Related Experiment Video
Updated: Dec 31, 2025
A Protocol for Using Gene Set Enrichment Analysis to Identify the Appropriate Animal Model for Translational Research
Published on: August 16, 2017
Predicting Human Clinical Outcomes Using Mouse Multi-Organ Transcriptome
Satoshi Kozawa1, Fumihiko Sagawa1, Satsuki Endo1
1Karydo TherapeutiX, Inc., Kyoto, Japan; ERATO Sato Live Bio-Forecasting Project, Kyoto, Japan; The Thomas N. Sato BioMEC-X Laboratories, Advanced Telecommunications Research Institute International, 2-2-2 Hikaridai, Seika-cho, Soraku-gun, Kyoto 619-0288, Japan.
Predicting drug success is challenging, but a new unbiased method uses mouse gene expression patterns and machine learning to accurately forecast clinical outcomes for various drug types, aiding drug development.
Area of Science:
- Biomedical Sciences
- Pharmacology
- Computational Biology
Background:
- High failure rates (approx. 90%) of pre-clinical drugs in clinical trials lead to significant financial losses.
- Translating pre-clinical findings into predictable clinical outcomes remains a critical challenge in drug development.
Purpose of the Study:
- To develop a modality-independent and unbiased approach for predicting drug clinical outcomes.
- To leverage multi-organ transcriptome patterns in mice and a machine learning-humanized database for outcome prediction.
Main Methods:
- Utilized multi-organ transcriptome patterns induced in mice.
- Employed a machine learning algorithm to "humanize" a mouse-transcriptome database with human clinical outcome data.
- Validated the approach across small-molecule, antibody, and peptide drugs.
Main Results:
- Successfully predicted known clinical outcomes for 5,519 adverse events and 11,312 therapeutic indications.
- Demonstrated the approach's adaptability in deducing potential molecular mechanisms of drug outcomes.
- Identified novel drug repositioning targets without requiring prior drug structural or mechanistic information.
Conclusions:
- The developed approach offers a versatile, unbiased, and effective tool for predicting drug clinical outcomes.
- This method can significantly aid the drug development process by improving prediction accuracy and identifying new therapeutic opportunities.
- The modality-independent nature allows broad applicability across diverse drug types and development stages.
