HKEY-AI4AI-Atlas™ 1.0 uses AI-assisted dataset curation to support model selection, endpoint design and cross-species interpretation
BOSTON and SUZHOU, China, July 22, 2026 /PRNewswire/ — HKeyBio is building HKEY-AI4AI-Atlas™ 1.0, its AI-Assisted Autoimmune Model Data Atlas. The internal capability-building initiative uses AI-assisted dataset curation, standardization and retrieval to organize HKeyBio’s accumulated experience in autoimmune and allergic disease models, in vivo pharmacodynamics, pathology, immune endpoints, PK/PD, NHP responses and translational endpoints.
HKEY-AI4AI-Atlas™ 1.0 at a Glance
What it is: HKEY-AI4AI-Atlas™ 1.0 is HKeyBio’s internal autoimmune translational data-atlas capability, supported by AI-assisted dataset curation.
Problem addressed: It helps standardize model selection, endpoint stability, positive-control response ranges and cross-species interpretation experience.
Use cases: The initiative supports study design, model recommendation, endpoint selection, N-size planning, business development materials and future AI-assisted preclinical research.
HKeyBio capability link: The atlas builds on HKeyBio’s 50+ NHP autoimmune and allergy models, 500+ autoimmune and allergy animal models, and experience supporting 500+ autoimmune IND applications.
The value lies in historical model response and interpretability
In autoimmune and allergic disease research, model selection depends not only on the availability of a model, but also on baseline behavior, positive-control response ranges, endpoint variability, pathology interpretation, biomarker combinations and cross-species translational context.
HKEY-AI4AI-Atlas™ 1.0 is built with clear data boundaries
1. Eligible data scope
May include HKeyBio-owned model validation data, internal methodology data, publicly available literature-derived data, authorized and de-identified data, positive-control mechanisms and model baseline data.
2. Excluded data scope
Will not include confidential client project raw data, unauthorized client drug results, identifiable client project information, data that could reveal client pipelines, or data not approved for AI-related use.
3. AI-assisted dataset curation, standardized templates and scoring
Subject to human review, AI may assist with field extraction, terminology mapping, duplicate detection and preliminary labeling across model materials. Standardized datasets may cover disease area, model name, species, induction method, study duration, recommended group size, positive control, primary endpoint, secondary endpoint, pathology endpoint, PK/PD feasibility, translational relevance and best-use scenario.
The atlas may support future market-facing knowledge assets
Externally, the atlas will not disclose raw data, but selected insights may be converted into white papers, model selection guides, LinkedIn content, webinars, website modules, BD materials and WeChat article series.
The long-term value is not only in individual models, but in the historical response patterns, endpoint stability, pathology interpretation and biomarker experience behind those models, said the Head of Translational Medicine at HKeyBio.
About HKeyBio
HKeyBio is a preclinical CRO specializing in autoimmune and allergic diseases, providing full-process in vivo pharmacodynamic services. Positioned as an autoimmune disease model expert, HKeyBio supports model selection, in vivo efficacy evaluation, biomarker analysis, pathology and translational interpretation.
For more information, please contact:
Website: www.hkeybio.com
Email: tech@hkeybio.com
