Behavioral Phenomics Database

Decode Behavioral Phenotypes, Map Brain Disease Atlases

Built on AI fine-grained behavior parsing, PhenomicsAtlas aggregates multi-species, multi-disease, and multimodal phenotypic data — an animal phenomics infrastructure for brain science and CNS drug discovery.

Enter PhenomicsAtlas
PhenomicsAtlas disease phenotype database interface
6,000+Animal samples
6Species
41+Disease models
1,600+Behavior hours

Continuously growing data assets
Last updated · 2026.07

Why Phenomics

From Behavior Observation to Disease Phenotype Decoding

Animal behavior is the integrated readout of neural function, disease progression, and drug action. PhenomicsAtlas turns long-duration behavior video into computable, comparable phenotypic data — infrastructure for brain science and CNS drug discovery.

Traditional ethology

  • Relies on a few predefined metrics
  • Struggles with continuous, dynamic phenotypes
  • No unified scale for cross-model comparison

Phenotype decoding

  • AI parses long-duration behavior video
  • Standardized behavioral units and sequences
  • Builds a searchable, comparable phenotype space

Key values

01

High-dimensional features

Go beyond a few predefined metrics — extract thousands of quantifiable features from long-duration behavior video.

02

Cross-disease atlases

Systematically compare species, disease models, and interventions in a unified phenotypic embedding space.

03

AI auto-parsing

Upgrade manual observation to an automated pipeline that continuously builds searchable, reusable behavioral phenomics data.

Pipeline

From Raw Behavior to a Computable Disease Phenotype Space

Transform non-computable behavior video into searchable, comparable, and mineable disease phenotype data.

01
Raw behavior video

Raw behavior video

Multi-view, long-duration recording

02
Behavioral units

Behavioral units

AI recognition of fine actions

03
Behavioral sequences

Behavioral sequences

Parse timing and transitions

04
Phenotypic fingerprints

Phenotypic fingerprints

Aggregate high-dimensional features

05
Disease embedding space

Disease embedding space

Compare disease and intervention states

Data Resources

Multi-Species, Multi-Disease & Multimodal Phenotypic Data

Data assets keep expanding — every record retains full experimental context, acquisition conditions, and analysis lineage.

View all datasets ↗

01 Species

MouseRatTree shrewNon-human primateCanineMore species

02 Disease models

Autism spectrum disorderDepressionAlzheimer's diseaseParkinson's diseaseSpinal cord injuryAging & cognitive impairment

03 Data modalities

AI fine spontaneous behavior CORE3D pose & kinematics Behavioral sequences Cognition · metabolism · physiology COMING

Disease Models

40 Disease Models Across Major Brain Disease Phenotypes

From psychiatric and neurodegenerative conditions to pain and intervention perturbations — models are mapped into a unified behavioral phenotype space for cross-model comparison and mechanism discovery.

Capabilities

From Data Retrieval to Interactive Phenotype Exploration

Search, compare, embed, and profile — built for research and drug discovery workflows.

01

Data search

Filter quickly by species, disease, model, intervention, age, and modality.

02

Cohort comparison

Select multiple animal cohorts to compare behavioral composition, kinematics, and temporal structure.

03

Phenotype embedding

Explore distribution distances among healthy, disease, and intervention groups in high-dimensional space.

04

Individual profiles

Inspect an animal’s behavior repertoire, temporal structure, and composite phenotypic fingerprint.

Standards

Standardized, Traceable Animal Phenotype Data

From experimental design to analysis results, full context is retained so every finding can be understood, verified, and reused.

01Standardized acquisition

Record instrument models, environments, and acquisition configs

02Unified behavior ontology

Normalize behavior names, hierarchy, and definitions

03Complete experimental metadata

Retain species, strain, modeling, and intervention info

04Data quality control

Assess video, pose, recognition, and feature data quality

Next Frontier

From Behavioral Phenomics DB to Multimodal Phenomics Infrastructure

Starting from AI fine behavior, we progressively fuse cognition, metabolism, physiology, neural activity, and molecular data into a cross-scale, cross-species, computable disease phenotype atlas.

Start Exploring

Start Exploring the Behavioral Phenotype Space of Disease

Search multi-species disease-model data, compare experimental cohorts, and uncover disease signatures hidden in behavior.

Enter PhenomicsAtlas