一种联合惩罚似然建模的免疫治疗预后评估系统
By using joint penalized likelihood modeling, the problems of complex heterogeneity and multi-endpoint data correlation in TMB assessment methods are solved, enabling more accurate prognostic assessment of immunotherapy, improving the accuracy and adaptability of the model, and making it suitable for personalized treatment strategies in tumor immunotherapy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
- Filing Date
- 2025-02-06
- Publication Date
- 2026-07-17
AI Technical Summary
Existing TMB assessment methods based on the total number of cell mutations fail to fully capture the inherent complex heterogeneity of tumors, resulting in inaccurate prediction of immunotherapy efficacy. Furthermore, they face challenges such as neglecting the correlation of multi-endpoint data, collinearity of clonal features, and limited sample size, which affect the accuracy of treatment strategy formulation.
We adopted a joint penalized likelihood modeling strategy based on a generalized linear mixture model (GLMM). By introducing random effects to handle multi-endpoint data integration, we used a penalized term to address collinearity among clonal mutation features, screened key variables, controlled for coefficient differences between subgroups, and integrated observational data from different immunotherapy cohorts to construct a more flexible and accurate prognostic assessment system.
It improves the accuracy and interpretability of prognostic assessment of immunotherapy, can more comprehensively reflect multidimensional treatment effects, enhances statistical power and scalability under limited sample size conditions, and provides more accurate clinical decision support.
Smart Images
Figure CN120108718B_ABST