ

Genetic testing process, sampling methods, report interpretation, NCDS model principles, API integration — common questions answered

LIVE
📰

News & Updates

Industry news, company updates, academic publications, conference events — latest in equine genomics

Make equine biology
computable.

We are building the biological intelligence infrastructure for the global equine ecosystem. Join us in making every equine decision begin with biological truth.

Decode to Decide.