Predictive models for Kidney Graft Survival

The Challenge

Long-term graft survival after kidney transplantation remains challenging despite advances in immunosuppressive therapies. The client needed a solution to predict kidney graft rejection earlier, and accurately evaluate donor-recipient compatibility, factoring in the importance of identifying non-invasive biomarkers and risk factors for personalized treatment.

Our Solution

Agilisium developed a Survival analysis model designed to bridge transcriptomics and retrospective data for predictive insights into kidney graft longevity. This solution focuses on the following key aspects:

Risk Score Prediction

AI/ML models were developed to predict risk scores and summarize kidney graft survival across various time points.

Biomarker Definition

Identified and defined key biomarkers responsible for graft rejection, providing actionable insights into their biological roles.

‍

Donor-Recipient Compatibility

Leveraged AI/ML to recommend optimal donor-recipient pairs, reducing likelihood of rejection & improving long-term outcomes.

‍

Predictive Model Development

Time-to-Event models such as Cox proportional models were developed to identify clinical parameters and biomarkers that influence kidney graft survival across various time points.

Key Impact

Improvement

in renal graft survival and donor-recipient match prediction

The Customer

A prestigious US-based healthcare and research institution globally recognized for its pioneering work in translating discoveries into life-saving therapies.

The Outcomes

Personalized Treatment Plans

Paves the way for highly personalized treatments by stratifying patients based on individualized risk factors, improving long-term graft survival rates.

Accelerated R&D for Novel Diagnostic Testing

Accelerates identification of clinically relevant biomarkers, supporting the development of new diagnostic tests for the assessment of renal transplant rejection risk.

‍